Brendan Eich: JavaScript, Firefox, Mozilla, and Brave

Lex Fridman (01:16.880)

but in my experience, good ideas are in abundance.

Brendan Eich (01:20.740)

Good execution, on the other hand, is rare.

Lex Fridman (01:23.720)

I think some mix of good timing, good idea,

Lex Fridman (01:26.040)

and good execution is essential.

Lex Fridman (01:28.160)

Getting that mix right is tough,

Lex Fridman (01:29.840)

and Brendan, somehow, multiple times in his career,

Lex Fridman (01:32.880)

did just that.

Brendan Eich (01:34.260)

I'm starting to believe it's more art than science,

Lex Fridman (01:36.640)

like most interesting things in life.

Brendan Eich (01:39.200)

If you enjoy this thing, subscribe on YouTube,

Lex Fridman (01:41.720)

review an Apple podcast, follow on Spotify,

Brendan Eich (01:44.680)

support on Patreon, or connect with me on Twitter

Lex Fridman (01:47.480)

at Lex Friedman.

Lex Fridman (01:48.840)

And now, here's my conversation with Brendan Eich.

Lex Fridman (01:53.360)

When did you first fall in love with programming?

Brendan Eich (01:56.000)

I didn't program a lot when I was in high school,

Lex Fridman (01:58.320)

but I had a friend who had a Commodore PET,

Lex Fridman (02:01.580)

and after we saw Star Wars, he said,

Lex Fridman (02:03.160)

"'Hey, let's make a basic program

Brendan Eich (02:05.660)

"'that does the Death Star Trench run.'"

Lex Fridman (02:08.040)

And it was just simple 2D graphics,

Lex Fridman (02:10.960)

and I didn't know what I was doing,

Lex Fridman (02:11.840)

so I just helped him out on the math and stuff like that.

Brendan Eich (02:14.320)

I was a math and science kid.

Lex Fridman (02:16.120)

I was really into the HP calculators of the early mid-'70s.

Brendan Eich (02:20.120)

These were the RPN.

Lex Fridman (02:21.600)

They were really strongly built,

Lex Fridman (02:23.440)

and as Arik Goldfinger said of gold, divinely heavy.

Lex Fridman (02:27.520)

There's probably some gold in them, too, gold metalization.

Lex Fridman (02:29.880)

But they were awesome calculators,

Lex Fridman (02:31.300)

and they had all the scientific functions,

Lex Fridman (02:33.160)

so I was really into that.

Lex Fridman (02:35.420)

So I aimed toward physics.

Brendan Eich (02:38.720)

I was a little late for the, I think,

Lex Fridman (02:40.640)

the 20th century golden age,

Lex Fridman (02:42.460)

and I read a lot of science fiction,

Lex Fridman (02:43.760)

so I was like, yeah, it's on the hyperdrives

Lex Fridman (02:45.880)

and warp drives, and physics was not gonna get there quickly,

Lex Fridman (02:50.360)

and I started hacking on computers

Brendan Eich (02:52.480)

while I was studying physics as an undergraduate

Lex Fridman (02:55.040)

at Santa Clara University,

Lex Fridman (02:56.560)

and I dodged the Fortran bullet

Lex Fridman (03:00.120)

because I was in the science department

Brendan Eich (03:02.040)

instead of the engineering department,

Lex Fridman (03:03.140)

where they still did Fortran card decks.

Brendan Eich (03:05.720)

I think they had an auto collator.

Lex Fridman (03:07.520)

But we were using Pascal,

Lex Fridman (03:09.320)

and I got one of the first portable C Compilers ports

Lex Fridman (03:15.200)

to the DEC minicomputers we were using,

Lex Fridman (03:17.760)

and I fell in love with programming

Lex Fridman (03:19.320)

just based on procedural abstraction, Pascal,

Brendan Eich (03:25.880)

just what now would be considered old school,

Lex Fridman (03:28.880)

like structured programming from the 70s.

Brendan Eich (03:31.600)

Niklaus Wirth, the creator of Pascal,

Lex Fridman (03:33.480)

was a good writer and a good pedagogue, right?

Brendan Eich (03:36.160)

He always at ETH would do these courses

Lex Fridman (03:38.220)

where it's like build your own computer,

Brendan Eich (03:40.120)

build your own compiler, build your own operating system.

Lex Fridman (03:42.720)

It was scratch. Yeah, kind of.

Lex Fridman (03:44.220)

And I know some people who were grad students under him

Lex Fridman (03:47.120)

and said he would torture the students

Brendan Eich (03:50.480)

with things like this custom email system

Lex Fridman (03:52.260)

that had 25 word limit and things like that.

Brendan Eich (03:56.820)

I unfortunately dodged both the Pascal

Lex Fridman (03:59.240)

and the Fortran bullets.

Lex Fridman (04:00.740)

Could you maybe linger on the Pascal?

Lex Fridman (04:04.760)

What kind of programming language was it?

Lex Fridman (04:06.320)

What is it reminiscent of today?

Lex Fridman (04:08.160)

Because it sounds like it may have had an impact

Brendan Eich (04:11.440)

on your own trajectory.

Lex Fridman (04:12.600)

Yeah, it was in the ALGOL family,

Lex Fridman (04:14.460)

and ALGOL was the big successful language design

Lex Fridman (04:20.960)

and compiler project in the 60s.

Brendan Eich (04:23.160)

It had a successor called ALGOL 68,

Lex Fridman (04:24.960)

which was ambitious but not as successful.

Lex Fridman (04:27.440)

But Pascal was kind of wordy procedures

Lex Fridman (04:31.000)

and functions language.

Brendan Eich (04:32.580)

It distinguished between functions

Lex Fridman (04:33.920)

which return a value and procedures which don't,

Brendan Eich (04:36.160)

which just compute.

Lex Fridman (04:37.880)

And you could say that whole ALGOL family went into ADA.

Brendan Eich (04:42.580)

Pascal had a second life thanks to Borland

Lex Fridman (04:44.560)

with Turbo Pascal, which was hugely successful.

Brendan Eich (04:48.920)

I think in large part due to Anders Helberg,

Lex Fridman (04:51.000)

who then went to Microsoft and did C Sharp and done that

Brendan Eich (04:53.820)

with his team there and has done really well

Lex Fridman (04:55.520)

doing TypeScript, type JavaScript.

Lex Fridman (04:58.000)

So yeah, there's a lineage here.

Lex Fridman (05:00.800)

But I was also interested in C and Unix

Brendan Eich (05:03.800)

by the time I was an undergrad

Lex Fridman (05:05.680)

because people were bringing Unix up

Brendan Eich (05:07.840)

on all sorts of hardware.

Lex Fridman (05:09.240)

I had some friends who were doing

Brendan Eich (05:10.200)

their own wirewrapped computers, 6820 maybe.

Lex Fridman (05:14.480)

And I was wirewrapping for my engineering course,

Brendan Eich (05:19.160)

6809 or something simpler, building a computer on a board.

Lex Fridman (05:22.480)

And I wanted to build a more ambitious one

Lex Fridman (05:24.720)

and port Unix to it, but I picked the wrong processor.

Lex Fridman (05:26.960)

I picked the National Semiconductor NS16032,

Brendan Eich (05:30.960)

which was this amazing, you know,

Lex Fridman (05:33.440)

CISC, complex instruction set computer,

Lex Fridman (05:37.200)

and not the reduced instruction set computers

Lex Fridman (05:39.360)

that were just being contemplated into the mid eighties.

Lex Fridman (05:43.740)

And RISC ultimately went out.

Lex Fridman (05:45.440)

RISC won in some ways, it dissolved into,

Brendan Eich (05:48.360)

you have both now, you have these super scalar architectures

Lex Fridman (05:50.800)

where like Intel has kept probably too much

Brendan Eich (05:54.160)

backward compatibility at the instruction level,

Lex Fridman (05:56.800)

but that's just, there's a front end that parses that

Brendan Eich (05:58.920)

into these, you know, these wide internal instructions.

Lex Fridman (06:01.240)

So, you know, the very long instruction word research

Brendan Eich (06:04.440)

that was also interesting at the time

Lex Fridman (06:06.900)

kind of became the micro architecture

Brendan Eich (06:08.560)

inside the backward compatible Intel.

Lex Fridman (06:11.440)

But I picked a National Semi chip

Lex Fridman (06:13.160)

and it never got made successfully.

Lex Fridman (06:14.840)

It was full of bugs and I never could have brought it up.

Lex Fridman (06:17.800)

But I went on out of physics after three years

Lex Fridman (06:21.160)

into math, computer science.

Lex Fridman (06:23.240)

And like I said, I did it because I saw,

Lex Fridman (06:25.740)

I was being sort of childlike and naive about physics.

Lex Fridman (06:28.420)

And I thought, meanwhile, the Valley is go go for computers.

Lex Fridman (06:32.120)

The Apple II, right?

Brendan Eich (06:33.700)

The PC, the Intel 8086, 8088 based PC,

Lex Fridman (06:40.720)

the IBM, you know, gave Microsoft the future for,

Brendan Eich (06:45.040)

you know, somewhat fishy deal.

Lex Fridman (06:46.800)

So it was wide open in the computing space,

Lex Fridman (06:48.640)

but in physics, you were as optimistic about physics as?

Lex Fridman (06:52.200)

No, I mean, I was one of three brothers

Brendan Eich (06:54.680)

who were all in the same grade.

Lex Fridman (06:56.060)

I have a twin and a younger brother who skipped second grade

Lex Fridman (07:00.920)

and was with us the whole time after that.

Lex Fridman (07:02.980)

And, you know, he went on,

Brendan Eich (07:04.480)

he actually studied under Kip Thorne at Caltech.

Lex Fridman (07:07.520)

But he also didn't, he ended up in software.

Brendan Eich (07:09.560)

He didn't talk about physics stuff.

Lex Fridman (07:12.340)

Does it make you sad that theoretical physics,

Brendan Eich (07:16.080)

even with string theory,

Lex Fridman (07:17.120)

hasn't really had any foundational breakthroughs

Lex Fridman (07:20.920)

in the latter part of the 20th century?

Lex Fridman (07:23.320)

Yeah, in fact, I'd say the problem is theory

Brendan Eich (07:25.220)

over experiment.

Lex Fridman (07:27.120)

I would say, you know, we need more Aristotle

Lex Fridman (07:29.920)

and less Plato.

Lex Fridman (07:31.240)

You know, mathematics is not all physical.

Brendan Eich (07:33.800)

There are lots of mathematics that cannot be realized

Lex Fridman (07:36.280)

as far as I know in this world.

Lex Fridman (07:38.360)

So to understand the world, you need to do experiments.

Lex Fridman (07:41.000)

You need to not just dream up inductive theories

Brendan Eich (07:45.200)

that could have lots of alternative theories

Lex Fridman (07:47.880)

competing with them, with no way to decide between them,

Brendan Eich (07:50.120)

except aesthetics, which is not a good guide in my opinion.

Lex Fridman (07:53.000)

Yeah, I don't know if you are friends

Brendan Eich (07:54.760)

or have a relationship with Elon Musk.

Lex Fridman (07:56.880)

Where's the, in terms of like what you would love to see

Brendan Eich (08:00.520)

our society investing in, building up,

Lex Fridman (08:02.800)

is it closer to Elon or is it closer to Feynman

Lex Fridman (08:07.080)

and Einstein and those?

Lex Fridman (08:09.120)

Well, those gentlemen are no longer with us

Lex Fridman (08:10.600)

and I think that's noticed.

Lex Fridman (08:12.000)

So like I said, the real glory days of physics,

Brendan Eich (08:14.480)

the famous pictures from Germany before the second war

Lex Fridman (08:17.280)

were just a fantastic assembly of brains,

Brendan Eich (08:21.760)

Schrodinger and Einstein.

Lex Fridman (08:23.120)

And physics, I think, took a wrong turn

Brendan Eich (08:26.320)

that maybe all of, I would say, Western science

Lex Fridman (08:30.040)

took in going for models over reality, right?

Brendan Eich (08:33.760)

You see this in all sorts of fields.

Lex Fridman (08:35.320)

Now, we can build models that are very predictive

Lex Fridman (08:38.000)

and generative and then we build actual devices

Lex Fridman (08:40.320)

or semiconductors, things like that.

Brendan Eich (08:42.200)

That's good, I'm not dismissing that.

Lex Fridman (08:44.440)

We need good models, we need to experiment

Lex Fridman (08:46.880)

and prove them and test them.

Lex Fridman (08:49.680)

But the problem I've seen in physics,

Brendan Eich (08:51.920)

which you see certainly in economics, the dismal science,

Lex Fridman (08:55.640)

and you see surprisingly in other so called hard sciences

Brendan Eich (08:59.080)

is models that don't really have to be

Lex Fridman (09:04.320)

tested against reality.

Brendan Eich (09:05.440)

They can instead become policy tools

Lex Fridman (09:07.640)

or they can become, like I said,

Brendan Eich (09:09.920)

one of a large family of alternate theories

Lex Fridman (09:12.960)

that could be as predictive

Lex Fridman (09:14.640)

but nobody's doing the winnowing out.

Lex Fridman (09:17.000)

That's such an interesting tension in society.

Brendan Eich (09:18.920)

You see this in even the software scientists

Lex Fridman (09:21.000)

which have a deep love for psychology.

Brendan Eich (09:24.440)

You see this in epidemiology, not with the virus.

Lex Fridman (09:27.200)

Absolutely.

Brendan Eich (09:28.040)

There's this tension of how much of the world

Lex Fridman (09:31.280)

can we understand through just a beautifully fit model?

Lex Fridman (09:35.040)

And then at the same time, my main work

Lex Fridman (09:37.640)

is in machine learning where it's like

Brendan Eich (09:41.000)

there is no provable thing usually.

Lex Fridman (09:44.400)

It's all about just getting the right data set

Lex Fridman (09:47.760)

and getting tricks and so on.

Lex Fridman (09:49.640)

And there's this tension even in my own soul

Brendan Eich (09:51.760)

of I grew up on theoretical computer science.

Lex Fridman (09:55.680)

I loved approximation algorithms,

Brendan Eich (09:58.920)

all of that different complexity classes,

Lex Fridman (10:01.920)

just those little puzzles.

Brendan Eich (10:03.960)

I mean, I don't know, to you as somebody

Lex Fridman (10:06.200)

who was in math and computer science

Lex Fridman (10:08.600)

and then ended up going into places

Lex Fridman (10:11.280)

where you engineer some of the most impactful things

Brendan Eich (10:13.880)

in this world, do you see the P versus NP,

Lex Fridman (10:18.680)

all that whole space is interesting at all?

Brendan Eich (10:21.280)

Yeah, it's not that useful in practice.

Lex Fridman (10:24.640)

People are using it with sort of crypto analysis

Brendan Eich (10:27.920)

or asymptotic arguments about can we have

Lex Fridman (10:31.760)

a quantum resistant crypto algorithm, things like that,

Lex Fridman (10:35.320)

which may not be practical, right?

Lex Fridman (10:36.560)

If you follow Mikhail Diakonov or Gil Kalai,

Brendan Eich (10:40.400)

there are big questions about how quantum computing

Lex Fridman (10:43.640)

will scale up, how practical it will be.

Lex Fridman (10:46.800)

Is that something that you think about quantum computing?

Lex Fridman (10:49.880)

Not except for spare time.

Brendan Eich (10:51.080)

Like you said, I'm not using this kind of computer science

Lex Fridman (10:53.560)

in practice because almost everything now is engineering

Lex Fridman (10:57.680)

and finding ways to get computers

Lex Fridman (11:02.160)

to be more useful for people,

Brendan Eich (11:03.320)

which goes from design problems,

Lex Fridman (11:06.160)

which are really kind of an art.

Brendan Eich (11:07.440)

Like you said, anything you can't automate is an art.

Lex Fridman (11:09.960)

Well, we can have machine learning compose music

Lex Fridman (11:13.360)

and it can imitate, you can train it,

Lex Fridman (11:14.800)

and it can sound kind of decent,

Lex Fridman (11:16.040)

but maybe lacking that je ne sais quoi.

Lex Fridman (11:18.680)

But user interface still, I think, requires human art.

Lex Fridman (11:23.560)

So speaking of things that didn't follow

Lex Fridman (11:26.440)

a perfect theory and model, JavaScript,

Lex Fridman (11:30.480)

so there's two things.

Lex Fridman (11:31.320)

One, it had an impact on the world at a huge scale,

Brendan Eich (11:34.360)

obviously, and it also still is one of probably

Lex Fridman (11:38.920)

the most popular programming language in the world.

Lex Fridman (11:41.400)

So can we go back to the origin story?

Lex Fridman (11:45.200)

Can you tell the story of how JavaScript was created?

Brendan Eich (11:47.840)

Yeah, I was at Silicon Graphics

Lex Fridman (11:49.960)

after graduate school for seven years,

Lex Fridman (11:53.080)

and it got to be big and successful

Lex Fridman (11:54.720)

and divisionalized and political,

Lex Fridman (11:56.800)

and I thought kind of boring.

Lex Fridman (12:00.480)

And a friend who'd been there went to one of the last

Brendan Eich (12:03.960)

of the super companies, the super startups

Lex Fridman (12:06.040)

in the early 90s.

Brendan Eich (12:07.640)

There were several.

Lex Fridman (12:08.720)

I suppose General Magic was a little after that

Brendan Eich (12:10.840)

around the same time.

Lex Fridman (12:11.920)

But Micro Unity was that company that I went to,

Lex Fridman (12:14.360)

and it was because my friend Jeff Weinstein

Lex Fridman (12:16.880)

had gone there from Silicon Graphics.

Brendan Eich (12:18.040)

He recruited me, and Micro Unity was doing everything.

Lex Fridman (12:21.960)

So this was like the ultimate sort of pretend grad school.

Brendan Eich (12:26.080)

It was doing a new fab, new semiconductor process.

Lex Fridman (12:29.360)

It was doing new analog and digital circuits

Brendan Eich (12:32.600)

on the same very large but not wafer scale chip.

Lex Fridman (12:36.080)

Originally, it was five centimeters on a side.

Brendan Eich (12:40.400)

It was really hot too, so I needed a water cooler.

Lex Fridman (12:43.880)

It was a Craykiller, and then they shrunk it,

Lex Fridman (12:46.280)

and they tried to do a home sort of media processor

Lex Fridman (12:49.360)

that was essentially a barrel processor.

Lex Fridman (12:51.920)

But you could think of trying to do all the things

Lex Fridman (12:55.440)

that we now see in modern architectures

Brendan Eich (12:58.160)

with short vector instructions

Lex Fridman (13:00.240)

and sort of wide instructions or multiple issue,

Lex Fridman (13:05.120)

and doing a lot of the stuff in software

Lex Fridman (13:07.520)

because the second iteration, the set top box,

Brendan Eich (13:10.120)

was really for avoiding the cost to the cable company

Lex Fridman (13:13.040)

of rolling the trucks out

Brendan Eich (13:14.520)

to replace your garbage General Atlantic set top box

Lex Fridman (13:17.560)

with a totally newer, less garbagey one.

Lex Fridman (13:21.000)

So if you could have software gradable set top boxes,

Lex Fridman (13:23.860)

the cable companies thought they could save a lot of money

Lex Fridman (13:26.200)

and add features.

Lex Fridman (13:27.160)

Is this assembly, or which level of the software?

Brendan Eich (13:30.240)

It was like, we were writing in, we were using GCC.

Lex Fridman (13:33.800)

We were writing C++ and C.

Brendan Eich (13:36.240)

Somebody I worked with there, really very smart guy,

Lex Fridman (13:39.880)

hired from a sort of Wall Street

Brendan Eich (13:41.680)

hotshot programming consultancy,

Lex Fridman (13:45.240)

did his own hardware design as well as software.

Lex Fridman (13:47.040)

And we were working on how to make

Lex Fridman (13:49.380)

not only a short vector units,

Lex Fridman (13:50.740)

but general bit shufflers and permuters.

Lex Fridman (13:53.080)

So you could do things like crypto algorithms efficiently,

Lex Fridman (13:57.440)

and you could do demodulation of the cable,

Lex Fridman (14:01.400)

complex quadrature amplitude modulated signal.

Lex Fridman (14:05.240)

So you're basically taking A to D converters,

Lex Fridman (14:07.980)

dumping things in buffers,

Lex Fridman (14:09.000)

and then doing the rest in software.

Lex Fridman (14:10.740)

All the framing and the Reed Solomon and Viterbi

Lex Fridman (14:13.480)

and all that error correction.

Lex Fridman (14:14.480)

So that was really great learning experience,

Lex Fridman (14:16.060)

but it was not gonna work.

Lex Fridman (14:16.920)

It was doing too many risky things at once, right?

Brendan Eich (14:18.660)

If you, as Jim Clark said to me,

Lex Fridman (14:20.460)

when I hopped to Netscape after three years at MicroUnity,

Brendan Eich (14:23.200)

he said, oh yeah, you do 10 things each,

Lex Fridman (14:26.040)

one in 10 odds, it's gonna be one in 10 billion, right?

Brendan Eich (14:29.840)

The multiplication principle.

Lex Fridman (14:31.080)

So, Netscape was already a rocket,

Lex Fridman (14:33.360)

and I passed the chance to go there in 1994.

Lex Fridman (14:36.760)

I knew the founders because I worked at SGI,

Brendan Eich (14:39.200)

Clark's company.

Lex Fridman (14:40.040)

Could you pause for a second in Netscape?

Lex Fridman (14:42.200)

When was the launch of this rocket?

Lex Fridman (14:44.800)

94.

Lex Fridman (14:45.840)

94 was the launch of Netscape?

Lex Fridman (14:47.320)

And I went there in early 95 in April.

Brendan Eich (14:49.720)

Okay, so you said you missed the launch.

Lex Fridman (14:52.200)

Well, I missed the first floor employment opportunity,

Lex Fridman (14:55.560)

but the IPO was August 1995, so I was there for that.

Lex Fridman (14:58.960)

How obvious was it that Netscape was like world changing?

Lex Fridman (15:02.280)

What was the layout?

Lex Fridman (15:04.080)

Was Netscape one of the first big browsers?

Brendan Eich (15:06.160)

Yes, so when I was at MicroUnity still in 93,

Lex Fridman (15:09.280)

we saw a browser called Mosaic.

Lex Fridman (15:11.720)

And up till then, we'd used email and we'd used Usenet,

Lex Fridman (15:14.760)

the NNTP protocol, we'd use news readers, we used FTP,

Brendan Eich (15:18.680)

we used all these old internet protocols,

Lex Fridman (15:20.860)

all relying on the DNS and TCP IP and UDP for that matter.

Brendan Eich (15:25.720)

When I was at Silicon Graphics,

Lex Fridman (15:26.800)

we brought up the whole stack, right?

Brendan Eich (15:28.140)

We had to discover how to find the ethernet addresses

Lex Fridman (15:31.760)

on your network and then find IP addresses for them,

Brendan Eich (15:34.720)

ARP protocol, all that stuff.

Lex Fridman (15:35.940)

And it was great because nobody knew in the 80s

Lex Fridman (15:38.680)

what was gonna win, all the proprietary stacks

Lex Fridman (15:41.480)

like IBM, SNA and DeckNet and all these other protocols

Brendan Eich (15:44.840)

were saying, we're gonna do it

Lex Fridman (15:46.320)

or it's gonna be heterogeneous future.

Lex Fridman (15:48.080)

And instead it was Berkeley Unix and the TCP IP stack

Lex Fridman (15:52.520)

that dated back to the ARPANET that won.

Lex Fridman (15:54.900)

And I think we knew it, we all knew it at SGI,

Lex Fridman (15:57.680)

but the salespeople didn't.

Lex Fridman (15:59.280)

And so they kept trying to get multiple network stacks

Lex Fridman (16:02.080)

interoperating, but in the end it won.

Lex Fridman (16:04.360)

And so that was the internet and it was email and texty

Lex Fridman (16:08.120)

and it was used and very texty.

Lex Fridman (16:09.520)

And then Tim Berners Lee did his thing,

Lex Fridman (16:12.540)

but I don't think I was paying attention.

Lex Fridman (16:14.120)

And I think the date when he first did it

Lex Fridman (16:16.400)

or when he wrote the famous emails and pushed back to 89,

Lex Fridman (16:19.580)

but I noticed a mosaic in 93 because one of the things

Lex Fridman (16:23.160)

that Mark Andreessen and Eric Bina did at NCSA

Brendan Eich (16:26.560)

was they innovated on the early HTML standard.

Lex Fridman (16:31.320)

They in particular Mark sent this email saying,

Brendan Eich (16:33.040)

hey everybody, we think you should be able

Lex Fridman (16:34.360)

to put an image in a page.

Lex Fridman (16:36.240)

And you know when he sent that Eric Bina had already

Lex Fridman (16:38.400)

written the code.

Lex Fridman (16:39.240)

And I talked to Tim Berners Lee more recently

Lex Fridman (16:41.920)

just a few years ago and he was like,

Brendan Eich (16:43.080)

oh, we had another way of doing it and it didn't work out

Lex Fridman (16:45.880)

because Mark shipped his in mosaic.

Lex Fridman (16:48.000)

And this convinced me of several things.

Lex Fridman (16:50.040)

One, the internet meant there was a huge first mover

Brendan Eich (16:53.840)

advantage and being fast, getting on first mattered a lot.

Lex Fridman (16:58.080)

And so Richard Gabriel of scheme and poetry fame

Brendan Eich (17:01.840)

has written about this, the famous.

Lex Fridman (17:03.160)

What's poetry? Well, he's a poet.

Brendan Eich (17:06.320)

Oh, actual poetry.

Lex Fridman (17:07.160)

Is he talking about some kind of something?

Brendan Eich (17:08.360)

No, no.

Lex Fridman (17:09.200)

I mean, he's the founder of Lucid,

Brendan Eich (17:10.200)

which is where Jamie Zawinski worked before Netscape.

Lex Fridman (17:12.920)

And Lucid was doing compilers and Lucid Emacs,

Brendan Eich (17:16.320)

which was a fork of Emacs,

Lex Fridman (17:18.240)

famously Jamie fighting against Richard Stallman, Stalmax.

Lex Fridman (17:22.680)

And so Richard Gabriel, very, very brainy computer guy,

Lex Fridman (17:25.260)

but also a poet, but he wrote a nice essay

Brendan Eich (17:27.360)

that gets abused all the time.

Lex Fridman (17:28.960)

In fact, Jamie's put a kind of warning in front

Brendan Eich (17:30.760)

of his version of it on his site,

Lex Fridman (17:32.640)

JWC.org called Worse is Better.

Lex Fridman (17:34.400)

And this is about survival advantage of software

Lex Fridman (17:38.240)

in the network world, in my opinion.

Brendan Eich (17:39.780)

It's about Unix.

Lex Fridman (17:41.440)

It started out being framed as Unix and Lisp,

Brendan Eich (17:43.720)

good news, bad news, because all the Lisp people,

Lex Fridman (17:46.000)

the MIT people were like, oh, you know,

Brendan Eich (17:48.320)

the crown jewel scheme, this Faberge egg or Common Lisp,

Lex Fridman (17:52.720)

this giant cathedral, of course we're going to win.

Brendan Eich (17:54.520)

This is civilization.

Lex Fridman (17:55.740)

And those, you know, those farmers in New Jersey

Brendan Eich (17:58.040)

to borrow from the Sopranos, those picks down at Bell Labs,

Lex Fridman (18:01.320)

they're just, you know, there's nothing sound there.

Brendan Eich (18:03.400)

It's all hacking.

Lex Fridman (18:05.280)

Well, guess what won?

Brendan Eich (18:06.840)

Wow, so you're saying this is a fundamental,

Lex Fridman (18:09.960)

like principle of the internet is moving fast wins.

Brendan Eich (18:14.800)

You could say in almost any network system,

Lex Fridman (18:16.960)

like in biological evolution,

Brendan Eich (18:18.260)

you see successful alleles sweep populations

Lex Fridman (18:20.720)

and they don't always have, you know,

Brendan Eich (18:23.440)

they aren't free of flaws.

Lex Fridman (18:24.480)

They're heterozygous advantage, right?

Brendan Eich (18:26.360)

You can get both parents give you the gene variant

Lex Fridman (18:30.200)

and you get sickle cell anemia, right?

Lex Fridman (18:31.800)

But if one of them does, you're more resistant to malaria.

Lex Fridman (18:34.680)

And so this isn't a beautiful process,

Brendan Eich (18:38.160)

except at large scale.

Lex Fridman (18:39.520)

And then you realize that because it moves fast

Lex Fridman (18:41.520)

and can adapt, it can win.

Lex Fridman (18:44.200)

And people still struggle with this.

Brendan Eich (18:46.600)

I used to struggle with this

Lex Fridman (18:47.880)

because JavaScript was done in such a hurry

Lex Fridman (18:50.120)

and the force of web compatibility meant

Lex Fridman (18:52.980)

early mistakes couldn't be fixed.

Lex Fridman (18:55.240)

And even the standards process injected new mistakes,

Lex Fridman (18:57.680)

as it will.

Lex Fridman (18:59.040)

But often standards bodies go back

Lex Fridman (19:00.400)

and making compatible changes.

Brendan Eich (19:01.680)

You can't do that with the web.

Lex Fridman (19:03.080)

It's more like, again, like biology,

Brendan Eich (19:05.200)

you preserve what still works.

Lex Fridman (19:07.120)

You don't want to break ATP metabolism or whatever.

Lex Fridman (19:10.360)

So you have to kind of resign yourself to the reality

Lex Fridman (19:15.140)

of worse is better being enshrined

Brendan Eich (19:20.000)

in actual design points you might not like.

Lex Fridman (19:23.120)

And that happened with JavaScript and I'm way over it,

Lex Fridman (19:25.780)

but it also, I think was a huge advantage.

Lex Fridman (19:28.880)

That's why JavaScript has kind of swept

Brendan Eich (19:31.160)

a lot of programming domains.

Lex Fridman (19:33.400)

People will say, oh, it's not because of merit.

Brendan Eich (19:34.840)

Well, you're right.

Lex Fridman (19:35.660)

But we also improved it over time in the standards body.

Brendan Eich (19:37.600)

I spent 20 years doing that.

Lex Fridman (19:39.320)

And you don't get that choice.

Brendan Eich (19:41.760)

It's like, I'm not saying that that was the best language.

Lex Fridman (19:45.360)

I'm just saying that was the right time to do it.

Lex Fridman (19:47.360)

And I like to say the alternative was not to do it.

Lex Fridman (19:50.280)

I could have told Netscape, I can't do this.

Brendan Eich (19:52.180)

It's too rushed.

Lex Fridman (19:53.200)

And it would have been visual basic script.

Lex Fridman (19:55.260)

And it would have been bad.

Lex Fridman (19:57.180)

So that's a good way to present the alternative.

Lex Fridman (19:59.780)

But so it was a Netscape and you have written it

Lex Fridman (1:00:00.520)

So this is the engineering mindset versus the theoretician.

Lex Fridman (1:00:03.500)

So you didn't want to create the perfect language,

Lex Fridman (1:00:05.120)

but one that was the popular and shipped

Lex Fridman (1:00:06.940)

and all that kind of stuff.

Lex Fridman (1:00:07.780)

And you could say there was,

Brendan Eich (1:00:08.620)

I was standing on the shoulders of giants.

Lex Fridman (1:00:09.940)

So there was a staged process

Brendan Eich (1:00:11.220)

where I had to hold back things

Lex Fridman (1:00:12.820)

that were well designed by others in other languages

Lex Fridman (1:00:15.900)

and I could imitate,

Lex Fridman (1:00:17.260)

but I couldn't do them all in the 10 days.

Lex Fridman (1:00:18.540)

So they came in 1996 and 97,

Lex Fridman (1:00:22.820)

and they came into the third edition of the standard,

Brendan Eich (1:00:25.140)

which was final finalized in 1999.

Lex Fridman (1:00:27.680)

But at that point,

Brendan Eich (1:00:28.520)

Netscape had been sold to AOL

Lex Fridman (1:00:31.220)

which was a decent exit considering

Lex Fridman (1:00:33.140)

and had previously been mercilessly crushed.

Lex Fridman (1:00:37.380)

Netscape was selling the browser along with server software

Brendan Eich (1:00:40.700)

that it had acquired after its IPO

Lex Fridman (1:00:42.740)

and Microsoft was just underpricing it.

Lex Fridman (1:00:44.560)

So there was no way to compete with that.

Lex Fridman (1:00:47.500)

Microsoft was also making Internet Explorer

Brendan Eich (1:00:51.420)

the default browser in Windows,

Lex Fridman (1:00:53.140)

which is called tying and antitrust law.

Lex Fridman (1:00:55.580)

And they were doing even more brutal things.

Lex Fridman (1:00:57.140)

There was a famous investor.

Brendan Eich (1:00:59.260)

He did very well on Google.

Lex Fridman (1:01:00.340)

So he's a billionaire, Ramshree Ram,

Lex Fridman (1:01:01.780)

and he was sales guy or head of sales at Netscape.

Lex Fridman (1:01:04.700)

And he got off the phone looking ashen faced

Brendan Eich (1:01:07.920)

after a compact called and said,

Lex Fridman (1:01:10.540)

Microsoft just told us they're gonna pull

Brendan Eich (1:01:12.060)

our Windows license if we ship Netscape

Lex Fridman (1:01:13.860)

as the default browser.

Brendan Eich (1:01:15.100)

Wow.

Lex Fridman (1:01:15.940)

So there is some bullying going on.

Brendan Eich (1:01:17.460)

It was totally immaterial in the antitrust case.

Lex Fridman (1:01:19.720)

But JavaScript escaped into the standard setting

Brendan Eich (1:01:23.060)

where there was fairly good cooperation.

Lex Fridman (1:01:24.900)

Microsoft had a really good guy on it

Lex Fridman (1:01:26.860)

and Guy Steele was there for a time

Lex Fridman (1:01:29.420)

and there was some good work.

Lex Fridman (1:01:30.900)

But after the antitrust case

Lex Fridman (1:01:33.460)

and Netscape kind of dissolving into AOL

Lex Fridman (1:01:37.500)

and not really going anywhere quickly,

Lex Fridman (1:01:38.940)

Mozilla took years to really bring up,

Brendan Eich (1:01:42.380)

the standard froze.

Lex Fridman (1:01:43.400)

And by 2003, even though they'd been sort of noodling around

Brendan Eich (1:01:46.300)

with advanced versions, JavaScript 2,

Lex Fridman (1:01:48.900)

I'd given the keys to the kingdom

Brendan Eich (1:01:50.300)

to another MIT grad, Baltimore Horwatt.

Lex Fridman (1:01:53.060)

Very big brain and still at Google, I think.

Brendan Eich (1:01:54.860)

He won the Putnam in 86

Lex Fridman (1:01:56.700)

and he's very mathematical.

Brendan Eich (1:01:58.380)

Legit.

Lex Fridman (1:01:59.640)

He designed this successor language, JavaScript 2,

Lex Fridman (1:02:02.980)

but it only showed up in mutated form

Lex Fridman (1:02:06.620)

in Microsoft's ASP.NET server side

Lex Fridman (1:02:09.560)

and it didn't last there.

Lex Fridman (1:02:11.100)

And it showed up in Flash

Lex Fridman (1:02:12.340)

and that's what became ActionScript 3.

Lex Fridman (1:02:14.220)

Ah, ActionScript, interesting.

Lex Fridman (1:02:16.860)

And then Flash of course declined.

Lex Fridman (1:02:18.740)

And so how did we arrive at ES6

Brendan Eich (1:02:21.300)

where it's like there's so many,

Lex Fridman (1:02:24.120)

where everyone, okay, there's this history of JavaScript

Brendan Eich (1:02:27.220)

that people were, it was just like cool

Lex Fridman (1:02:29.220)

when you're like having beers to talk crap about JavaScript.

Brendan Eich (1:02:32.540)

Everyone loves to hate,

Lex Fridman (1:02:33.900)

like people who are married say, ah, marriage sucks,

Brendan Eich (1:02:36.460)

is they just wanna get, let off some steam

Lex Fridman (1:02:38.900)

even though everyone uses the language.

Lex Fridman (1:02:40.900)

But ES6, it's become this like reputable,

Lex Fridman (1:02:45.940)

like it fixed major pain points, I think.

Brendan Eich (1:02:49.180)

It added things to the language

Lex Fridman (1:02:50.420)

and added something that was already ES5 strict mode,

Lex Fridman (1:02:53.580)

but made it implicit in class bodies and module bodies.

Lex Fridman (1:02:57.260)

It was a big jump,

Lex Fridman (1:02:58.500)

but it accumulated some of the ES4 designs

Lex Fridman (1:03:01.300)

that we'd done with Adobe

Brendan Eich (1:03:03.640)

for what we hoped would be the fourth edition of ECMAScript

Lex Fridman (1:03:07.660)

that were supposed to fold in some of these old JavaScript 2

Brendan Eich (1:03:10.740)

ideas that had come into ActionScript 3.

Lex Fridman (1:03:13.060)

So you look at the family tree and you see these forks

Lex Fridman (1:03:15.220)

and the main ones are the ones that go into Adobe Flash

Lex Fridman (1:03:20.700)

acquired from Macromedia

Lex Fridman (1:03:21.900)

and the one that went into the service side

Lex Fridman (1:03:23.580)

of Microsoft's stack, which kind of died.

Lex Fridman (1:03:27.100)

And then trying to bring them back into the standard

Lex Fridman (1:03:28.900)

and not quite succeeding, ES4 was mothballed,

Lex Fridman (1:03:32.300)

but all the good parts that everyone liked made it into ES6.

Lex Fridman (1:03:35.700)

And so that was a success.

Lex Fridman (1:03:37.060)

And I said earlier, I had the wrong year,

Lex Fridman (1:03:38.380)

I think it's 2015, so it's off by.

Brendan Eich (1:03:40.380)

Four, ES6.

Lex Fridman (1:03:42.420)

Yeah, it was finalized in 2015.

Brendan Eich (1:03:44.300)

It took a little longer than we hoped,

Lex Fridman (1:03:45.620)

but because ES5 was 2009

Lex Fridman (1:03:48.580)

and that was a smaller increment from ES3.

Lex Fridman (1:03:51.580)

We skipped four again, we mothballed it.

Lex Fridman (1:03:53.500)

And we had a split in the committee where some people said,

Lex Fridman (1:03:56.620)

ES4 is too big, we're gonna work on incremental improvements,

Brendan Eich (1:03:59.540)

no new syntax in particular, they promised.

Lex Fridman (1:04:02.580)

Not quite true, but they added a bunch of interesting APIs,

Brendan Eich (1:04:06.300)

Alan Wiersbrock, my coauthor of the Hobble paper.

Lex Fridman (1:04:09.880)

And he was at Microsoft at the time.

Brendan Eich (1:04:11.140)

I ended up hiring with Mozilla.

Lex Fridman (1:04:12.300)

He wanted to get to Mozilla and keep doing

Brendan Eich (1:04:16.620)

his sort of editor job of the JavaScript standard,

Lex Fridman (1:04:19.660)

ECMAScript.

Lex Fridman (1:04:21.300)

And when we got ES6 done, it was a little late, 2015,

Lex Fridman (1:04:25.660)

and we switched to year numbers.

Lex Fridman (1:04:27.180)

So people still call it ES6, I call it ES6.

Lex Fridman (1:04:30.100)

But if you remember, off by nine plus 2000.

Brendan Eich (1:04:33.100)

Yeah, I mean, ES6 is such a big job.

Lex Fridman (1:04:35.780)

I mean, like you said, there's a third that connects all of it,

Lex Fridman (1:04:38.300)

but ES6 is when it became this language

Lex Fridman (1:04:41.100)

that almost feels ready to take over the world completely.

Brendan Eich (1:04:45.180)

More programming and the large features,

Lex Fridman (1:04:46.660)

more features you need for larger teams.

Brendan Eich (1:04:49.380)

Software engineering.

Lex Fridman (1:04:50.300)

Microsoft did something smart, too.

Brendan Eich (1:04:52.220)

Anders and company, Luke Hoban, who's left Microsoft,

Lex Fridman (1:04:56.220)

also did TypeScript.

Lex Fridman (1:04:57.800)

And they realized something, I think,

Lex Fridman (1:05:00.400)

that Gilad Barak has also popularized,

Lex Fridman (1:05:02.760)

and he was involved in Dart at Google.

Lex Fridman (1:05:05.740)

If you, don't worry about soundness in the type system.

Brendan Eich (1:05:08.860)

You don't try to enforce the type checks

Lex Fridman (1:05:10.920)

at runtime in particular.

Brendan Eich (1:05:11.860)

Just use it as sort of a warning system,

Lex Fridman (1:05:13.460)

a tool time type system.

Brendan Eich (1:05:15.020)

You can still have a lot of value for developers,

Lex Fridman (1:05:17.660)

especially in large projects.

Lex Fridman (1:05:19.260)

So TypeScript's been a roaring success for Microsoft.

Lex Fridman (1:05:21.260)

What do you think about TypeScript?

Lex Fridman (1:05:24.100)

Is it adding confusion, or is it ultimately beneficial?

Lex Fridman (1:05:27.380)

I think it's beneficial.

Brendan Eich (1:05:28.220)

Now, it's technically a superset of JavaScript,

Lex Fridman (1:05:30.920)

so of course I love it, right?

Brendan Eich (1:05:33.380)

The shortest JavaScript program

Lex Fridman (1:05:35.540)

is still a TypeScript program.

Brendan Eich (1:05:36.660)

Any JavaScript program is a TypeScript program,

Lex Fridman (1:05:38.540)

which is brilliant,

Brendan Eich (1:05:39.380)

because then you can start incrementally adding

Lex Fridman (1:05:40.860)

type annotations, getting warnings,

Brendan Eich (1:05:43.020)

learning how to use them.

Lex Fridman (1:05:44.420)

Microsoft's had to kind of look around corners

Brendan Eich (1:05:47.180)

at the standards body and guess how their version of modules

Lex Fridman (1:05:50.780)

or decorators should work.

Lex Fridman (1:05:52.500)

And the standards body then may change things a bit.

Lex Fridman (1:05:55.380)

So I think they're obligated with TypeScript

Brendan Eich (1:05:57.620)

either to carry their own version

Lex Fridman (1:05:59.340)

or to bring it back with incompatible changes

Brendan Eich (1:06:01.680)

towards the standard over time.

Lex Fridman (1:06:03.100)

And I think they've played generally fair there.

Brendan Eich (1:06:05.420)

There's some sentiment that,

Lex Fridman (1:06:06.460)

why don't they standardize TypeScript?

Brendan Eich (1:06:07.980)

Well, they've been clear they don't want to.

Lex Fridman (1:06:09.920)

They have a proprietary investment, it's valuable.

Brendan Eich (1:06:11.880)

They have control of the ball.

Lex Fridman (1:06:13.420)

And in some ways, you can say the same thing

Brendan Eich (1:06:15.820)

to any of the other big companies in the standards body.

Lex Fridman (1:06:18.100)

Why doesn't Google standardize its stuff?

Lex Fridman (1:06:20.580)

So you think it'll continue being

Lex Fridman (1:06:22.260)

like a kind of dance partner to JavaScript,

Lex Fridman (1:06:24.620)

to the base JavaScript?

Lex Fridman (1:06:25.860)

There's a hope that at some point,

Brendan Eich (1:06:27.620)

if they keep reconverging it and the standard

Lex Fridman (1:06:29.740)

doesn't break them and goes in a good direction,

Brendan Eich (1:06:32.020)

we will get at least the annotation syntax

Lex Fridman (1:06:34.780)

and some semantics around them.

Brendan Eich (1:06:37.020)

Because when you're talking about type annotations,

Lex Fridman (1:06:39.020)

they're generally on parameters and return values

Lex Fridman (1:06:42.340)

and variable declarations.

Lex Fridman (1:06:43.940)

They're cast operators.

Brendan Eich (1:06:45.540)

You want that syntax to be reserved

Lex Fridman (1:06:47.260)

and you want it to work the same in all engines.

Lex Fridman (1:06:49.980)

And this is where ideas like Gilad's

Lex Fridman (1:06:51.540)

pluggable type systems might be good,

Brendan Eich (1:06:54.240)

though then you could create the same problem

Lex Fridman (1:06:55.820)

you have with Lisp and Scheme,

Brendan Eich (1:06:56.960)

where there's a bunch of macro libraries

Lex Fridman (1:06:58.540)

and they don't agree and you have conflicts between them.

Lex Fridman (1:07:01.700)

But pluggable type systems could be one way to standardize.

Lex Fridman (1:07:04.940)

What do you think about the giant ecosystem

Lex Fridman (1:07:07.060)

of frameworks in JavaScript?

Lex Fridman (1:07:09.820)

It feels like, because, I mean,

Brendan Eich (1:07:12.860)

this is a side effect of how many people use JavaScript,

Lex Fridman (1:07:16.060)

a lot of entrepreneurial spirit

Brendan Eich (1:07:19.220)

create their own JavaScript frameworks

Lex Fridman (1:07:22.300)

and they're actually awesome in all different ways.

Lex Fridman (1:07:27.260)

And this is an interesting question

Lex Fridman (1:07:31.220)

about almost like philosophically

Brendan Eich (1:07:33.180)

about biological system and evolution,

Lex Fridman (1:07:35.260)

all that kind of stuff.

Lex Fridman (1:07:36.140)

Do you see that as good or should it like,

Lex Fridman (1:07:38.600)

should some of them die out quicker?

Brendan Eich (1:07:40.380)

I think maybe they should, now jQuery was a very clever

Lex Fridman (1:07:44.480)

thing, John Resig made this library

Brendan Eich (1:07:45.980)

that was sort of query and do

Lex Fridman (1:07:47.600)

and blended sort of CSS selector syntax

Brendan Eich (1:07:49.980)

with JavaScript sort of object graph or DOM querying

Lex Fridman (1:07:53.020)

and made it very easy for people to do things

Brendan Eich (1:07:54.940)

almost like they were learning jQuery as its own language,

Lex Fridman (1:07:57.660)

domain specific language.

Lex Fridman (1:07:58.980)

And that I think reflected in part the difficulty

Lex Fridman (1:08:03.220)

of using the document object model,

Brendan Eich (1:08:04.960)

these APIs that were originally designed in the 90s

Lex Fridman (1:08:07.020)

for Java as well as JavaScript.

Brendan Eich (1:08:08.460)

They're very object oriented or even procedural.

Lex Fridman (1:08:11.740)

They're very kind of verbose.

Lex Fridman (1:08:13.360)

And it took like a constructor call

Lex Fridman (1:08:15.420)

and three different, you know, hokey pokey dances

Brendan Eich (1:08:17.820)

to do something where as in jQuery, it's just one line.

Lex Fridman (1:08:20.500)

Right, so that fed back finally into the standards.

Brendan Eich (1:08:24.500)

It didn't mean we standardized jQuery.

Lex Fridman (1:08:25.940)

It wasn't quite that concise,

Lex Fridman (1:08:27.140)

but you find now with the modern standards

Lex Fridman (1:08:29.660)

that we were working on in the HTML5 sort of effort

Brendan Eich (1:08:34.780)

that things became simpler, the fetch API

Lex Fridman (1:08:37.100)

and the query selector API, document.querySelector.

Brendan Eich (1:08:40.140)

A lot of things can be done now in raw JavaScript

Lex Fridman (1:08:42.540)

that you would make more concise and terse in jQuery,

Lex Fridman (1:08:46.660)

but it's not bad.

Lex Fridman (1:08:47.980)

It's pretty good.

Brendan Eich (1:08:48.820)

Whereas in the old DOM of 15 years ago,

Lex Fridman (1:08:51.100)

it was just too verbose.

Lex Fridman (1:08:51.940)

So maybe the frameworks were born kind of

Lex Fridman (1:08:55.560)

because JavaScript lacks some of the features of jQuery.

Lex Fridman (1:08:59.020)

And so like now that JavaScript is swallowing

Lex Fridman (1:09:02.700)

what jQuery was, then the frameworks will,

Brendan Eich (1:09:05.220)

only the ones that truly add value will stick around

Lex Fridman (1:09:07.840)

and the other ones will die out.

Lex Fridman (1:09:08.980)

And that highlights this also this division

Lex Fridman (1:09:10.700)

between the core language JavaScript,

Brendan Eich (1:09:12.300)

which can show up in other places

Lex Fridman (1:09:13.580)

like Node.js on the server side

Lex Fridman (1:09:15.680)

and the browser specific APIs

Lex Fridman (1:09:17.660)

or the document object model APIs,

Brendan Eich (1:09:19.720)

which are even managed by the W3C,

Lex Fridman (1:09:21.580)

the standards body that was off in XML La La Land

Brendan Eich (1:09:24.460)

when we were doing real JavaScript standards in ECMA.

Lex Fridman (1:09:26.940)

And you have this division of labor,

Brendan Eich (1:09:30.260)

division of responsibility and division of style

Lex Fridman (1:09:33.260)

and sort of aesthetics and also speed.

Lex Fridman (1:09:37.260)

So the document object model really stagnated

Lex Fridman (1:09:39.620)

after Microsoft kind of deinvested in the web.

Lex Fridman (1:09:43.580)

And Microsoft did something in their haste

Lex Fridman (1:09:45.860)

in the spirit of Netscape,

Brendan Eich (1:09:46.900)

doing things quickly and getting on first called DHTML.

Lex Fridman (1:09:49.780)

And some of their innovations

Brendan Eich (1:09:51.340)

that were like an alternative document object model

Lex Fridman (1:09:53.700)

didn't really get standardized until HTML5

Brendan Eich (1:09:56.860)

when we pragmatists at Opera at the time,

Lex Fridman (1:10:00.020)

Ian Hickson who went to Google,

Brendan Eich (1:10:01.720)

Apple and Mozilla said let's,

Lex Fridman (1:10:05.100)

XML is not gonna replace HTML, HTML4 is too old.

Brendan Eich (1:10:08.700)

Let's standardize HTML5 based on all this good stuff

Lex Fridman (1:10:12.140)

including that DHTML variant dynamic HTML5.

Brendan Eich (1:10:14.900)

HTML5, it feels like to me, maybe you can correct me,

Lex Fridman (1:10:18.140)

like a beautiful piece of design work.

Brendan Eich (1:10:21.140)

It was, it's not often with web stuff,

Lex Fridman (1:10:24.860)

you have this breath of just like,

Brendan Eich (1:10:27.320)

oh, whoever did this is, it just feels good.

Lex Fridman (1:10:31.740)

Is that, what are your thoughts about HTML?

Lex Fridman (1:10:33.660)

Is the, am I being too romantic?

Lex Fridman (1:10:35.860)

A little bit, a little bit.

Brendan Eich (1:10:36.780)

Are there flaws, fundamental flaws to it

Lex Fridman (1:10:38.660)

that I'm just not aware of?

Brendan Eich (1:10:40.180)

My old friend Hixie did a great job.

Lex Fridman (1:10:42.540)

He was another renegade physics student.

Lex Fridman (1:10:45.660)

And he was basically a QA guy at Opera

Lex Fridman (1:10:47.980)

but he obviously trained physics student

Lex Fridman (1:10:52.500)

and someone who could write British

Lex Fridman (1:10:54.620)

or he developed test suites

Lex Fridman (1:10:57.820)

and he started thinking about them more axiomatically.

Lex Fridman (1:11:00.920)

Now this is, this can be good

Brendan Eich (1:11:02.520)

because you can sort of systematize

Lex Fridman (1:11:03.980)

in a way that makes a better HTML

Brendan Eich (1:11:06.020)

or you can get caught in the pragmatism of saying,

Lex Fridman (1:11:08.100)

well, we have to handle all of these edge cases

Lex Fridman (1:11:09.820)

so we're just gonna have sort of a test matrix.

Lex Fridman (1:11:12.580)

And if the matrix is large,

Brendan Eich (1:11:13.980)

it will not be beautiful by many people's lights.

Lex Fridman (1:11:15.860)

Everyone likes to minimize along their preferred dimensions,

Brendan Eich (1:11:19.340)

the seven special forms and scheme or whatever.

Lex Fridman (1:11:21.660)

But reality is HTML needs to be big.

Brendan Eich (1:11:27.140)

It's kind of shambolic, it's a creative multi paradigm.

Lex Fridman (1:11:30.400)

And Hixie did a good job, I would say, with a bunch of it.

Brendan Eich (1:11:34.380)

Other people came in in the spirit of Ian Hickson

Lex Fridman (1:11:36.620)

to do HTML5 work and they've carried on that effort.

Lex Fridman (1:11:40.260)

And so it's a mix of pragmatism,

Lex Fridman (1:11:42.820)

de facto standards from the past being sort of combined

Brendan Eich (1:11:46.280)

or written down for the first time

Lex Fridman (1:11:47.580)

and then rethought in a way that has a simpler syntax,

Brendan Eich (1:11:50.300)

like the fetch API instead of XML HTTP request.

Lex Fridman (1:11:54.420)

This video too as well, it ultimately,

Brendan Eich (1:11:56.540)

it feels like maybe you can correct me,

Lex Fridman (1:11:58.100)

it feels like it was the nail in the coffin of Flash.

Brendan Eich (1:12:01.180)

Steve Jobs saying no Flash on the iPhone,

Lex Fridman (1:12:03.020)

in my opinion, was the actual sting through the heart.

Brendan Eich (1:12:05.460)

But, well, I'm not sure what trope you wanna use.

Lex Fridman (1:12:09.140)

Flash was a zombie until just this year, right?

Brendan Eich (1:12:12.060)

Or last year, I think last year

Lex Fridman (1:12:13.300)

was the end of Flash in main browsers.

Lex Fridman (1:12:16.380)

But Jobs really did the death blow.

Lex Fridman (1:12:18.280)

And yeah, you're right, we had to make HTML5 competitive.

Brendan Eich (1:12:23.020)

I still don't think we got

Lex Fridman (1:12:23.900)

that beautiful timeline animation.

Brendan Eich (1:12:26.780)

The timeline thing, so you like the time.

Lex Fridman (1:12:28.620)

I mean, me from, I used to animate

Brendan Eich (1:12:31.180)

all kinds of stuff inside Flash,

Lex Fridman (1:12:32.780)

plus there's a programming element.

Brendan Eich (1:12:34.940)

It was a little bit, I don't know if you can comment on that,

Lex Fridman (1:12:38.440)

but to me, it was a little bit like go to statement,

Brendan Eich (1:12:40.780)

like in a sense that it was a little bit too chaotic.

Lex Fridman (1:12:44.900)

Like it didn't, that OCD part of me as a programmer

Brendan Eich (1:12:48.380)

wasn't satisfied by Flash.

Lex Fridman (1:12:50.180)

It feels like there was bugs that were introduced

Brendan Eich (1:12:52.760)

through the animation process that I couldn't debug easily.

Lex Fridman (1:12:55.340)

Yes, I heard that too.

Brendan Eich (1:12:56.620)

I didn't use it,

Lex Fridman (1:12:57.460)

so I'm doing the grass is greener thing here.

Brendan Eich (1:12:59.820)

The thing I liked about the animation model

Lex Fridman (1:13:02.100)

was that it was this immutable function of time.

Lex Fridman (1:13:03.900)

So you could time warp and you could,

Lex Fridman (1:13:06.220)

if you dodged these bugs or worked carefully,

Brendan Eich (1:13:08.320)

you could really make it sing in ways

Lex Fridman (1:13:09.900)

that I think still a little challenging

Brendan Eich (1:13:11.700)

with the web animation standards.

Lex Fridman (1:13:14.620)

But, or just using raw canvas and WebGL.

Lex Fridman (1:13:18.140)

But there's so many tools now that maybe it doesn't matter.

Lex Fridman (1:13:20.740)

And yet we had to do video,

Brendan Eich (1:13:23.980)

we had to do WebGL and then evolve it.

Lex Fridman (1:13:27.820)

We had to do web audio.

Lex Fridman (1:13:30.060)

But once we did all these things that helped Flash die,

Lex Fridman (1:13:34.140)

thanks to Steve Gubbs,

Brendan Eich (1:13:35.700)

we had something that people didn't realize.

Lex Fridman (1:13:39.580)

We had that vision that Market Vision had,

Brendan Eich (1:13:41.180)

this graphics capable to the metal portable runtime.

Lex Fridman (1:13:46.620)

And we at Mozilla realized this

Lex Fridman (1:13:49.360)

and we saw JavaScript was something

Lex Fridman (1:13:52.420)

that you could compile to.

Brendan Eich (1:13:53.340)

Adobe had somebody in the Adobe Labs doing this too.

Lex Fridman (1:13:55.620)

He had a project called Alchemy.

Brendan Eich (1:13:57.220)

We had somebody who's now Google,

Lex Fridman (1:13:59.820)

Alon Zakai, who did his own LLVM based compiler

Brendan Eich (1:14:03.260)

that would take C or C++ and it would emit JavaScript.

Lex Fridman (1:14:06.140)

And you would think this is crazy.

Brendan Eich (1:14:07.660)

You're going from this sort of machine types,

Lex Fridman (1:14:09.820)

low level, controlled memory allocation language

Brendan Eich (1:14:13.100)

to this garbage collected, dynamically typed,

Lex Fridman (1:14:16.380)

high level, higher level language.

Lex Fridman (1:14:18.060)

But Alon sort of just phenomenologically carved nature

Lex Fridman (1:14:22.460)

of the joint and found the forms

Brendan Eich (1:14:23.740)

that were fast in JavaScript.

Lex Fridman (1:14:25.380)

And then with Dave Herman,

Brendan Eich (1:14:27.300)

who I'd recruited from Northeastern University,

Lex Fridman (1:14:29.580)

who was a type theorist,

Lex Fridman (1:14:31.280)

and Luke Wagner, who's still at Mozilla,

Lex Fridman (1:14:33.840)

who was the compiler guy and the JIT guy,

Brendan Eich (1:14:36.860)

they figured out how to codify what Alon had done

Lex Fridman (1:14:40.240)

into a typed subset of JavaScript called Asm.js.

Lex Fridman (1:14:44.320)

And this is a strange thing to think about

Lex Fridman (1:14:45.700)

because it doesn't have new syntax.

Brendan Eich (1:14:47.320)

The types are casts that occur in dominator positions

Lex Fridman (1:14:52.780)

in the control flow graph.

Lex Fridman (1:14:54.180)

So it's like a hack on JavaScript and it's a subset

Lex Fridman (1:14:56.740)

and it uses those bitwise operators

Brendan Eich (1:14:58.380)

that I talked about copying from Java

Lex Fridman (1:15:00.420)

to basically cast numeric types,

Brendan Eich (1:15:04.500)

which are double position flowing point, into integers.

Lex Fridman (1:15:06.940)

And so inside JavaScript, in the kernel semantics,

Brendan Eich (1:15:10.600)

are integers.

Lex Fridman (1:15:11.700)

And if you use these operators,

Brendan Eich (1:15:13.040)

if a compiler emits them in the right places,

Lex Fridman (1:15:15.600)

you can then treat them as typed values,

Brendan Eich (1:15:18.500)

typed memory locations, and you can type check your program.

Lex Fridman (1:15:22.380)

You can not only type check it, you can compile it.

Lex Fridman (1:15:24.940)

And this is all in sort of linear time, Alon.

Lex Fridman (1:15:27.480)

You can compile it to have deterministic performance.

Brendan Eich (1:15:30.420)

It doesn't touch the garbage collector.

Lex Fridman (1:15:32.580)

It calls a bunch of functions that come from the C functions

Brendan Eich (1:15:35.500)

or C++ code that you're compiling.

Lex Fridman (1:15:37.500)

And you can make the Epic Unreal Engine

Brendan Eich (1:15:40.980)

go in 30 frames a second.

Lex Fridman (1:15:43.620)

And when we did this in 2013 in the fall,

Brendan Eich (1:15:46.900)

Tim Sweeney and I didn't think it could be done quickly.

Lex Fridman (1:15:50.180)

Thought it would take years.

Lex Fridman (1:15:51.580)

And the team went to Raleigh, to Epic,

Lex Fridman (1:15:54.100)

and in four days they had Unreal Engine ported

Brendan Eich (1:15:57.460)

by pressing a compile button.

Lex Fridman (1:15:59.980)

But they had to have WebGL,

Brendan Eich (1:16:01.700)

which came from OpenGL.

Lex Fridman (1:16:03.220)

Yes, it came to OpenGL, which came from Silicon Graphics GL.

Brendan Eich (1:16:06.020)

They had to have Web Audio so they could map OpenAL,

Lex Fridman (1:16:09.460)

which was another audio library standard to Web Audio,

Brendan Eich (1:16:11.740)

which was kind of a Chrome idiosyncratic thing.

Lex Fridman (1:16:14.940)

But they could make it work.

Lex Fridman (1:16:16.300)

And they had to have Asm.js for fast C++ to JavaScript.

Lex Fridman (1:16:21.100)

And if you didn't have that fast compiler step,

Brendan Eich (1:16:25.260)

the JavaScript you'd write by hand

Lex Fridman (1:16:27.380)

trying to do an Unreal game would be too big and too slow.

Brendan Eich (1:16:29.820)

It would touch the garbage collector.

Lex Fridman (1:16:31.420)

It would not keep up with 30 frames a second

Brendan Eich (1:16:33.220)

on the hardware, 2013 hardware.

Lex Fridman (1:16:35.580)

So we demoed that at,

Brendan Eich (1:16:38.540)

this must have been fall 2012 now that I think about it.

Lex Fridman (1:16:40.420)

Because we demoed it at GDC,

Brendan Eich (1:16:42.340)

Game Developer Conference 2013.

Lex Fridman (1:16:44.420)

And people were stunned.

Brendan Eich (1:16:45.420)

That's like Unreal Engine, Unreal Tournament

Lex Fridman (1:16:47.380)

running in my browser window.

Brendan Eich (1:16:48.860)

No plugin, no Flash, no Java, no.

Lex Fridman (1:16:52.660)

So were those the early days of,

Brendan Eich (1:16:54.420)

because JavaScript now is able to run basically on par

Lex Fridman (1:16:58.340)

with a lot of the C++.

Lex Fridman (1:17:01.860)

And even before then,

Lex Fridman (1:17:02.780)

you had the fast JavaScript VMs in 2008

Brendan Eich (1:17:05.340)

when Chrome came out.

Lex Fridman (1:17:06.520)

Just before it came out, Mozilla,

Brendan Eich (1:17:08.740)

my friend Andreas Gal and I and others

Lex Fridman (1:17:11.660)

hacked out Trace Monkey, our Trace based JIT.

Brendan Eich (1:17:14.420)

The Squirrel Fish Extreme team at Apple did their JIT.

Lex Fridman (1:17:18.700)

And we were all competing

Brendan Eich (1:17:19.660)

on these crazy performance benchmarks.

Lex Fridman (1:17:21.860)

It was a little bit too much tuning to the benchmark.

Lex Fridman (1:17:23.520)

But JavaScript started getting fast

Lex Fridman (1:17:25.180)

and developers started noticing it.

Lex Fridman (1:17:26.660)

But it was still kind of its own high level language

Lex Fridman (1:17:29.740)

with garbage collection.

Brendan Eich (1:17:30.740)

The Asm.js step helped us go further

Lex Fridman (1:17:33.220)

because until we really proved the concept,

Brendan Eich (1:17:37.060)

people were still saying, well, JavaScript's okay.

Lex Fridman (1:17:39.340)

It's getting faster, thanks to V8.

Brendan Eich (1:17:40.980)

Everybody gave Google credit, especially Google.

Lex Fridman (1:17:43.300)

But we need something to kill Flash.

Brendan Eich (1:17:44.900)

Let's use the portable native client code

Lex Fridman (1:17:47.260)

that Google had acquired, native client.

Brendan Eich (1:17:50.260)

Which is a separate lineage for taking basically C code,

Lex Fridman (1:17:52.940)

compiling it into a software fault isolated container

Brendan Eich (1:17:56.260)

of some sort using some kind of virtualization technique.

Lex Fridman (1:17:59.740)

And maybe it can even be in process

Lex Fridman (1:18:01.620)

and still be memory safe, that would be awesome.

Lex Fridman (1:18:03.080)

But they ended up using process isolation too.

Lex Fridman (1:18:04.800)

And that kind of weakened it.

Lex Fridman (1:18:06.300)

And in the end, it was like portable native client,

Brendan Eich (1:18:08.540)

okay, meet the new boss, same as the old boss.

Lex Fridman (1:18:11.060)

This is the Google Flash, right?

Lex Fridman (1:18:13.260)

But when we did Asm.js and we showed Unreal Engine working,

Lex Fridman (1:18:17.180)

I think it was only a matter of time

Brendan Eich (1:18:18.660)

before Google threw in the towel.

Lex Fridman (1:18:19.820)

And in fact, everybody agreed in spring of 2015,

Brendan Eich (1:18:23.160)

we're gonna take what was proven by Asm.js

Lex Fridman (1:18:25.500)

and make a new syntax, a binary syntax, it's efficient,

Brendan Eich (1:18:29.460)

that loads into the same JavaScript VM

Lex Fridman (1:18:31.580)

that JavaScript loads into.

Lex Fridman (1:18:32.620)

So there'll be two source languages, one VM,

Lex Fridman (1:18:34.780)

very important, one garbage collector,

Brendan Eich (1:18:36.220)

one memory manager, one set of compiler stages.

Lex Fridman (1:18:40.780)

And that's called WebAssembly.

Lex Fridman (1:18:42.300)

And that's the successor to Asm.js.

Lex Fridman (1:18:44.260)

And it's important that it have binary syntax

Brendan Eich (1:18:46.340)

because at the end of the day, especially on mobile,

Lex Fridman (1:18:48.780)

if you're downloading JavaScript,

Brendan Eich (1:18:50.100)

even if you're using LZ compression on the wire,

Lex Fridman (1:18:52.040)

that's cool, but you've got to blow it out into memory

Lex Fridman (1:18:53.940)

and then parse the silly eight character function keyword

Lex Fridman (1:18:56.860)

that I picked, when I should have used something shorter,

Brendan Eich (1:19:00.820)

I picked it because of awk, the Unix tool.

Lex Fridman (1:19:04.540)

So anyways.

Brendan Eich (1:19:05.380)

I want to, but I'm not following the awk thread.

Lex Fridman (1:19:07.940)

Yeah, don't worry about it.

Brendan Eich (1:19:09.780)

Is it surprising to you that,

Lex Fridman (1:19:12.260)

how damn fast JavaScript is these days?

Brendan Eich (1:19:14.660)

I mean, like, you've been through the whole journey.

Lex Fridman (1:19:16.900)

I know every step of the way, but is it like,

Brendan Eich (1:19:19.640)

I mean, it feels incredible.

Lex Fridman (1:19:22.140)

It does, but I knew, so the funny thing is,

Lex Fridman (1:19:24.220)

computer science is this big karmic wheel, right?

Lex Fridman (1:19:26.860)

Wheel of Fortuna.

Lex Fridman (1:19:29.140)

And in the, it's about the 97,

Lex Fridman (1:19:32.460)

I was loaned by Netscape to do due diligence for Sun

Brendan Eich (1:19:37.100)

in their acquisition of Anamorphic,

Lex Fridman (1:19:39.300)

which was David Unger and friends, people,

Brendan Eich (1:19:45.180)

Craig, I'm forgetting his name, he went to Microsoft.

Lex Fridman (1:19:47.700)

These Stanford language buffs who'd taken Smalltalk

Lex Fridman (1:19:51.180)

and then David create itself

Lex Fridman (1:19:53.500)

as a simpler sort of Smalltalk language

Lex Fridman (1:19:56.260)

and made really fast, just in time compiling VMs for them.

Lex Fridman (1:20:00.620)

And they, you know, well ahead of Java hotspot

Brendan Eich (1:20:04.320)

or JavaScript V8 or any of these modern VMs,

Lex Fridman (1:20:07.660)

figured out how to make dynamic code fast

Lex Fridman (1:20:10.780)

because Smalltalk is dynamic language, right?

Lex Fridman (1:20:13.100)

It has classes, it has, I think more lockdown declarative

Brendan Eich (1:20:16.420)

syntax than JavaScript, but it's fundamentally dynamic.

Lex Fridman (1:20:18.500)

You don't declare the types.

Lex Fridman (1:20:19.900)

But you could infer the types as the program runs

Lex Fridman (1:20:24.180)

and you start to form these ideas

Brendan Eich (1:20:26.100)

about what types are actually flowing through key operations

Lex Fridman (1:20:29.100)

and you form little so called polymorphic inline caches

Brendan Eich (1:20:32.740)

that are optimized machine code.

Lex Fridman (1:20:35.660)

The cache is the machine code that assumes,

Brendan Eich (1:20:38.340)

does a quick check to make sure the type is right

Lex Fridman (1:20:40.180)

and if it's not right, it bails to the interpreter.

Lex Fridman (1:20:42.420)

And if it is right, you go pretty fast.

Lex Fridman (1:20:44.500)

And that short test is a predicted branch,

Lex Fridman (1:20:47.100)

so things are pretty quick.

Lex Fridman (1:20:48.660)

All that amazing stuff I knew about in the 90s

Lex Fridman (1:20:51.340)

and I didn't have time to do it

Lex Fridman (1:20:53.780)

and Anamorphic got bought by Sun and they did hotspot.

Lex Fridman (1:20:57.220)

And you needed that even in Java because at scale,

Lex Fridman (1:21:00.060)

Java has some dynamic aspects due to invoke interface.

Brendan Eich (1:21:04.340)

You can have basically collections of Java code

Lex Fridman (1:21:06.740)

where you don't know at the time each module

Brendan Eich (1:21:10.900)

or package is compiled exactly what's being called,

Lex Fridman (1:21:13.900)

what subclass or what implementation

Brendan Eich (1:21:16.300)

of an interface is being called.

Lex Fridman (1:21:17.820)

And so you want to optimize using this sort of dynamic

Brendan Eich (1:21:20.100)

polymorphic caching there too.

Lex Fridman (1:21:21.740)

And they did that and hotspot, it's amazing beast.

Brendan Eich (1:21:24.620)

I've met like 13 people who all claim they created it.

Lex Fridman (1:21:27.740)

I think one of them may deserve credit more than others.

Lex Fridman (1:21:31.620)

But I didn't get to do that in JavaScript.

Lex Fridman (1:21:34.660)

And when we knew that Google was going to do their own

Brendan Eich (1:21:38.420)

browser, which we knew at Mozilla around 2006,

Lex Fridman (1:21:42.420)

I also met the team that did V8.

Lex Fridman (1:21:45.140)

And it turns out it was Lars Bach,

Lex Fridman (1:21:46.580)

who was one of the young engineers from Anamorphic

Brendan Eich (1:21:49.300)

that got acquired by Sun.

Lex Fridman (1:21:50.660)

And so Lars is like one of the world's expert

Brendan Eich (1:21:52.940)

on these kinds of virtual machines.

Lex Fridman (1:21:54.940)

And he picked my brains about JavaScript.

Brendan Eich (1:21:56.780)

I could tell he didn't like it at the time,

Lex Fridman (1:21:58.100)

but he had to do it.

Brendan Eich (1:21:59.920)

And...

Lex Fridman (1:22:00.760)

Oh, really interesting.

Brendan Eich (1:22:01.580)

Yeah, in 2006 lunch at Google's campus.

Lex Fridman (1:22:03.780)

And then I had another friend who was DevRel at Chrome

Lex Fridman (1:22:07.020)

and he said, yeah, we don't know what they're doing.

Lex Fridman (1:22:08.340)

This is getting 2007 to fall, getting toward 2008.

Brendan Eich (1:22:12.020)

We're trying to get Chrome out

Lex Fridman (1:22:13.020)

and we don't know what's going with the V8 team.

Brendan Eich (1:22:14.500)

They're off in Aarhus, Denmark,

Lex Fridman (1:22:16.900)

rewriting their engine four times, which is good.

Brendan Eich (1:22:19.300)

That's the right way to do this kind of development.

Lex Fridman (1:22:22.300)

They were learning JavaScript, including all its quirks,

Brendan Eich (1:22:24.280)

which they came to hate the fire of a thousand suns,

Lex Fridman (1:22:27.020)

which is one of the reasons that Lars and company did Dart,

Brendan Eich (1:22:29.820)

their own language.

Lex Fridman (1:22:30.940)

But they also made the language fast.

Lex Fridman (1:22:32.900)

And meanwhile, we knew this was happening.

Lex Fridman (1:22:35.260)

So we got our act together with Trace Monkey,

Brendan Eich (1:22:38.500)

our tracing JIT at Mozilla and Apple I think was also aware.

Lex Fridman (1:22:41.860)

And so they were doing their own JIT.

Lex Fridman (1:22:43.020)

So the era of JITed fast JavaScript in 2008

Lex Fridman (1:22:46.420)

had this prehistory going back to Smalltalk itself

Lex Fridman (1:22:49.580)

and Anamorphic.

Lex Fridman (1:22:50.980)

And again, the lineage is interesting

Brendan Eich (1:22:53.100)

because you had Lars and Anamorphic

Lex Fridman (1:22:54.500)

and then he ends up at Google.

Brendan Eich (1:22:55.980)

Yeah, and today we have an incredibly fast language

Lex Fridman (1:23:00.160)

that like you said, still,

Brendan Eich (1:23:02.860)

without hate, you can't have love.

Lex Fridman (1:23:04.780)

So I think there's both love and hate for this dance,

Brendan Eich (1:23:09.820)

this rich complex dance of JavaScript

Lex Fridman (1:23:12.100)

throughout its history.

Brendan Eich (1:23:12.940)

There's a dialectic for sure.

Lex Fridman (1:23:15.300)

Today, JavaScript is the most popular language in the world.

Lex Fridman (1:23:20.140)

Why by many measures?

Lex Fridman (1:23:22.980)

Why do you think that is?

Brendan Eich (1:23:24.700)

Is there some fundamental ideas

Lex Fridman (1:23:28.420)

that you've already spoke to a little bit

Lex Fridman (1:23:29.740)

but sort of broader

Lex Fridman (1:23:30.820)

that you think is the most popular language in the world?

Lex Fridman (1:23:32.820)

So I think I did, by doing first class functions

Lex Fridman (1:23:36.220)

and taking the good parts of the C operator hierarchy

Lex Fridman (1:23:40.780)

and just keeping things simple enough,

Lex Fridman (1:23:44.060)

maybe it could have been simpler

Lex Fridman (1:23:45.100)

but I had to make it look like Java

Lex Fridman (1:23:47.380)

and interoperate with Java

Brendan Eich (1:23:48.780)

that there was inherent goodness,

Lex Fridman (1:23:52.860)

Aristotelian quality there.

Lex Fridman (1:23:54.420)

And people perceive that

Lex Fridman (1:23:56.100)

even through all the quirks and warts.

Lex Fridman (1:23:57.740)

And then over time working on it with the standards body,

Lex Fridman (1:23:59.940)

working on it not only as a core language

Lex Fridman (1:24:02.800)

but in the context of HTML5 and making the browser better,

Lex Fridman (1:24:06.660)

listening to developers, thinking about,

Brendan Eich (1:24:08.940)

this is something that Nick Thompson wrote nicely

Lex Fridman (1:24:10.860)

about on Hacker News, I was very flattered.

Brendan Eich (1:24:12.620)

He said, Java was this thing

Lex Fridman (1:24:14.300)

where the experts were writing the code

Lex Fridman (1:24:16.000)

and it was compiled and you had to declare all your types

Lex Fridman (1:24:18.020)

and Sun didn't really give a damn about

Brendan Eich (1:24:20.980)

the average programmer who wanted to build real web apps,

Lex Fridman (1:24:23.820)

dynamic things.

Lex Fridman (1:24:25.060)

And I was in there meanwhile doing a bunch of people's jobs

Lex Fridman (1:24:28.820)

making JavaScript survive those early years

Brendan Eich (1:24:31.580)

when it was kind of touch and go.

Lex Fridman (1:24:33.620)

JavaScript was considered a Mickey Mouse language.

Brendan Eich (1:24:36.140)

It was for annoyances like the scrolling text

Lex Fridman (1:24:38.660)

at the bottom of the browser in the status bar.

Lex Fridman (1:24:40.780)

But I kept listening to developers working with them

Lex Fridman (1:24:43.060)

and trying to make it run in that single threaded event loop

Brendan Eich (1:24:46.860)

in a useful way.

Lex Fridman (1:24:47.720)

And I think that forged something

Brendan Eich (1:24:49.880)

that people have come to love.

Lex Fridman (1:24:51.180)

Now you don't always love the best thing, right?

Brendan Eich (1:24:53.700)

I talked about Shakespeare sonnet about

Lex Fridman (1:24:58.620)

I'm Mr. Sizer, nothing like the sun.

Brendan Eich (1:25:00.180)

Or the scene from Josh Whedon's film Serenity at the end

Lex Fridman (1:25:04.860)

where the actual piece in the score by David Newman

Brendan Eich (1:25:08.340)

is called Love where Captain Mal is teaching River Tam

Lex Fridman (1:25:12.860)

about how to pilot the ship.

Lex Fridman (1:25:13.860)

And she's a super genius, super soldier.

Lex Fridman (1:25:15.380)

She knows how to do it already.

Lex Fridman (1:25:17.100)

And he's basically talking about how you have to love

Lex Fridman (1:25:19.500)

the ship because if you don't, it's going to kill you.

Lex Fridman (1:25:22.020)

And then the piece falls off the ship.

Lex Fridman (1:25:23.860)

It's kind of like JavaScript.

Brendan Eich (1:25:26.060)

You have to love it.

Lex Fridman (1:25:26.900)

You have to love it because now people say we're stuck

Brendan Eich (1:25:30.020)

with it because it got this priority of place.

Lex Fridman (1:25:31.620)

But there's love underpinning that.

Lex Fridman (1:25:33.340)

And actually the listening to developers,

Lex Fridman (1:25:35.860)

that's kind of beautiful.

Brendan Eich (1:25:36.700)

There's most successful products in this world

Lex Fridman (1:25:41.140)

with all the messes, with all the flaws.

Brendan Eich (1:25:43.100)

Perhaps the flaws themselves are actual features,

Lex Fridman (1:25:45.360)

but that's a whole nother, that's a discussion about love.

Lex Fridman (1:25:48.180)

But underneath it, there's something

Lex Fridman (1:25:50.220)

that just connects with people.

Lex Fridman (1:25:52.420)

And it has to keep connecting.

Lex Fridman (1:25:53.480)

If JavaScript kind of went off in this,

Brendan Eich (1:25:55.640)

people sometimes complain about ES6.

Lex Fridman (1:25:57.340)

Oh, you put classes in JavaScript.

Brendan Eich (1:25:58.700)

I hate classes.

Lex Fridman (1:25:59.540)

You've ruined it.

Lex Fridman (1:26:00.420)

But it's not true.

Lex Fridman (1:26:01.420)

It's a dynamic language.

Brendan Eich (1:26:02.660)

Smalltalk had classes.

Lex Fridman (1:26:04.820)

Python has classes.

Brendan Eich (1:26:07.060)

There are lots of Lisp variants that have classy systems.

Lex Fridman (1:26:12.740)

So people who don't reject it based on some sort

Brendan Eich (1:26:16.780)

of fashion judgment do use it and do interact

Lex Fridman (1:26:22.120)

with the standards body.

Brendan Eich (1:26:23.380)

The standards body is competing browser vendors mainly,

Lex Fridman (1:26:25.840)

but also now big companies that use JavaScript heavily,

Brendan Eich (1:26:28.480)

the Paypal's and other such companies, Salesforce.

Lex Fridman (1:26:32.980)

And they have to cater to web developers.

Brendan Eich (1:26:36.300)

They have to hire developers who know JavaScript.

Lex Fridman (1:26:39.020)

They have to keep their engines up to the latest standard.

Lex Fridman (1:26:43.100)

And this creates all this sort of social structure

Lex Fridman (1:26:45.900)

around JavaScript that is unusual.

Brendan Eich (1:26:47.820)

I mean, you get C++ buffs that follow the inner workings

Lex Fridman (1:26:51.420)

of C++, what is it now?

Brendan Eich (1:26:54.460)

21 something, I don't know.

Lex Fridman (1:26:56.580)

I've lost track.

Lex Fridman (1:26:57.500)

But it's a more rarefied group.

Lex Fridman (1:26:59.020)

It's more like the old language, gray hairs.

Brendan Eich (1:27:02.860)

Whereas JavaScript is a younger

Lex Fridman (1:27:04.500)

and more vibrant and large crowd.

Brendan Eich (1:27:07.380)

There's a community feel to it.

Lex Fridman (1:27:09.260)

There's echoes, perhaps I don't wanna draw too many

Brendan Eich (1:27:13.180)

similarities, maybe you can comment on it.

Lex Fridman (1:27:15.420)

There's C++ is like Wall Street,

Lex Fridman (1:27:18.980)

and the JavaScript is like Wall Street bets

Lex Fridman (1:27:21.220)

from the recent events.

Brendan Eich (1:27:22.620)

It's like there's a chaotic community of all,

Lex Fridman (1:27:24.940)

and there's some power from that distributed crowd

Brendan Eich (1:27:28.820)

of people that ultimately.

Lex Fridman (1:27:30.260)

It's more thematic, it's more of the people.

Brendan Eich (1:27:32.660)

It lets people in without requiring these credentials.

Lex Fridman (1:27:35.860)

I remember in the late 90s into the 90s,

Brendan Eich (1:27:38.420)

people were all getting Java credentials.

Lex Fridman (1:27:39.900)

And I knew people, and friends knew people

Brendan Eich (1:27:42.700)

who became Java programmers, and you knew

Lex Fridman (1:27:44.580)

they really should have been like nature guides or pilots.

Brendan Eich (1:27:47.260)

They hated programming, but they thought,

Lex Fridman (1:27:49.020)

I gotta make money, I'm gonna become a Java programmer.

Lex Fridman (1:27:51.780)

Do you have some, because it's such a monumental moment

Lex Fridman (1:27:55.060)

in our current history, as a quick aside,

Lex Fridman (1:27:58.060)

do you have thoughts about this huge distributed crowdsourced

Lex Fridman (1:28:03.260)

financial happenings with Wall Street bets?

Brendan Eich (1:28:06.740)

That's like nobody could have, well,

Lex Fridman (1:28:08.780)

you could have predicted, but the scale and the impact

Brendan Eich (1:28:11.420)

of this kind of emergent behavior

Lex Fridman (1:28:13.260)

from independent parties that could happen.

Brendan Eich (1:28:15.780)

Like I said, my own experience with the dismal science,

Lex Fridman (1:28:19.900)

as with physics, led me to reject a lot of bad models.

Brendan Eich (1:28:23.500)

Economics was always compromised by politics,

Lex Fridman (1:28:26.260)

political economy.

Brendan Eich (1:28:28.220)

You could also argue that it was,

Lex Fridman (1:28:29.460)

it used to be a branch of moral philosophy,

Lex Fridman (1:28:30.980)

so it was concerned with the good,

Lex Fridman (1:28:32.460)

and it became divorced and became sort of

Brendan Eich (1:28:34.740)

in this quasi Newtonian way, just about,

Lex Fridman (1:28:37.700)

everything's just running by itself, don't worry about it.

Brendan Eich (1:28:40.220)

This monopoly's crushing your Netscape company,

Lex Fridman (1:28:42.060)

but that's just nature.

Lex Fridman (1:28:43.700)

And economics couldn't, or doesn't really have good models

Lex Fridman (1:28:46.580)

for the Wall Street bets subreddit.

Lex Fridman (1:28:49.500)

You know, they know how to squeeze a short, right?

Lex Fridman (1:28:52.300)

So the amazing thing is you have Robinhood app,

Brendan Eich (1:28:56.460)

which was, again, supposedly for the demos, for the people,

Lex Fridman (1:29:00.740)

and eliminated the fee through various kinds of straddles

Brendan Eich (1:29:06.300)

or some kind of spread operation

Lex Fridman (1:29:09.220)

that helped them eliminate the fee or eat the fee.

Lex Fridman (1:29:11.860)

And in fact, as a broker in these days,

Lex Fridman (1:29:14.540)

because it takes two days to settle,

Brendan Eich (1:29:15.820)

there's counterparty risk, as they found out.

Lex Fridman (1:29:18.620)

And so the Wall Street bets people,

Brendan Eich (1:29:22.980)

you know, the memes are like the Terminator robot

Lex Fridman (1:29:25.700)

with the $600 STEMI check and the hedge funds

Brendan Eich (1:29:30.140)

that make little girl hiding under the desk.

Lex Fridman (1:29:34.060)

There is a problem, which I talked about

Brendan Eich (1:29:36.060)

in a recent podcast, which I'm conscious of

Lex Fridman (1:29:38.300)

from the history of the web, and that is,

Brendan Eich (1:29:40.020)

you could say it's monopoly,

Lex Fridman (1:29:41.260)

which antitrust wasn't enforced

Brendan Eich (1:29:42.980)

after USB Microsoft for a long time.

Lex Fridman (1:29:45.140)

And a lot of this was due to the money interests

Brendan Eich (1:29:48.020)

buying control of politicians.

Lex Fridman (1:29:50.900)

And, you know, in Plato's five regimes, that's oligarchy.

Brendan Eich (1:29:54.460)

That's where we are.

Lex Fridman (1:29:56.060)

And now we're seeing a fight against the oligarchs.

Brendan Eich (1:29:58.780)

I don't know if it'll work,

Lex Fridman (1:29:59.620)

but you're definitely seeing it.

Lex Fridman (1:30:00.780)

And it's also kind of hackerish, right?

Lex Fridman (1:30:02.420)

It's got a hacker ethos.

Brendan Eich (1:30:04.100)

You know, hey, Robinhood, no fees.

Lex Fridman (1:30:05.660)

Oh, interesting.

Brendan Eich (1:30:06.500)

Hey, you know, I could buy a fraction of a share

Lex Fridman (1:30:09.180)

in this thing, or I can keep buying with my stimulus check.

Lex Fridman (1:30:12.300)

So I mentioned Hegel seeing Napoleon on the horse, right?

Lex Fridman (1:30:16.980)

Hegel also talked about the cunning of reason

Brendan Eich (1:30:19.220)

that you have this sort of, you know,

Lex Fridman (1:30:21.380)

God sees history in full, and if you believe in God,

Brendan Eich (1:30:24.100)

or, you know, we don't know the future,

Lex Fridman (1:30:25.940)

but there's always this sort of fly in the ointment,

Brendan Eich (1:30:28.380)

this unintended consequence that confounds

Lex Fridman (1:30:32.140)

the best plans of the powers that be.

Lex Fridman (1:30:35.660)

And we're living through it.

Lex Fridman (1:30:37.060)

I'm glad it's not, you know, street warfare

Brendan Eich (1:30:39.700)

or mechanized warfare, because it has been in the past.

Lex Fridman (1:30:43.380)

It's more like soft power, and people are fighting back.

Lex Fridman (1:30:50.620)

Do you think it's possible?

Lex Fridman (1:30:51.740)

So JavaScript used to be for the front end of the web.

Brendan Eich (1:30:57.100)

It's now increasingly so being used for back end,

Lex Fridman (1:31:01.620)

like running stuff that's like behind the scenes.

Lex Fridman (1:31:06.060)

And it's also starting to be used quite a bit

Lex Fridman (1:31:10.260)

for things like TensorFlow.js.

Lex Fridman (1:31:12.260)

So starting to actually use like these heavy duty

Lex Fridman (1:31:14.700)

applications that are using neural networks,

Brendan Eich (1:31:17.140)

machine learning, and so on in the browser.

Lex Fridman (1:31:19.380)

Is it possible in 10, 20, 30 years

Lex Fridman (1:31:24.140)

that basically most of the world runs on JavaScript?

Lex Fridman (1:31:28.660)

This is a dystopia and a nightmare to some people.

Brendan Eich (1:31:32.380)

When we did Asm.js and WebAssembly,

Lex Fridman (1:31:34.940)

I would joke and mean people with scenes like

Brendan Eich (1:31:37.340)

Neo waking up in his pod in the matrix,

Lex Fridman (1:31:39.500)

and he's all skinny and weak and hairless.

Brendan Eich (1:31:42.380)

And, you know, you realize in the future

Lex Fridman (1:31:46.220)

that you're living in some simulation

Brendan Eich (1:31:48.020)

that it's all running on JavaScript,

Lex Fridman (1:31:49.180)

and you just scream forever.

Brendan Eich (1:31:51.980)

It's possible.

Lex Fridman (1:31:52.820)

Gary Bernhardt does these funny talks.

Brendan Eich (1:31:54.860)

He did watch.js, and then he did this

Lex Fridman (1:31:58.020)

life and death of JavaScript, I think it's called,

Brendan Eich (1:31:59.540)

where he took some clever ideas that actually have

Lex Fridman (1:32:02.580)

a thread of credibility to them.

Lex Fridman (1:32:05.340)

But I mentioned software fault isolation.

Lex Fridman (1:32:09.220)

In the old days, when we were using computers,

Brendan Eich (1:32:10.620)

we said we're gonna use the Unix monolithic monitor,

Lex Fridman (1:32:14.140)

and it's the privileged program.

Brendan Eich (1:32:15.980)

This is before you even had hardware rings of protection.

Lex Fridman (1:32:18.420)

Those, some of the early 60s operating systems

Brendan Eich (1:32:20.460)

used hardware protection zones.

Lex Fridman (1:32:22.700)

But Unix is privileged, and the program

Brendan Eich (1:32:25.900)

that runs user code in a process is hosted.

Lex Fridman (1:32:30.460)

It's the guest, in the host, and you get to suspend it.

Brendan Eich (1:32:33.660)

You get to kill it.

Lex Fridman (1:32:35.220)

If it crashes, it doesn't take down the whole OS.

Brendan Eich (1:32:37.300)

It's a wonderful idea, but the call into the kernel

Lex Fridman (1:32:41.940)

is expensive, the system call, so called.

Lex Fridman (1:32:43.820)

And this has even been optimized now

Lex Fridman (1:32:45.180)

for things like getting the time of day,

Lex Fridman (1:32:46.320)

so it doesn't actually enter the kernel.

Lex Fridman (1:32:49.580)

And meanwhile, hardware architectures

Lex Fridman (1:32:52.780)

and virtualization techniques have gone

Lex Fridman (1:32:54.460)

in a different direction, even to the point

Brendan Eich (1:32:56.380)

where you can do software fault isolation very cheaply

Lex Fridman (1:32:58.840)

without entering the operating system kernel,

Lex Fridman (1:33:00.740)

and so you get unikernels and exokernels

Lex Fridman (1:33:02.940)

and very lightweight VMs.

Lex Fridman (1:33:04.460)

And so Gary took this idea and said, JavaScript

Lex Fridman (1:33:08.060)

will take over computing, because the system call

Brendan Eich (1:33:10.540)

boundary's too expensive, so everything ends up

Lex Fridman (1:33:12.340)

in JavaScript with these lighter weight

Brendan Eich (1:33:14.220)

isolation enforcement mechanisms.

Lex Fridman (1:33:16.940)

It's not totally beyond belief.

Brendan Eich (1:33:18.580)

It'd be WebAssembly too.

Lex Fridman (1:33:22.100)

It's nice to ask you sort of for advice to,

Brendan Eich (1:33:25.380)

there's so many people that are interested

Lex Fridman (1:33:26.980)

in starting to learning about programming,

Brendan Eich (1:33:29.180)

getting into this world.

Lex Fridman (1:33:31.180)

Is there some number of languages,

Brendan Eich (1:33:34.980)

three to five programming languages

Lex Fridman (1:33:36.460)

that you would recommend people learn,

Brendan Eich (1:33:38.220)

or maybe a broader advice on how

Lex Fridman (1:33:42.060)

to get started in programming?

Brendan Eich (1:33:43.940)

Well, so you asked about machine learning,

Lex Fridman (1:33:45.380)

and JavaScript is a general purpose language,

Lex Fridman (1:33:47.820)

and it's a language that's not that great

Lex Fridman (1:33:53.100)

for doing matrix operations or doing parallel programming,

Brendan Eich (1:33:58.100)

I would say, without using some extensions

Lex Fridman (1:34:01.100)

or some libraries that have some magic in them.

Lex Fridman (1:34:03.620)

So if someone wanted to learn,

Lex Fridman (1:34:08.580)

there are amazing languages in sort of the APL family

Brendan Eich (1:34:11.100)

that are very useful for, I would say, linear algebra,

Lex Fridman (1:34:14.060)

which gets to a lot of the kernels in machine learning.

Lex Fridman (1:34:16.820)

And so APL had like J and then K and their K variants

Lex Fridman (1:34:22.240)

because the guy that did K is still going,

Lex Fridman (1:34:24.660)

and they're proprietary, but he's still innovating there.

Lex Fridman (1:34:28.900)

There are, you know, Python is used.

Lex Fridman (1:34:30.420)

So people talk about TensorFlow.js.

Lex Fridman (1:34:31.740)

Well, it's not that surprising

Brendan Eich (1:34:33.460)

because Python was heavily used for machine learning,

Lex Fridman (1:34:36.380)

and Python was always, you know,

Brendan Eich (1:34:38.740)

they didn't have this fast just in time compiler tradition.

Lex Fridman (1:34:41.540)

There were some projects that tried this,

Lex Fridman (1:34:43.180)

and some of them were interesting.

Lex Fridman (1:34:45.340)

PyPy was interesting,

Lex Fridman (1:34:46.300)

but the philosophy with Python was,

Lex Fridman (1:34:50.020)

oh, you need to go fast, write a C plugin,

Lex Fridman (1:34:51.780)

and drop into C code.

Lex Fridman (1:34:53.300)

So I think people should look at multiple languages

Brendan Eich (1:34:57.500)

because there are different tools in the belt.

Lex Fridman (1:34:59.500)

If you're trying to do supervision or rapid prototyping,

Brendan Eich (1:35:02.820)

you want a dynamic language.

Lex Fridman (1:35:03.900)

You want to throw things together and see what works.

Brendan Eich (1:35:06.500)

If you are trying to go down to the metal very fast,

Lex Fridman (1:35:09.280)

well, I'm an old C hacker,

Lex Fridman (1:35:10.740)

but I was also the executive sponsor of Rust at Mozilla,

Lex Fridman (1:35:14.300)

and Rust has now escaped, you know,

Brendan Eich (1:35:16.180)

from that sort of nest where it was born

Lex Fridman (1:35:18.540)

to be adopted by a bunch of companies

Brendan Eich (1:35:21.460)

that have a foundation in the works.

Lex Fridman (1:35:23.220)

Some of the key core team members

Brendan Eich (1:35:24.780)

are working now at Amazon and other places.

Lex Fridman (1:35:27.700)

So it looks like Rust has reached escape velocity,

Lex Fridman (1:35:30.100)

and Rust is an interesting language

Lex Fridman (1:35:31.380)

because one of our goals there,

Brendan Eich (1:35:33.060)

one of the reasons I sponsored it was we were all tired

Lex Fridman (1:35:35.300)

of seeing those remote code execution vulnerabilities

Brendan Eich (1:35:38.100)

due to C and C++, and we thought,

Lex Fridman (1:35:41.240)

can we have a sort of safety property

Brendan Eich (1:35:43.840)

through a type and effect system or an ownership system,

Lex Fridman (1:35:46.920)

and Rust has that.

Lex Fridman (1:35:48.300)

And that ownership system is interesting

Lex Fridman (1:35:50.020)

because it doesn't just give you memory safety.

Brendan Eich (1:35:52.140)

There's a sort of theorem for free,

Lex Fridman (1:35:53.500)

a dual that falls out for protection against data races.

Lex Fridman (1:35:57.220)

So Rust is better for low level programming.

Lex Fridman (1:35:59.740)

You delimit your unsafe code

Brendan Eich (1:36:01.100)

where you do have to be unsafe,

Lex Fridman (1:36:02.900)

and you can prove certain facts about memory safety

Lex Fridman (1:36:05.300)

and race condition avoidance.

Lex Fridman (1:36:08.820)

And so I think people should learn these new languages.

Brendan Eich (1:36:11.340)

I think Go is a great language.

Lex Fridman (1:36:13.380)

I admire, you know, the Unix people who did that.

Brendan Eich (1:36:16.540)

Ken Stoll was involved, Rob Pike, of course, David,

Lex Fridman (1:36:21.420)

what's his name, and other people.

Brendan Eich (1:36:23.440)

Go is a huge success, really on the server side,

Lex Fridman (1:36:27.100)

anywhere you have a lot of networking to do,

Lex Fridman (1:36:30.140)

and it's garbage collected, but it's also very pragmatic.

Lex Fridman (1:36:33.780)

It has that sort of C flavor.

Brendan Eich (1:36:35.420)

As an old C hacker, I can't get used to the fact

Lex Fridman (1:36:37.020)

that they swapped the type and declarator

Brendan Eich (1:36:39.620)

in the declaration order.

Lex Fridman (1:36:41.140)

I haven't used Rust, but this is one of the most respected

Lex Fridman (1:36:44.180)

and loved languages currently, so it's interesting.

Lex Fridman (1:36:46.260)

Yeah, and it's still young.

Brendan Eich (1:36:47.380)

You look at these things, JavaScript is now considered old.

Lex Fridman (1:36:50.980)

It's gone through so many versions

Brendan Eich (1:36:52.500)

that you can fall in love with it all over again.

Lex Fridman (1:36:54.860)

25 plus years, you know, it's an adult.

Brendan Eich (1:36:58.740)

It should be out of the house.

Lex Fridman (1:36:59.800)

But it could be around another 25 years.

Brendan Eich (1:37:03.420)

Cannot rule it out.

Lex Fridman (1:37:04.240)

So Rust will be around for a long time.

Brendan Eich (1:37:06.260)

The longer you're around, the more likely you're Lindy,

Lex Fridman (1:37:08.900)

and you're around your wife.

Brendan Eich (1:37:10.400)

A lot of people ask me, like,

Lex Fridman (1:37:13.500)

I'm often torn between recommending either Python

Brendan Eich (1:37:15.780)

or JavaScript as the first language to play with,

Lex Fridman (1:37:18.100)

because, I mean, it's difficult,

Brendan Eich (1:37:20.900)

because it's so easy to do JavaScript incorrectly.

Lex Fridman (1:37:25.180)

It's much easier to do it correctly these days,

Brendan Eich (1:37:29.460)

or like, well, learn about programming.

Lex Fridman (1:37:31.620)

But the cool thing about JavaScript

Brendan Eich (1:37:33.940)

is that you can create stuff

Lex Fridman (1:37:36.100)

that will put a smile on your face.

Brendan Eich (1:37:38.140)

Like, as a developer, you can create stuff,

Lex Fridman (1:37:41.460)

and it'll visually look like something,

Lex Fridman (1:37:43.060)

and it'll do stuff, and it makes you feel good.

Lex Fridman (1:37:45.700)

It makes you fall in love with programming.

Brendan Eich (1:37:47.340)

With Python, you could do the same.

Lex Fridman (1:37:48.700)

It's a little slower.

Lex Fridman (1:37:49.540)

And with C++, it takes five to 10 years

Lex Fridman (1:37:51.580)

to write a program that actually does something.

Brendan Eich (1:37:54.100)

So, like, there's that tension between

Lex Fridman (1:37:56.820)

is JavaScript the right first step, or is it Python?

Lex Fridman (1:37:59.460)

And I've been going back and forth on those two.

Lex Fridman (1:38:02.060)

My Python, right, it came from a lineage of ABC,

Brendan Eich (1:38:05.300)

which was a pedagogical language in the Netherlands.

Lex Fridman (1:38:09.580)

And it, you know,

Brendan Eich (1:38:12.140)

it was a good teaching language, too.

Lex Fridman (1:38:16.700)

I think it is a good teaching language,

Lex Fridman (1:38:17.980)

and it's a little more restrictive

Lex Fridman (1:38:19.100)

in that if you misspell something

Brendan Eich (1:38:21.980)

in a way that JavaScript might let run,

Lex Fridman (1:38:25.220)

let reach runtime, it'll get stopped

Brendan Eich (1:38:26.860)

at syntax check in Python.

Lex Fridman (1:38:30.600)

That's good for beginners.

Brendan Eich (1:38:31.700)

I think the sloppiness that some people object to,

Lex Fridman (1:38:34.760)

like, people were just tweeting at me,

Brendan Eich (1:38:37.140)

having just learned JavaScript.

Lex Fridman (1:38:38.660)

They said, I can take a number, and I can index into it,

Lex Fridman (1:38:41.160)

and get undefined out of it as a property.

Lex Fridman (1:38:43.180)

Why is that?

Brendan Eich (1:38:44.020)

A number's not an object.

Lex Fridman (1:38:44.860)

And I explained why it is,

Brendan Eich (1:38:46.180)

because like in Java, the primitive types,

Lex Fridman (1:38:48.340)

which unfortunately are not objects,

Brendan Eich (1:38:49.840)

can be automatically boxed or wrapped by an object.

Lex Fridman (1:38:53.260)

And I made that implicit.

Brendan Eich (1:38:55.440)

In Java, it's typed, and you have to declare things,

Lex Fridman (1:38:58.240)

and you'll get type errors.

Lex Fridman (1:39:00.500)

But there are cases in Java where you get auto boxing

Lex Fridman (1:39:03.080)

or auto wrapping, because you've declared that you want it.

Brendan Eich (1:39:06.020)

In JavaScript, it just happens.

Lex Fridman (1:39:07.580)

And so once I explained it, like, oh, wow, I get it.

Lex Fridman (1:39:10.300)

But it also means that you can commit a blunder that just.

Lex Fridman (1:39:13.540)

You don't get punished for it, you don't detect.

Brendan Eich (1:39:15.360)

You get an undefined value,

Lex Fridman (1:39:16.800)

and you don't know where it came from.

Brendan Eich (1:39:17.860)

Right.

Lex Fridman (1:39:19.900)

I've been reading a lot about military history recently.

Lex Fridman (1:39:23.340)

And one way to paint the picture of browsers,

Lex Fridman (1:39:26.780)

internet browsers, is through the various wars

Brendan Eich (1:39:29.280)

throughout its history.

Lex Fridman (1:39:30.620)

I don't know if that's a useful way to look at it,

Lex Fridman (1:39:32.300)

but we've already talked a little bit about Netscape

Lex Fridman (1:39:34.660)

and Internet Explorer in the early days.

Lex Fridman (1:39:38.420)

Can you tell the story of the different wars,

Lex Fridman (1:39:41.260)

if that's at all an interesting way to look at it,

Lex Fridman (1:39:43.420)

of the 90s and to today?

Lex Fridman (1:39:46.460)

Yeah.

Lex Fridman (1:39:47.300)

So I mentioned that Microsoft, you know,

Lex Fridman (1:39:49.100)

which was convicted for it, did abuse its monopoly,

Lex Fridman (1:39:51.500)

but they had a pretty good team by the time they did IE4.

Lex Fridman (1:39:55.100)

And Netscape, unfortunately, I was like second floor,

Lex Fridman (1:39:57.900)

and I was friends with all the first floor people,

Lex Fridman (1:39:59.540)

the front end guys who did the JavaScript event hookup

Lex Fridman (1:40:02.700)

and things like that,

Lex Fridman (1:40:04.700)

that that team was fairly burnt out.

Lex Fridman (1:40:07.580)

And I think having gone public,

Lex Fridman (1:40:10.900)

the upper management wanted to buy a bunch of companies

Brendan Eich (1:40:13.300)

to try to go head to head with Microsoft.

Lex Fridman (1:40:15.340)

Didn't work, but buying a bunch of companies

Brendan Eich (1:40:17.740)

usually doesn't work.

Lex Fridman (1:40:19.380)

I think the modern sort of approach roughly

Brendan Eich (1:40:22.800)

is like Mark Zuckerberg took,

Lex Fridman (1:40:25.260)

which is to keep them at arm's length

Lex Fridman (1:40:26.860)

and let them do their thing.

Lex Fridman (1:40:28.300)

And now that he's pulling WhatsApp in

Lex Fridman (1:40:29.900)

and people are fleeing it

Lex Fridman (1:40:31.420)

because it's tied into the ad surveillance.

Brendan Eich (1:40:33.480)

But, you know, for a while,

Lex Fridman (1:40:36.260)

they're keeping it separate really does work

Brendan Eich (1:40:37.660)

because you bought it for its value,

Lex Fridman (1:40:39.020)

it's complimentary, and you're not messing with it.

Brendan Eich (1:40:41.140)

With Netscape, when they bought a bunch of companies,

Lex Fridman (1:40:43.600)

they had some of the first floor people

Brendan Eich (1:40:45.500)

or the founders burned out.

Lex Fridman (1:40:46.940)

They had newcomers who wanted their turn to do the browser,

Lex Fridman (1:40:51.460)

and they hadn't really done browsers or understood them.

Lex Fridman (1:40:54.540)

And so Netscape 4 was originally supposed to be 3,

Lex Fridman (1:40:57.340)

and it was so late, they renumbered it.

Lex Fridman (1:40:58.820)

We did a 3 release.

Brendan Eich (1:41:00.560)

Jamie and a few others put some extra effort into.

Lex Fridman (1:41:03.060)

SecureMine was supported in the built in mail program.

Lex Fridman (1:41:08.340)

And Netscape 4 was late,

Lex Fridman (1:41:10.200)

and it was only on Windows at first,

Lex Fridman (1:41:11.860)

and Microsoft had really started doing better,

Lex Fridman (1:41:14.920)

like they do.

Brendan Eich (1:41:15.760)

They copy, and the first version's trash,

Lex Fridman (1:41:17.300)

and the second one, you're starting to feel threatened.

Brendan Eich (1:41:19.100)

The third one, you can tell what's gonna happen,

Lex Fridman (1:41:21.460)

and the fourth one's good.

Lex Fridman (1:41:22.460)

And plus there's the benefit, like you said,

Lex Fridman (1:41:24.260)

that it comes as a default browser.

Brendan Eich (1:41:26.120)

Yes, and yet Netscape's screwing it up,

Lex Fridman (1:41:28.780)

and Microsoft really putting some quality people on it.

Brendan Eich (1:41:32.540)

IE4 was good.

Lex Fridman (1:41:33.380)

On Windows, it was good.

Lex Fridman (1:41:34.220)

And they did the dynamic HTML innovations.

Lex Fridman (1:41:38.300)

Scott Isaac's my old buddy,

Brendan Eich (1:41:39.700)

a former accountant who programmed in BASIC

Lex Fridman (1:41:42.360)

and became what Microsoft calls a program manager,

Brendan Eich (1:41:45.500)

which is kind of an elevated position.

Lex Fridman (1:41:48.760)

You can be a programmer or an engineer and track,

Lex Fridman (1:41:50.500)

but you switch to it,

Lex Fridman (1:41:51.340)

and you sort of lead a lot of design and standards efforts.

Lex Fridman (1:41:54.700)

And so Scott Isaac put in a lot of those funky DHTML APIs

Lex Fridman (1:41:58.940)

that didn't quite have the same flavor

Brendan Eich (1:42:01.020)

as the stuff that I did, and neither of them

Lex Fridman (1:42:03.260)

was like the later sort of verbose Java,

Brendan Eich (1:42:05.540)

like DOM W3C standardized.

Lex Fridman (1:42:08.540)

But IE4 was pretty darn good.

Brendan Eich (1:42:10.540)

I remember a friend, Scott Furman and I,

Lex Fridman (1:42:12.460)

got invited by Scott Isaac to Gordon Beers in San Jose.

Brendan Eich (1:42:16.700)

They were doing a preview of IE4.

Lex Fridman (1:42:19.340)

This must have been 1997.

Lex Fridman (1:42:21.500)

And Scott said, yeah, here's the new graphics stuff

Lex Fridman (1:42:25.140)

we're doing.

Brendan Eich (1:42:25.980)

We've got something like your Netscape layers.

Lex Fridman (1:42:28.220)

We've got VML, a vector markup language.

Brendan Eich (1:42:31.740)

We can do virtual reality.

Lex Fridman (1:42:33.420)

And Scott and I looked at each other and said,

Lex Fridman (1:42:35.260)

we're doomed, right?

Lex Fridman (1:42:36.820)

Microsoft was starting to fire on all cylinders.

Lex Fridman (1:42:38.780)

So I have to give them credit for that,

Lex Fridman (1:42:40.460)

even though they abused their market power.

Lex Fridman (1:42:42.940)

And maybe I shouldn't give them credit

Lex Fridman (1:42:44.320)

for having the resources to hire talented people,

Lex Fridman (1:42:45.940)

but they did a credible job on IE4.

Lex Fridman (1:42:48.900)

What really was bad was that phase of the browser wars

Brendan Eich (1:42:51.940)

ended with monopoly and perhaps due to the antitrust case,

Lex Fridman (1:42:56.100)

perhaps due to regulation in Europe,

Brendan Eich (1:42:59.340)

perhaps just due to Microsoft not liking

Lex Fridman (1:43:01.100)

dealing with standardization, they let it rot.

Brendan Eich (1:43:05.420)

They just abandoned it, IE5, 5.5, IE6 later,

Lex Fridman (1:43:10.520)

but these were not well maintained.

Brendan Eich (1:43:12.300)

They had a lot of security bugs.

Lex Fridman (1:43:14.420)

It felt really closed and outdated too,

Brendan Eich (1:43:16.660)

even though it's getting updated, it's just weird.

Lex Fridman (1:43:19.020)

Browsers like Mozilla and then Firefox were adding tabs.

Brendan Eich (1:43:22.100)

Opera had a version of tabs and they didn't add tabs.

Lex Fridman (1:43:25.940)

And they pop up blocking,

Brendan Eich (1:43:27.560)

something I should have done from the start.

Lex Fridman (1:43:28.860)

People realized that you can tell

Brendan Eich (1:43:30.160)

when the user clicks something

Lex Fridman (1:43:31.540)

and it goes in JavaScript to open a little window,

Brendan Eich (1:43:34.300)

that you can sort of inspect the stack

Lex Fridman (1:43:36.220)

and see that the click originated that,

Lex Fridman (1:43:38.140)

and it's probably okay.

Lex Fridman (1:43:39.420)

Whereas if you're just loading a script

Lex Fridman (1:43:40.820)

and it opens a new window, that's a spam technique

Lex Fridman (1:43:43.860)

and you should block it.

Brendan Eich (1:43:45.220)

Tabs were a brilliant innovation.

Lex Fridman (1:43:47.380)

Like you said, Opera had it,

Lex Fridman (1:43:48.380)

but I remember I fully switched to Firefox the moment.

Lex Fridman (1:43:52.060)

I remember the moments of first using tabs in Firefox

Lex Fridman (1:43:57.060)

and not liking it for the first few minutes,

Lex Fridman (1:44:00.220)

and then like, wait a minute.

Brendan Eich (1:44:02.460)

You get the groove, yeah.

Lex Fridman (1:44:03.300)

You get the groove and you understand.

Lex Fridman (1:44:05.580)

So that timing, what year was this?

Lex Fridman (1:44:08.540)

Because also as a aspiring web designer,

Brendan Eich (1:44:12.100)

I use Table, so we didn't mention Layout or CSS much.

Lex Fridman (1:44:16.260)

There's also a change in the way

Brendan Eich (1:44:18.760)

the frames were going away.

Lex Fridman (1:44:21.100)

So there's a change in the way websites looked

Lex Fridman (1:44:23.620)

and behaved and all that kind of stuff.

Lex Fridman (1:44:25.300)

CSS finally, which Microsoft embraced with IA4

Lex Fridman (1:44:28.420)

and Netscape never really did right.

Lex Fridman (1:44:30.700)

CSS became a better standard over time

Brendan Eich (1:44:33.440)

for doing Table Layout that relieved you

Lex Fridman (1:44:36.140)

of the need to use what are called spacer GIFs,

Lex Fridman (1:44:39.260)

spacer GIFs, right?

Lex Fridman (1:44:40.300)

Images, you would throw into space at tables.

Brendan Eich (1:44:44.180)

The typographic power of the web has gotten better,

Lex Fridman (1:44:47.460)

but it's still not on the level of PDF

Lex Fridman (1:44:49.320)

and you can't do advanced typography,

Lex Fridman (1:44:51.580)

but it's gotten really better.

Lex Fridman (1:44:53.180)

And even then, tables were getting better.

Lex Fridman (1:44:54.920)

If you were using Firefox, that would have been 2004

Brendan Eich (1:44:57.260)

because it was called Firebird until earlier that year.

Lex Fridman (1:45:00.260)

No, yeah, I think it wasn't.

Brendan Eich (1:45:01.860)

2003.

Lex Fridman (1:45:02.700)

I don't remember, it was a Firebird, which had tabs?

Brendan Eich (1:45:06.100)

We had tabs the whole way.

Lex Fridman (1:45:07.060)

So it started out as Mozilla slash browser in 2002,

Brendan Eich (1:45:10.420)

became Phoenix.

Lex Fridman (1:45:12.080)

There's a BIOS that has an embedded version of IE

Lex Fridman (1:45:14.700)

and they said, we're called Phoenix Technologies,

Lex Fridman (1:45:16.420)

you can't use Phoenix.

Lex Fridman (1:45:17.620)

And so we said, okay, we'll call it Firebird.

Lex Fridman (1:45:19.180)

And then this Australian centered

Brendan Eich (1:45:21.220)

open source database project started really like

Lex Fridman (1:45:24.900)

in the true Mad Max style, just screaming at us saying,

Brendan Eich (1:45:27.340)

you can't use Firebird.

Lex Fridman (1:45:28.380)

And I had to sort of be the ambassador and say,

Brendan Eich (1:45:30.660)

okay, we're gonna rename.

Lex Fridman (1:45:31.500)

And they're like, we don't believe you,

Brendan Eich (1:45:32.340)

you shouldn't have used it, we hate you.

Lex Fridman (1:45:33.700)

And then we renamed it to Firefox.

Lex Fridman (1:45:36.100)

And they're like, ah, we love you.

Lex Fridman (1:45:38.060)

And then I haven't heard of them ever since.

Lex Fridman (1:45:40.140)

But Firefox was a clever name.

Lex Fridman (1:45:41.780)

We had to think of something distinctive.

Brendan Eich (1:45:43.660)

We wanted to keep the fire going.

Lex Fridman (1:45:45.740)

And it turns out there's a red panda, right?

Brendan Eich (1:45:48.180)

That's the nickname for it.

Lex Fridman (1:45:49.460)

So that's the second set of browser wars.

Lex Fridman (1:45:51.980)

So how was Firefox born, how was Mozilla born?

Lex Fridman (1:45:59.660)

There's a long story there too.

Lex Fridman (1:46:01.060)

So Netscape got acquired by AOL,

Lex Fridman (1:46:02.980)

which as I say was a reasonable happy ending

Brendan Eich (1:46:05.500)

for a lot of people,

Lex Fridman (1:46:06.580)

because Netscape otherwise was gonna go out of business

Brendan Eich (1:46:08.500)

because Microsoft was just killing its market.

Lex Fridman (1:46:10.740)

There was no way to charge for a browser.

Brendan Eich (1:46:12.540)

Windows came with IE, IE4 was pretty good

Lex Fridman (1:46:15.980)

and Netscape 4 wasn't that good.

Brendan Eich (1:46:17.580)

It took a while to get better.

Lex Fridman (1:46:20.180)

But the Netscape executive said,

Brendan Eich (1:46:22.340)

let's do an open source escape pod.

Lex Fridman (1:46:25.660)

And like in Star Wars, A New Hope,

Brendan Eich (1:46:27.180)

the gunner won't shoot it

Lex Fridman (1:46:28.340)

because there's no life forms on board, right?

Brendan Eich (1:46:30.180)

It's not a threat.

Lex Fridman (1:46:31.460)

And so we did Mozilla in 1998

Lex Fridman (1:46:33.740)

and it looked like it was going to initially

Lex Fridman (1:46:37.980)

just give the world an open source browser.

Lex Fridman (1:46:40.140)

But it's really hard to get people to work

Lex Fridman (1:46:42.020)

on this sort of hairball that had been hacked up

Brendan Eich (1:46:43.900)

over by that point four years.

Lex Fridman (1:46:46.100)

And it also couldn't have the crypto module

Brendan Eich (1:46:49.140)

for secure sockets, so called,

Lex Fridman (1:46:51.020)

or now transport layer security.

Brendan Eich (1:46:52.980)

That was an electronic munition.

Lex Fridman (1:46:54.380)

We were not allowed to release that

Brendan Eich (1:46:56.100)

in the full 1024 bit key strength.

Lex Fridman (1:47:00.460)

And yet people, one of whom I happened to meet previously

Brendan Eich (1:47:05.100)

at SGI when I went on a sales support engineering trip,

Lex Fridman (1:47:08.980)

Tim Hudson in Brisbane, Australia,

Lex Fridman (1:47:11.100)

and Eric A. Young did what became open SSL.

Lex Fridman (1:47:16.460)

It was called SSL EAY after Eric's initials.

Lex Fridman (1:47:19.500)

And Tim and Eric took their open SSL

Lex Fridman (1:47:23.180)

outside of the purview of the NSA

Lex Fridman (1:47:25.420)

and the Department of Commerce,

Lex Fridman (1:47:27.300)

and they stuck it into Mozilla's code.

Lex Fridman (1:47:28.940)

And that was perhaps the best hack that was done

Lex Fridman (1:47:31.060)

in the first few months after we open sourced the browser.

Brendan Eich (1:47:34.540)

We had other problems.

Lex Fridman (1:47:35.420)

The politics inside Netscape were riven

Brendan Eich (1:47:37.420)

by these acquisitions.

Lex Fridman (1:47:38.340)

So the one acquisition that kind of messed up Netscape

Brendan Eich (1:47:41.020)

for also wanted to keep doing a proprietary mail

Lex Fridman (1:47:44.340)

and groupware program, not Jamie Zawinski's mail program

Brendan Eich (1:47:48.860)

that was in Netscape two and three.

Lex Fridman (1:47:50.180)

And they held it back from open source.

Brendan Eich (1:47:51.860)

We didn't have a mail program, it was just a browser.

Lex Fridman (1:47:54.320)

We didn't know what AOL will do to us.

Brendan Eich (1:47:56.700)

Turns out they didn't interfere with us for a long time.

Lex Fridman (1:47:59.260)

But Netscape wasn't the best steward of Mozilla.

Brendan Eich (1:48:02.980)

We were operating Mozilla as a pirate ship

Lex Fridman (1:48:05.340)

without a legal entity.

Lex Fridman (1:48:06.420)

So most of us worked for Netscape

Lex Fridman (1:48:08.900)

under a separate organization.

Lex Fridman (1:48:10.900)

And initially the first engineering manager,

Lex Fridman (1:48:15.400)

Tom Paquin of Netscape was the Mozilla founding manager.

Lex Fridman (1:48:19.360)

But he left pretty quickly

Lex Fridman (1:48:20.820)

and he left me as the acting manager,

Brendan Eich (1:48:23.120)

which is more like method acting in my case.

Lex Fridman (1:48:25.500)

And that was my first management stint.

Lex Fridman (1:48:28.780)

But then someone who'd written the licenses,

Lex Fridman (1:48:32.140)

Mitchell Baker, she was a lawyer at Netscape.

Brendan Eich (1:48:34.100)

She was involved in the open source license decision making

Lex Fridman (1:48:38.180)

and the actual writing and construction of those licenses.

Brendan Eich (1:48:40.380)

That was Mitchell's job, Netscape public license

Lex Fridman (1:48:42.980)

and the truly open Mozilla public license.

Lex Fridman (1:48:45.700)

And there were two because Netscape needed,

Lex Fridman (1:48:47.380)

because of some encumbered code, needed some special rights

Lex Fridman (1:48:49.580)

but that went away over time.

Lex Fridman (1:48:51.100)

Mitchell was always interested in Mozilla

Lex Fridman (1:48:52.820)

and she came back from maternity leave

Lex Fridman (1:48:54.220)

and she said, I'll be the manager if you want.

Lex Fridman (1:48:55.900)

And Jamie and I said, sure.

Lex Fridman (1:48:57.420)

And then Jamie quit, he quit after a year.

Brendan Eich (1:48:59.380)

He said, this didn't work, I'm sorry.

Lex Fridman (1:49:02.580)

He acted like it was a total failure

Brendan Eich (1:49:04.340)

because Mozilla didn't restart the browser market.

Lex Fridman (1:49:08.100)

But there's no way it could have, right?

Brendan Eich (1:49:09.260)

Netscape was still shipping variants of Netscape 4,

Lex Fridman (1:49:13.740)

which was based on the old code.

Brendan Eich (1:49:15.700)

Mozilla was trying to react to the code

Lex Fridman (1:49:18.360)

to make greenfield for developers.

Lex Fridman (1:49:19.860)

So it was one of my big goals.

Lex Fridman (1:49:20.900)

It wasn't a technical goal so much as again, a social goal.

Brendan Eich (1:49:24.180)

People wanted a more standard spaced browser.

Lex Fridman (1:49:26.380)

They wanted less of a hairball that had been hacked on

Brendan Eich (1:49:29.380)

by ex grad students starting four years prior.

Lex Fridman (1:49:32.020)

So we said, we're gonna make a modular code base.

Brendan Eich (1:49:35.100)

We're gonna use a variant or an open source version

Lex Fridman (1:49:37.500)

of Microsoft's component object model,

Brendan Eich (1:49:39.980)

has reference counting and standardized V tables,

Lex Fridman (1:49:43.420)

virtual calls and C++.

Lex Fridman (1:49:45.180)

And we're gonna use JavaScript.

Lex Fridman (1:49:47.300)

We're gonna have a bridge between those two

Lex Fridman (1:49:48.700)

so you can script those components

Lex Fridman (1:49:50.140)

just like Java components.

Brendan Eich (1:49:52.780)

We're going to make a portable front end

Lex Fridman (1:49:54.800)

with a markup language for the user interface.

Brendan Eich (1:49:57.220)

Not tables, not HTML, but custom menus

Lex Fridman (1:50:00.380)

and dropdowns and toolbars.

Lex Fridman (1:50:02.880)

And that was called Zool, XML user interface language.

Lex Fridman (1:50:06.060)

And some real talent on the Netscape side delivered that.

Brendan Eich (1:50:08.220)

Dave Hyatt, who was instrumental in Zool,

Lex Fridman (1:50:11.320)

Chris Watterson, Joe Hewitt, Blake Ross.

Lex Fridman (1:50:15.700)

And Blake was an intern.

Lex Fridman (1:50:17.820)

He was like a high school aged intern at Netscape.

Lex Fridman (1:50:20.500)

And at some point we were innovating rapidly

Lex Fridman (1:50:23.920)

in the Mozilla world and Netscape was still caught up

Brendan Eich (1:50:26.960)

in this management mess from these acquisitions

Lex Fridman (1:50:29.340)

and it wasn't delivering.

Lex Fridman (1:50:30.420)

And every year they were wondering if ALO was gonna come

Lex Fridman (1:50:32.800)

and start beheading the executives

Brendan Eich (1:50:34.260)

because it didn't do anything useful.

Lex Fridman (1:50:36.100)

And there was this thought

Brendan Eich (1:50:36.940)

you should take the Netscape browser engine

Lex Fridman (1:50:38.940)

and put it in the Windows ALO client,

Brendan Eich (1:50:40.740)

which was the dial up client

Lex Fridman (1:50:41.980)

that all the increasingly aging users of ALO were using.

Brendan Eich (1:50:45.420)

Never happened.

Lex Fridman (1:50:46.260)

It would have been too big a change.

Lex Fridman (1:50:48.140)

So it wasn't clear why ALO bought Netscape,

Lex Fridman (1:50:49.820)

but as I said, they left it alone.

Lex Fridman (1:50:50.940)

But Netscape didn't leave Mozilla alone.

Lex Fridman (1:50:53.300)

And so in 2001, Mitchell called me up and said,

Brendan Eich (1:50:58.780)

I'm no longer employed.

Lex Fridman (1:50:59.820)

And I was like, what?

Lex Fridman (1:51:00.660)

You quit?

Lex Fridman (1:51:01.480)

No, no, this wasn't my choice.

Lex Fridman (1:51:02.320)

And there was a layoff which maybe accidentally

Lex Fridman (1:51:05.300)

or on purpose got rid of Mitchell.

Lex Fridman (1:51:06.820)

But the funny thing was we had an open source project.

Lex Fridman (1:51:08.660)

We had a lot of the engineers on staff on our side

Lex Fridman (1:51:11.220)

and we had people we'd hired through the Mozilla community

Lex Fridman (1:51:15.220)

who were top notch.

Brendan Eich (1:51:16.300)

They'd risen, they came in high quality, they knew the code

Lex Fridman (1:51:19.260)

and they actually were better than the average

Brendan Eich (1:51:21.660)

or median hire of Netscape.

Lex Fridman (1:51:23.340)

And so the funny thing was the executive

Brendan Eich (1:51:26.940)

who thought they'd gotten rid of Mitchell in the layoff

Lex Fridman (1:51:29.940)

on the next week's community call around Mozilla

Lex Fridman (1:51:32.700)

and what to do, there's Mitchell.

Lex Fridman (1:51:35.220)

And so this showed you can kind of transcend

Brendan Eich (1:51:36.820)

your boundaries of corporate open source

Lex Fridman (1:51:40.420)

if you get a project that has enough loyalty,

Brendan Eich (1:51:42.700)

even among the paid staff.

Lex Fridman (1:51:43.820)

Because we had outside people contributing.

Brendan Eich (1:51:45.580)

We had people at Red Hat and a few other places,

Lex Fridman (1:51:47.980)

but the majority of the hackers were employed by Netscape.

Lex Fridman (1:51:50.620)

But a lot of them at that point had come from the community

Lex Fridman (1:51:53.380)

and others got the community and wanted to work with it.

Lex Fridman (1:51:55.820)

And it was really the weakest engineers at Netscape

Lex Fridman (1:51:57.900)

who didn't like Mozilla and didn't like the crucible

Brendan Eich (1:52:01.100)

of competing with the better programmers.

Lex Fridman (1:52:04.220)

So if the project is good enough, it will rise,

Brendan Eich (1:52:06.640)

the Phoenix will rise out of the...

Lex Fridman (1:52:08.220)

That's exactly right.

Lex Fridman (1:52:09.060)

And so we had this Mozilla code base

Lex Fridman (1:52:10.880)

that was getting better.

Brendan Eich (1:52:11.720)

In fact, I think at some point in 2002

Lex Fridman (1:52:13.540)

when we declared Mozilla 1.0, I engineered a roadmap

Brendan Eich (1:52:17.380)

that successively through similar sort of six week,

Lex Fridman (1:52:20.100)

five week releases, like we all do with browser releases

Brendan Eich (1:52:23.000)

nowadays, Chrome does and Firefox braved us three weeks.

Lex Fridman (1:52:26.460)

We got to a point where we said, you know what?

Brendan Eich (1:52:29.700)

It doesn't suck.

Lex Fridman (1:52:30.540)

This is like the 1.0 that you want to release

Brendan Eich (1:52:33.240)

because if you hold it back any longer to polish it,

Lex Fridman (1:52:35.340)

you're denying others the ability to use it.

Brendan Eich (1:52:36.980)

It's like pro engineer, the mechanical CAD tool

Lex Fridman (1:52:39.400)

embedded the code, they embedded the layout engine.

Lex Fridman (1:52:43.660)

And Mozilla 1.0 was like a Netscape communication suite.

Lex Fridman (1:52:46.980)

We had at that point gotten male people

Brendan Eich (1:52:49.560)

to reintegrate mail and news and we had an editor for HTML.

Lex Fridman (1:52:53.640)

And it felt like a 90s suite, suiteware.

Lex Fridman (1:52:57.620)

And it felt kind of bloated.

Lex Fridman (1:52:58.800)

And the people who were taking that Mozilla open source

Lex Fridman (1:53:02.140)

and then adding Netscape flavor to it

Lex Fridman (1:53:05.020)

were not calling the shots right.

Lex Fridman (1:53:06.540)

And they were also under AOL's thumb a little bit

Lex Fridman (1:53:08.460)

and that they said, well, we should probably put

Brendan Eich (1:53:10.540)

the AOL instant messenger chicklet on the toolbar.

Lex Fridman (1:53:13.180)

We should put the ICQ, the other messaging system

Brendan Eich (1:53:16.220)

that AOL had acquired.

Lex Fridman (1:53:17.340)

We should put the ICQ button on the toolbar.

Lex Fridman (1:53:19.580)

And pretty soon Netscape looked like a bit of a NASCAR

Lex Fridman (1:53:23.500)

badged version of Mozilla.

Lex Fridman (1:53:24.900)

And that also made Mozilla more popular.

Lex Fridman (1:53:27.180)

And yet they had contrived to fire or lay off the leader

Lex Fridman (1:53:35.460)

and we carried on with an open source structure

Lex Fridman (1:53:38.540)

where Mozilla was still, you know, Mitchell was calling

Brendan Eich (1:53:40.860)

sort of management or project level shots

Lex Fridman (1:53:44.480)

and I was calling technical shots.

Lex Fridman (1:53:46.100)

And we had a popular suite, but we thought,

Lex Fridman (1:53:51.200)

why not make it just a browser?

Brendan Eich (1:53:52.740)

Because it'll be simpler, it'll do one job well.

Lex Fridman (1:53:55.380)

And even then we can strip it down by having extensions.

Lex Fridman (1:53:58.780)

So Dave Hyatt and Blake Ross, the high school aged intern

Lex Fridman (1:54:03.740)

did the first version, which was called Mozilla slash browser.

Brendan Eich (1:54:07.260)

It was very, the small group of us, Ian Hicks

Lex Fridman (1:54:09.220)

and Asa Dotzler, me and Joe Hewitt and Hyatt and Blake.

Lex Fridman (1:54:14.340)

And Hyatt was really the senior hacker.

Lex Fridman (1:54:16.540)

He'd done all these things like amazing cross platform menus

Brendan Eich (1:54:20.500)

through the user interface, markup language.

Lex Fridman (1:54:23.500)

And he knew how to do tab browsing.

Brendan Eich (1:54:24.740)

He'd implemented it natively on Mac OS at the time

Lex Fridman (1:54:28.260)

in Camino originally called Chimera.

Brendan Eich (1:54:31.740)

He'd written multiple implementations,

Lex Fridman (1:54:34.300)

which was a thing programmers should do.

Brendan Eich (1:54:36.340)

It's like the V8 team did for those missing years

Lex Fridman (1:54:38.440)

when the rest of the Chrome team's like, where's V8?

Brendan Eich (1:54:41.980)

In fact, Dave's wife, Rebecca told me a story

Lex Fridman (1:54:44.440)

about when they were at UIUC,

Brendan Eich (1:54:45.860)

they were also University of Illinois grad students.

Lex Fridman (1:54:49.460)

There was an assignment, it was a programming assignment

Brendan Eich (1:54:51.300)

it was supposed to do at the end of the semester.

Lex Fridman (1:54:53.300)

And Dave's friend was this, I'm gonna go think

Lex Fridman (1:54:56.820)

and I'm gonna design and I'm gonna make this

Lex Fridman (1:54:59.460)

platonic perfect form of the program.

Lex Fridman (1:55:01.300)

And then I'm gonna write it at the end when it's due.

Lex Fridman (1:55:03.580)

And Hyatt just went there and started hacking.

Brendan Eich (1:55:05.140)

He wrote one version, he wrote a second version,

Lex Fridman (1:55:06.660)

a third version, end of the semester comes around.

Brendan Eich (1:55:09.220)

The friend's not doing too well.

Lex Fridman (1:55:11.140)

It wasn't perfect and it wasn't written.

Brendan Eich (1:55:13.000)

I'm not sure how that story ended for him,

Lex Fridman (1:55:14.820)

but Dave's version was a fifth iteration, it was great.

Lex Fridman (1:55:18.100)

And so he'd done that with everything you need

Lex Fridman (1:55:20.700)

in a tabbed browser.

Lex Fridman (1:55:21.700)

And this really showed well in Phoenix,

Lex Fridman (1:55:24.360)

what we called Phoenix and I had to rename two more times.

Lex Fridman (1:55:28.120)

And Blake went to Stanford, he became a Stanford student

Lex Fridman (1:55:32.260)

and couldn't work on it.

Brendan Eich (1:55:34.060)

Dave Hyatt went to Apple in 2001.

Lex Fridman (1:55:36.620)

He was one of the founding Safari team members.

Brendan Eich (1:55:39.580)

Interesting, wow.

Lex Fridman (1:55:40.460)

But he was still blogging about tabbed browsing.

Brendan Eich (1:55:43.540)

I think Apple at some point said you should.

Lex Fridman (1:55:44.860)

Safari have tabbed browsing?

Brendan Eich (1:55:46.140)

Yeah, but it was because of Hyatt.

Lex Fridman (1:55:48.220)

Hyatt was quite a feather in their cap.

Brendan Eich (1:55:49.820)

Don Melton, who had been the engineering manager

Lex Fridman (1:55:53.420)

for Safari from the beginning, had been in Netscape also.

Lex Fridman (1:55:56.540)

And so there's this diaspora of talent

Lex Fridman (1:55:58.980)

and yet Hyatt was still kind of writing blog posts

Brendan Eich (1:56:01.620)

about how to do tabs right.

Lex Fridman (1:56:03.260)

And at some point Apple said, don't blog about that.

Brendan Eich (1:56:05.360)

That's our proprietary tab technology.

Lex Fridman (1:56:07.580)

And I was like, no, it's not.

Brendan Eich (1:56:08.620)

It was an opera and I've refined it.

Lex Fridman (1:56:10.580)

So we had to replace people and we had Ben Goodger,

Brendan Eich (1:56:15.420)

a New Zealander we hired at Netscape.

Lex Fridman (1:56:18.020)

And he stepped in to be the Firefox lead.

Lex Fridman (1:56:21.260)

And we also had this weird circumstance

Lex Fridman (1:56:23.340)

where AOL finally did notice that Netscape

Brendan Eich (1:56:26.900)

was kind of an albatross,

Lex Fridman (1:56:28.380)

that they bought it for no particular benefit.

Lex Fridman (1:56:31.060)

And even then the AOL politics were also heinous,

Lex Fridman (1:56:33.820)

sort of East Coast politics.

Brendan Eich (1:56:35.380)

I remember taking two trips there

Lex Fridman (1:56:36.660)

because I was a principal engineer.

Lex Fridman (1:56:37.820)

And so us principal engineers got trotted out

Lex Fridman (1:56:39.560)

to do dog and pony shows in Dallas, Virginia.

Lex Fridman (1:56:42.740)

And the AOL opera management was very East Coast in flavor.

Lex Fridman (1:56:46.400)

And they were at that time merging with Time Warner,

Brendan Eich (1:56:49.100)

which did not go well.

Lex Fridman (1:56:50.500)

So one of these years we went out there

Lex Fridman (1:56:52.020)

and we were all doing dog and pony shows

Lex Fridman (1:56:53.420)

and there were these characters

Brendan Eich (1:56:55.100)

that were sort of like marketing guys.

Lex Fridman (1:56:56.340)

One of them was wearing a cravat

Lex Fridman (1:56:57.700)

and one was named Reggie.

Lex Fridman (1:56:59.460)

And they were very you rather than non you.

Brendan Eich (1:57:05.940)

Or they were like what's what's Stoneman's

Lex Fridman (1:57:07.480)

metropolitan film, UHB, urban haute bourgeoisie.

Brendan Eich (1:57:13.980)

They were haute bourgeoisie.

Lex Fridman (1:57:15.820)

They were funny and they were kind of useless

Lex Fridman (1:57:19.100)

and kind of preppy.

Lex Fridman (1:57:20.140)

And then the next year we went back

Lex Fridman (1:57:21.460)

and I said, where's Reggie?

Lex Fridman (1:57:22.340)

And it's like, oh, Reggie's not here anymore

Brendan Eich (1:57:24.140)

because Time Warner realized that the merger

Lex Fridman (1:57:26.300)

wasn't in their interest either.

Lex Fridman (1:57:27.520)

And then the sort of knives came out.

Lex Fridman (1:57:29.140)

And these mergers rarely work, right?

Brendan Eich (1:57:31.820)

This is very difficult.

Lex Fridman (1:57:32.680)

You get these giant companies

Lex Fridman (1:57:33.820)

and they think there's gonna be synergy.

Lex Fridman (1:57:35.300)

That was the 90s, late 90s watch word.

Lex Fridman (1:57:37.540)

And there wasn't synergy with AOL buying Netscape

Lex Fridman (1:57:39.340)

and there wasn't synergy with Time Warner and AOL.

Lex Fridman (1:57:41.260)

But did AOL ever really work?

Lex Fridman (1:57:43.020)

Was it ever really cool?

Brendan Eich (1:57:44.260)

Like the same kind of fire and excitement

Lex Fridman (1:57:46.460)

that Firefox eventually created,

Lex Fridman (1:57:48.740)

was that ever there in AOL?

Lex Fridman (1:57:51.100)

AOL was the right time to do a dial up service

Brendan Eich (1:57:55.020)

that got distribution by basically

Lex Fridman (1:57:56.660)

leaflet bombing compact discs on the country.

Lex Fridman (1:58:00.700)

And they beat out CompuServe and the other ones,

Lex Fridman (1:58:04.420)

Prodigy, and then the web happened.

Lex Fridman (1:58:06.860)

And so you had almost like this isolated continent,

Lex Fridman (1:58:10.940)

like some of the evolutionary biologists I follow

Brendan Eich (1:58:14.180)

make fun of the funny large marsupial mammals of Australia,

Lex Fridman (1:58:19.900)

how silly they are.

Lex Fridman (1:58:21.500)

And so AOL is like Australia.

Lex Fridman (1:58:23.100)

Yeah.

Lex Fridman (1:58:23.940)

And you saw it over time because they kept aging

Lex Fridman (1:58:26.420)

and they were using AOL to get online

Lex Fridman (1:58:28.180)

and they couldn't really use a web browser.

Lex Fridman (1:58:30.440)

And it became sort of a valued cohort

Brendan Eich (1:58:33.380)

because they still have relatively high

Lex Fridman (1:58:35.140)

socioeconomic status and they have grandchildren,

Lex Fridman (1:58:37.200)

but it's going away, it's dying at some point.

Lex Fridman (1:58:39.500)

Towards the end of the aughts, that decade,

Lex Fridman (1:58:42.100)

and then to the decade 2010 plus,

Lex Fridman (1:58:45.980)

that Firefox became this incredible,

Brendan Eich (1:58:48.620)

I forget when Chrome came out, but.

Lex Fridman (1:58:50.260)

2008, September.

Brendan Eich (1:58:51.100)

2008, but Firefox was the sexy cool thing

Lex Fridman (1:58:55.060)

that represented a lot of the cutting edge technologies

Lex Fridman (1:58:57.500)

and all that kind of stuff.

Lex Fridman (1:58:58.340)

Web 2, it was amazing.

Brendan Eich (1:59:00.140)

Kim O'reilly and John Battelle did the first Web 2 conference

Lex Fridman (1:59:03.380)

which eventually became huge and they split it.

Lex Fridman (1:59:05.420)

But that was in 2004, it was right when Firefox was out.

Lex Fridman (1:59:08.540)

Craigslist was huge, it was killing classified revenue

Brendan Eich (1:59:11.640)

for newspapers, but there was just this ferment.

Lex Fridman (1:59:14.420)

People starting.

Brendan Eich (1:59:15.260)

Wikipedia along there somewhere.

Lex Fridman (1:59:16.660)

Gmail was already done and it was an impressive web mail.

Brendan Eich (1:59:19.820)

There were others before it like Hotmail,

Lex Fridman (1:59:21.140)

but Gmail was really impressive from Google.

Lex Fridman (1:59:22.860)

And Google Maps, people started seeing what could be done.

Lex Fridman (1:59:26.020)

They thought how can you drag the map around

Lex Fridman (1:59:28.100)

and how does that work?

Lex Fridman (1:59:29.700)

And it was all JavaScript and images and.

Lex Fridman (1:59:32.460)

So Gmail was 2003, four?

Lex Fridman (1:59:34.500)

Yeah, it actually started quite early.

Brendan Eich (1:59:36.220)

It might've been 2002 or three,

Lex Fridman (1:59:37.340)

but by the time we started dealing with Google and Firefox

Brendan Eich (1:59:40.100)

to get the search deal,

Lex Fridman (1:59:41.040)

which was the main revenue source for Mozilla,

Lex Fridman (1:59:43.420)

and still is, 2004, early, Sergey Brinz,

Lex Fridman (1:59:47.740)

one of his trusted engineer guys, Fritz Schneider,

Brendan Eich (1:59:50.420)

made contact with me at Mozilla and we started talking

Lex Fridman (1:59:53.380)

and we realized search and browser need each other.

Lex Fridman (1:59:56.980)

And this is deeply true, right?

Lex Fridman (1:59:58.460)

This is still true.

Brendan Eich (1:59:59.780)

This is why a lot of the search engines

Lex Fridman (20:03.020)

in how many days and why was it only that many days?

Lex Fridman (20:05.860)

And what was the goal and the underlying principles

Lex Fridman (20:08.880)

in your mind at the time?

Lex Fridman (20:09.720)

So the whole, I'm sort of describing worse is better

Lex Fridman (20:11.740)

in a frenetic way because it fit the model of Netscape.

Brendan Eich (20:14.980)

When it was known that Jim Clark

Lex Fridman (20:18.420)

and Marc Inves were founding Netscape

Lex Fridman (20:20.020)

and they did the first release in 1994,

Lex Fridman (20:22.880)

that browser took over from Mosaic.

Brendan Eich (20:24.820)

In fact, that's why Mozilla is called that.

Lex Fridman (20:27.100)

It's the Mosaic killer.

Brendan Eich (20:28.260)

It's like the giant monster that kills Mosaic.

Lex Fridman (20:30.140)

That's awesome.

Lex Fridman (20:30.980)

And they knew they could, it wasn't that, again,

Lex Fridman (20:34.180)

it's not like you're doing advanced scientific research

Brendan Eich (20:36.220)

that is changing the world.

Lex Fridman (20:37.380)

You're more like taking down the last iteration

Brendan Eich (20:41.020)

on the browser, Marc did, which had images

Lex Fridman (20:43.180)

and other importances before he stopped working on it.

Lex Fridman (20:45.900)

And you're making Netscape the new thing that has images,

Lex Fridman (20:49.060)

plugins, which was the way to do video back in the day.

Brendan Eich (20:51.720)

It had something that's kind of died now for tiled windows

Lex Fridman (20:54.260)

called frames and frame sets.

Brendan Eich (20:56.000)

HTML tables, that was new.

Lex Fridman (20:59.280)

Eric Bina did tables in Netscape 1.1.

Lex Fridman (21:01.640)

So when I got there, they were heading toward IPO.

Lex Fridman (21:05.040)

Clark wanted the IPO early, I think his instinct was right.

Lex Fridman (21:07.640)

And that kicked off the whole dot com era, right?

Lex Fridman (21:10.280)

There was a recession in the US in 91.

Brendan Eich (21:13.280)

You can see old law and order reruns

Lex Fridman (21:15.140)

where they talk about the recession

Lex Fridman (21:16.520)

and how hard it's hitting New Yorkers.

Lex Fridman (21:18.080)

And after that, Greenspan really goosed things

Brendan Eich (21:20.440)

at the Federal Reserve and technology had been sort of

Lex Fridman (21:23.840)

fermenting in a way that came together with the internet.

Lex Fridman (21:26.480)

And Netscape made it possible to do pets.com,

Lex Fridman (21:30.040)

to do eBay, to get people to recognize a URL on a billboard

Lex Fridman (21:34.280)

and then type it in when they get home.

Lex Fridman (21:36.040)

And that was huge.

Brendan Eich (21:38.280)

That was so fast moving a rocket that Marc

Lex Fridman (21:42.120)

and the engineering team there thought,

Brendan Eich (21:45.060)

we need to make this a programmable browser,

Lex Fridman (21:47.920)

not just a document viewer, not just a video.

Brendan Eich (21:50.840)

It was all HTML with images and tables and also,

Lex Fridman (21:54.120)

like you said, frames.

Brendan Eich (21:55.320)

There was no dynamic element at all.

Lex Fridman (21:57.800)

Yeah, the most dynamism we get was from a plugin,

Brendan Eich (21:59.680)

which there are a few of them then.

Lex Fridman (22:01.640)

Flash didn't exist at that point.

Brendan Eich (22:03.960)

It was, I think.

Lex Fridman (22:05.960)

Java Applets yet or no?

Brendan Eich (22:07.400)

Well, that's the thing we did to deal with Sun.

Lex Fridman (22:09.440)

In fact, I was recruited to go do Scheme in the browser.

Brendan Eich (22:12.600)

Remember Guy Steele and Gerald Sussman's

Lex Fridman (22:14.800)

beautiful Lisp variant?

Brendan Eich (22:16.760)

I was gonna do it in the browser

Lex Fridman (22:17.920)

because my friends from SG, I thought,

Brendan Eich (22:19.480)

hey, we like Scheme, you like Scheme.

Lex Fridman (22:20.920)

And I'm like, I hardly ever use Scheme.

Brendan Eich (22:22.140)

It's not really used in industry,

Lex Fridman (22:23.360)

except in sort of silos, but I like it.

Brendan Eich (22:27.080)

Okay, I'll come do Scheme in the browser.

Lex Fridman (22:29.760)

I have a slide from my 2017 talk

Brendan Eich (22:32.720)

where I have Bruce Willis crawling through the duct

Lex Fridman (22:34.640)

in Die Hard.

Brendan Eich (22:35.480)

He's like, come out to the coast, have a lot of fun.

Lex Fridman (22:38.300)

Come on, do Scheme in the browser.

Lex Fridman (22:40.560)

But when I got there, there was no Scheme in the browser

Lex Fridman (22:42.160)

because they'd started a deal with Sun Microsystems.

Lex Fridman (22:45.240)

And my best contact there was Bill Joy,

Lex Fridman (22:47.040)

who I admired as a Berkeley Unix founder.

Brendan Eich (22:48.960)

And, you know, Sun founder.

Lex Fridman (22:50.600)

And Bill got the idea of making the browser programmable too.

Lex Fridman (22:53.920)

And so the main idea was to put the Java VM,

Lex Fridman (22:57.240)

which at that point was not really easy to embed,

Brendan Eich (23:00.400)

into Netscape, including the Netscape version on Windows

Lex Fridman (23:04.280)

that was still most popular,

Brendan Eich (23:05.660)

which was the 16 bit Windows 3.1, which was going away.

Lex Fridman (23:09.560)

Microsoft was coming out with Windows 95

Lex Fridman (23:11.360)

and everyone was afraid they were gonna do

Lex Fridman (23:13.620)

Internet Explorer, I guess, two at that point,

Brendan Eich (23:16.200)

three the next year.

Lex Fridman (23:17.080)

They already bought or invested in somehow Spyglass,

Brendan Eich (23:20.960)

this other company that shot out from NCSA

Lex Fridman (23:24.440)

at University of Illinois.

Lex Fridman (23:26.000)

And in fact, Microsoft had tried to buy Netscape

Lex Fridman (23:29.740)

in late 94, before I got there.

Lex Fridman (23:31.400)

And I heard about this later.

Lex Fridman (23:32.280)

I heard they offered way too little money.

Lex Fridman (23:35.040)

And so, you know, Jim Barksdale and Jim Clark said,

Lex Fridman (23:37.480)

get out of here, you know, pound sand.

Lex Fridman (23:39.820)

But then they realized, oh, this is going to hurt us

Lex Fridman (23:43.780)

because now they're gonna copy us.

Brendan Eich (23:45.520)

Didn't happen right away.

Lex Fridman (23:46.340)

I'm not sure when Gates internet title wave memo

Brendan Eich (23:49.080)

was written.

Lex Fridman (23:50.240)

That's the famous memo he wrote when Bill Gates realized

Brendan Eich (23:53.180)

that Microsoft was going down this old copy AOL path

Lex Fridman (23:56.840)

or copy CompuServe path, a project called Blackbird,

Brendan Eich (24:01.100)

presumably after the SR71, I don't know.

Lex Fridman (24:03.240)

But they were gonna make a, you know,

Brendan Eich (24:04.680)

dial up service with a custom content language stack

Lex Fridman (24:07.600)

and custom rendering.

Brendan Eich (24:09.040)

It wasn't the web.

Lex Fridman (24:10.680)

You know, they could have content partners.

Brendan Eich (24:13.200)

They have a lot of money,

Lex Fridman (24:14.400)

but it still wasn't to scale the web.

Brendan Eich (24:15.920)

It wasn't gonna be compelling.

Lex Fridman (24:16.880)

And Gates realized this,

Lex Fridman (24:18.200)

and he turned the company on a dime

Lex Fridman (24:20.000)

and they couldn't buy Netscape.

Brendan Eich (24:22.400)

Again, I'm not sure the timing,

Lex Fridman (24:23.320)

so they decided to copy it.

Lex Fridman (24:24.540)

And once we realized that everybody inside Netscape

Lex Fridman (24:27.400)

felt even more urgency and more of a frenetic mood.

Lex Fridman (24:30.920)

And so my chance to do scheme disappeared

Lex Fridman (24:33.620)

when the Java deal started brewing.

Lex Fridman (24:35.840)

But there was still a chance to do a companion language

Lex Fridman (24:38.800)

to Java because Java was a compiled,

Brendan Eich (24:41.080)

is a compiled language.

Lex Fridman (24:42.620)

It's evolved and improved quite a lot since then too,

Lex Fridman (24:44.960)

but it was for sort of serious advanced programmers

Lex Fridman (24:48.560)

that cost a certain salary or hourly rate.

Lex Fridman (24:51.400)

And people observed, Bill Joy observed,

Lex Fridman (24:53.880)

and Mark Andreessen and I observed

Brendan Eich (24:55.760)

that in a mature stack like Microsoft,

Lex Fridman (24:57.900)

you really benefit from having a scripting language

Brendan Eich (25:00.020)

like Visual Basic,

Lex Fridman (25:01.680)

which became Visual Basic script in IE3,

Lex Fridman (25:03.680)

but didn't take over and kill JavaScript,

Lex Fridman (25:07.120)

that you need two languages.

Brendan Eich (25:08.340)

One is for the component writers

Lex Fridman (25:10.160)

who are higher price and more expert.

Lex Fridman (25:12.560)

And the other is for scripters,

Lex Fridman (25:17.000)

certified public accountants, designers,

Brendan Eich (25:19.480)

graphic designers with some programming inclination,

Lex Fridman (25:21.720)

anybody, amateurs, doesn't matter.

Brendan Eich (25:23.460)

There's a much more demotic approach there

Lex Fridman (25:26.600)

for programming the components together,

Brendan Eich (25:28.660)

gluing them together.

Lex Fridman (25:29.880)

Some people say duct tape language, which I don't really like.

Lex Fridman (25:32.800)

But we saw, Bill Joy and Mark Andreessen and I,

Lex Fridman (25:36.080)

we saw the need for a companion language.

Lex Fridman (25:38.020)

And the gleam in our eye was to call it JavaScript.

Lex Fridman (25:40.400)

I didn't like it, that was marketing's plan.

Brendan Eich (25:42.480)

Mark called it Mocha, which I liked.

Lex Fridman (25:45.080)

And Netscape Marketing, I think, didn't like that.

Lex Fridman (25:47.080)

So they said, oh, there's some trademark

Lex Fridman (25:48.520)

and some software somewhere that uses Mocha,

Lex Fridman (25:50.800)

so we can't use that.

Lex Fridman (25:52.000)

And they tried LiveScript in August and that didn't last.

Lex Fridman (25:54.960)

And then finally we got the trademark license

Lex Fridman (25:56.680)

in December 1995.

Lex Fridman (25:58.720)

But the work I did to prove that it could be done

Lex Fridman (26:01.600)

was important because I came in in April

Lex Fridman (26:05.200)

and even then Netscape was growing so fast

Lex Fridman (26:08.320)

that they couldn't find an open hiring requisition

Brendan Eich (26:11.280)

in the client team for me.

Lex Fridman (26:12.320)

So they hired me into the server team.

Lex Fridman (26:14.280)

And I worked for a month on server team

Lex Fridman (26:16.680)

on what became HTTP 1.1.

Lex Fridman (26:18.600)

So I was actually, I had done protocol work

Lex Fridman (26:20.280)

at Silicon Graphics with Greg Chesson,

Brendan Eich (26:22.640)

former Bell Labs intern, grad student intern

Lex Fridman (26:25.360)

who knew all the Unix founders.

Lex Fridman (26:27.020)

And Greg was very interested in taking protocols

Lex Fridman (26:30.880)

to the next level with VLSI,

Brendan Eich (26:32.840)

because he thought that CPUs wouldn't scale up.

Lex Fridman (26:35.640)

He was mistaken in that, unfortunately.

Brendan Eich (26:37.480)

Moore's law more than kept up.

Lex Fridman (26:39.060)

And you have gigabit ethernet running

Brendan Eich (26:41.400)

with conventional processors.

Lex Fridman (26:42.520)

But I worked on protocols at SGI

Brendan Eich (26:45.680)

as well as Unix kernel hacking and NFS and things like that.

Lex Fridman (26:49.000)

So I came into Netscape to work on the server side

Brendan Eich (26:53.040)

for a month, but I was thinking the whole time,

Lex Fridman (26:55.260)

what should this language be like?

Lex Fridman (26:56.360)

Should it be easy to use?

Lex Fridman (26:58.480)

Might its syntax even be more like natural language

Brendan Eich (27:00.680)

like HyperTalk, which is Bill Atkinson's language

Lex Fridman (27:04.840)

in HyperCard, if you ever used HyperCard on an early Mac.

Lex Fridman (27:08.560)

And I thought, well, I'd like to do that,

Lex Fridman (27:12.080)

but my management is saying, make it look like Java,

Brendan Eich (27:15.160)

which looks like C from a distance.

Lex Fridman (27:17.480)

What does that mean?

Lex Fridman (27:18.320)

Is it braces?

Lex Fridman (27:19.280)

We're talking about visually?

Brendan Eich (27:20.400)

Does that mean like, management,

Lex Fridman (27:23.060)

do they understand what they think about?

Brendan Eich (27:24.640)

Marketing didn't know, but management did.

Lex Fridman (27:26.560)

Like Rick Schell, the VP of engineering, knew.

Lex Fridman (27:29.680)

And we had a plan even that was,

Lex Fridman (27:31.720)

if you have this companion language,

Brendan Eich (27:32.920)

you're going to glue things together

Lex Fridman (27:34.560)

between Java and JavaScript.

Lex Fridman (27:35.680)

So you're going to have commerce in memory, in the heap

Lex Fridman (27:38.800)

with data types.

Lex Fridman (27:40.240)

So you're going to want some of the data types

Lex Fridman (27:41.760)

in Java to reflect in the JavaScript.

Brendan Eich (27:43.880)

You're going to want the primitive types

Lex Fridman (27:45.700)

that Java unfortunately separated from objects.

Lex Fridman (27:47.540)

So at least some of them, double, let's call it

Lex Fridman (27:50.880)

in Java's terms from the C term

Brendan Eich (27:52.560)

for double precision floating point,

Lex Fridman (27:55.640)

or strings or Booleans and objects.

Lex Fridman (28:01.200)

And so right away there was this constraint

Lex Fridman (28:04.000)

that looking like Java meant kind of a C curly brace syntax

Lex Fridman (28:07.920)

but also some of the data types and objects.

Lex Fridman (28:09.800)

Like objects and so on, all that kind of stuff.

Brendan Eich (28:12.000)

Comparison operator.

Lex Fridman (28:13.000)

Garbage collection, all that stuff.

Brendan Eich (28:15.480)

Even the bitwise operators and the shift operators

Lex Fridman (28:17.640)

including the unsigned right shift,

Brendan Eich (28:19.580)

which Java had because it didn't have

Lex Fridman (28:21.420)

unsigned integer types.

Brendan Eich (28:22.720)

It said, if you want to do unsigned operations,

Lex Fridman (28:24.600)

use an operator.

Lex Fridman (28:25.440)

And that turned out to be important much later.

Lex Fridman (28:27.560)

I'll tell that story five time.

Lex Fridman (28:28.560)

But JavaScript inherited a set of operators,

Lex Fridman (28:35.080)

the expression grammar, the statement grammar

Brendan Eich (28:37.160)

up to a point from Java.

Lex Fridman (28:39.360)

But I wanted a functional language.

Brendan Eich (28:40.720)

I wanted scheme, a little bit of scheme,

Lex Fridman (28:42.840)

even though it wasn't as clean as scheme.

Brendan Eich (28:44.600)

I wanted.

Lex Fridman (28:45.440)

So you had a love, sorry to interrupt.

Brendan Eich (28:46.260)

You had a love for scheme and list

Lex Fridman (28:47.600)

but that functional language landscape.

Brendan Eich (28:49.760)

Yes, I wanted first class functions

Lex Fridman (28:51.640)

because I saw the need for callbacks in the browser

Brendan Eich (28:54.000)

where it's a single threaded program.

Lex Fridman (28:55.880)

All the early browsers were single threaded

Lex Fridman (28:57.360)

and it's the right model for users.

Lex Fridman (28:58.640)

Most users weren't ready for mutual exclusion

Lex Fridman (29:00.600)

and threading.

Lex Fridman (29:01.680)

So in a single threaded world,

Brendan Eich (29:03.680)

you cannot block the user interface.

Lex Fridman (29:05.440)

So you have to use a callback and run later.

Lex Fridman (29:07.520)

And without getting too fancy

Lex Fridman (29:09.920)

and trying to capture the continuation

Brendan Eich (29:11.640)

like call CC does in scheme,

Lex Fridman (29:13.340)

I thought I'll just make it easy to have fun arcs.

Brendan Eich (29:16.640)

First class functions you pass downward

Lex Fridman (29:18.860)

and it can call back, it'd be called back.

Lex Fridman (29:21.800)

And Java didn't have that at the time.

Lex Fridman (29:23.640)

It took forever to get proper first class functions

Brendan Eich (29:27.280)

or lambdas now into Java, Java seven or eight, I think.

Lex Fridman (29:30.720)

It did have concurrency, right?

Brendan Eich (29:32.120)

Yes.

Lex Fridman (29:32.960)

From the very beginning.

Lex Fridman (29:33.780)

But you were thinking that the JavaScript

Lex Fridman (29:35.680)

in the browser would not have the luxury

Brendan Eich (29:38.160)

of being concurrent.

Lex Fridman (29:39.360)

That's right.

Lex Fridman (29:40.200)

And the reason was Java was gonna run in the plugins.

Lex Fridman (29:42.480)

So it could fork threads and go to town.

Lex Fridman (29:44.560)

But the main action in the browser

Lex Fridman (29:46.880)

was in the single threaded program,

Brendan Eich (29:48.280)

the single Unix process on Unix or Windows.

Lex Fridman (29:51.880)

And it was where you had to service the event loop

Lex Fridman (29:54.400)

and then go do things.

Lex Fridman (29:56.160)

Respond to the network layout, some HTML, render it,

Brendan Eich (29:59.360)

turn widths into heights by filling containers, boxes,

Lex Fridman (2:00:01.140)

have their own browsers.

Brendan Eich (2:00:02.240)

Yeah, so in case people don't know,

Lex Fridman (2:00:03.660)

the main revenue source for the browser

Brendan Eich (2:00:05.580)

is the default search engine,

Lex Fridman (2:00:07.240)

which is kind of incredible to think about

Brendan Eich (2:00:09.260)

that that is a revenue source.

Lex Fridman (2:00:11.460)

It's a little bit sad.

Brendan Eich (2:00:12.580)

Yeah, it leads to this capture or kill effect

Lex Fridman (2:00:14.580)

where you have the search engine own its own browser

Lex Fridman (2:00:17.020)

and other browsers may struggle to get the distribution

Lex Fridman (2:00:21.260)

we talked about earlier.

Lex Fridman (2:00:22.620)

So where, and you said you've figured out

Lex Fridman (2:00:26.800)

that Google is working on its own browser

Brendan Eich (2:00:29.100)

at some point there.

Lex Fridman (2:00:29.940)

2006, yeah.

Brendan Eich (2:00:30.780)

2006, so would you say Firefox versus,

Lex Fridman (2:00:33.220)

was Internet Explorer part of the war here

Lex Fridman (2:00:35.380)

or was the Firefox versus Chrome?

Lex Fridman (2:00:37.300)

So Firefox didn't quite cause Microsoft to reconvene IE.

Brendan Eich (2:00:40.900)

They did do IE7 and I remember being on a plane

Lex Fridman (2:00:44.580)

back from the standards meeting,

Brendan Eich (2:00:46.220)

JavaScript standards meeting from Seattle, from Redmond,

Lex Fridman (2:00:48.980)

and there was some Microsoft guy in front of me.

Brendan Eich (2:00:52.100)

Turns out my wife knew him from her past life

Lex Fridman (2:00:54.940)

before we married and he was just this bearded big guy

Lex Fridman (2:00:58.200)

and he was like,

Lex Fridman (2:00:59.040)

we should have just killed Firefox in the cradle.

Brendan Eich (2:01:01.500)

All we needed to do was add pop up blocking in tabs

Lex Fridman (2:01:04.160)

and we could have made Internet Explorer kill Firefox.

Lex Fridman (2:01:06.180)

And it's like, shoulda, coulda, woulda, pal.

Lex Fridman (2:01:07.820)

And I was right behind him during this.

Lex Fridman (2:01:10.300)

But they didn't, they were slow

Lex Fridman (2:01:11.580)

and IE7 wasn't that great.

Lex Fridman (2:01:13.540)

And what really got them started I think was Chrome.

Lex Fridman (2:01:18.440)

And I talked to Larry Page in 2005,

Brendan Eich (2:01:21.500)

I think I said, we're talking about the Firefox relationship

Lex Fridman (2:01:24.500)

but he was also saying, what about WebKit?

Brendan Eich (2:01:26.120)

This was Apple's version of the old KHTML engine from Linux,

Lex Fridman (2:01:32.020)

the KDE side of Linux that was used in the Conqueror browser

Brendan Eich (2:01:35.780)

also with Ks that Apple had forked.

Lex Fridman (2:01:38.140)

And in 2005 was when Apple's principals

Brendan Eich (2:01:41.700)

including Dave Hyatt, Maciej Stokowiak,

Lex Fridman (2:01:43.400)

some of my friends who are still there said,

Brendan Eich (2:01:45.580)

we must stop patch bombing this poor KHTML project.

Lex Fridman (2:01:48.540)

We should make a proper Mozilla like organization,

Brendan Eich (2:01:50.900)

webkit.org.

Lex Fridman (2:01:51.780)

Now it wasn't a separate nonprofit or anything.

Brendan Eich (2:01:53.700)

It was still Apple, it was Apple controlled

Lex Fridman (2:01:55.220)

but they made their fork first class

Lex Fridman (2:01:58.180)

and they made it be something that they all worked in

Lex Fridman (2:02:00.500)

and lived in.

Lex Fridman (2:02:01.340)

And that was before Chrome.

Lex Fridman (2:02:03.040)

And then Chrome, Larry Page said, what about WebKit?

Brendan Eich (2:02:05.660)

I said, yeah, it's nice.

Lex Fridman (2:02:06.540)

I have friends who work on it.

Brendan Eich (2:02:08.000)

You might use that if you do your own browser.

Lex Fridman (2:02:09.560)

Why don't you do your own browser?

Brendan Eich (2:02:10.460)

Don't worry about Firefox.

Lex Fridman (2:02:11.980)

You should do your own browser.

Brendan Eich (2:02:12.940)

You can have your own opinion of how it should work.

Lex Fridman (2:02:15.260)

And sure enough they did.

Lex Fridman (2:02:17.060)

So by 2006, we knew they'd been working on it.

Lex Fridman (2:02:19.180)

Some of my friends who'd been at Netscape

Brendan Eich (2:02:21.320)

did the original demo.

Lex Fridman (2:02:22.420)

And the demo wasn't what you thought.

Brendan Eich (2:02:24.020)

It didn't have the fast JavaScript yet.

Lex Fridman (2:02:25.340)

That was still off in Denmark on a farm.

Lex Fridman (2:02:29.540)

Did it have tabs?

Lex Fridman (2:02:30.500)

It had tabs because all browsers had tabs at this point.

Lex Fridman (2:02:32.820)

And it had this software fault isolation I mentioned.

Lex Fridman (2:02:37.460)

It was through process isolation.

Lex Fridman (2:02:39.060)

So in theory, each tab has some operating system process.

Lex Fridman (2:02:43.560)

And so what's gonna take your tab down?

Brendan Eich (2:02:45.580)

Well, WebKit has bugs that can crash it

Lex Fridman (2:02:48.380)

but Flash was still big then.

Brendan Eich (2:02:49.880)

All the restaurant sites remember.

Lex Fridman (2:02:51.220)

And Flash crashed a lot.

Lex Fridman (2:02:53.100)

So the demo that I heard about,

Lex Fridman (2:02:55.020)

my friends at Netscape as a lot of people did,

Brendan Eich (2:02:57.980)

inside Google was the sad tab.

Lex Fridman (2:03:00.460)

They showed an early version of Chrome

Brendan Eich (2:03:02.120)

which is just this bare bones tab browser.

Lex Fridman (2:03:04.200)

They loaded a site with a known Flash volume

Lex Fridman (2:03:06.700)

and then suddenly Flash crashes.

Lex Fridman (2:03:08.540)

And everyone expected the whole browser to go down.

Lex Fridman (2:03:10.860)

But instead you got this little sad face in the tab

Lex Fridman (2:03:13.080)

and you could reload it and there it is again.

Lex Fridman (2:03:14.740)

So this was an improvement.

Lex Fridman (2:03:16.740)

It was a real move for security.

Brendan Eich (2:03:18.740)

It was based on a company they acquired

Lex Fridman (2:03:21.420)

called Green Border.

Brendan Eich (2:03:22.660)

They had some really big brains like Olfar Erlingsson

Lex Fridman (2:03:25.060)

I think was involved.

Lex Fridman (2:03:25.880)

And they had done some exotic security stuff

Lex Fridman (2:03:28.880)

but they ended up simplifying it to this process isolation.

Lex Fridman (2:03:32.180)

And it was good.

Lex Fridman (2:03:34.960)

And Firefox didn't have it at the time.

Lex Fridman (2:03:36.740)

So we were still struggling with security bugs.

Lex Fridman (2:03:41.760)

So we knew Chrome was coming

Lex Fridman (2:03:42.980)

but it took two more years to come out.

Lex Fridman (2:03:45.060)

And we were still getting the Google search revenue

Lex Fridman (2:03:48.300)

and we were still making Google the default engine

Lex Fridman (2:03:51.860)

and Firefox was still growing.

Brendan Eich (2:03:52.980)

Firefox grew I think until 2011.

Lex Fridman (2:03:55.260)

That was when it peaked.

Lex Fridman (2:03:56.900)

And as it started falling, it was because of Chrome.

Lex Fridman (2:03:59.900)

Chrome came out in 2008 and it had a comic book

Brendan Eich (2:04:02.900)

that leaked accidentally that showed some of the people

Lex Fridman (2:04:04.820)

who worked on it.

Brendan Eich (2:04:05.660)

Lars Bock was in there and so on.

Lex Fridman (2:04:07.260)

It was kind of soft launch

Brendan Eich (2:04:08.980)

because they didn't market it heavily.

Lex Fridman (2:04:10.380)

They didn't push distribution.

Lex Fridman (2:04:12.580)

But Google had reason to worry about distribution

Lex Fridman (2:04:14.820)

because Microsoft was doing a search engine, Bing,

Brendan Eich (2:04:19.220)

since 2007.

Lex Fridman (2:04:20.500)

In fact, when they came out with Bing,

Brendan Eich (2:04:22.020)

Google was worried that Microsoft would just brute force

Lex Fridman (2:04:25.220)

switch the default browser in everyone's Internet Explorer

Brendan Eich (2:04:28.540)

or even Firefox on Windows to Bing from Google.

Lex Fridman (2:04:32.580)

And Microsoft wasn't I think ready

Brendan Eich (2:04:34.860)

to dare the antitrust cops that way

Lex Fridman (2:04:37.100)

even though they'd gone to sleep.

Lex Fridman (2:04:39.020)

And I don't think Bing was ready either.

Lex Fridman (2:04:42.020)

But just in case it happened, Sundar Pichai,

Brendan Eich (2:04:45.500)

who rose very well based on this work,

Lex Fridman (2:04:47.740)

was sort of in charge of getting distribution deals.

Lex Fridman (2:04:50.580)

And he got Google toolbar

Lex Fridman (2:04:52.180)

and Google desktop search distribution.

Lex Fridman (2:04:55.500)

And if you remember those pieces of software,

Lex Fridman (2:04:56.780)

those were like desktop extensions,

Brendan Eich (2:04:59.460)

toolbars or operating system extensions

Lex Fridman (2:05:01.940)

for doing desktop search, searching your local files.

Lex Fridman (2:05:03.820)

Kind of like Mac OS Spotlight, right?

Lex Fridman (2:05:05.900)

Sadly, it died.

Brendan Eich (2:05:07.300)

It all died.

Lex Fridman (2:05:08.140)

And there were some features that we still missed

Brendan Eich (2:05:09.220)

that didn't make it into Chrome.

Lex Fridman (2:05:10.580)

But Sundar got OEMs to bundle those.

Lex Fridman (2:05:14.020)

And then he got enough of those deals

Lex Fridman (2:05:15.500)

that by 2007 or eight, Google felt,

Brendan Eich (2:05:18.000)

well, if Bing, Microsoft does the worst

Lex Fridman (2:05:20.500)

and tries to force Bing,

Brendan Eich (2:05:21.860)

we can reach in and reset it with that point of presence.

Lex Fridman (2:05:25.500)

So that was good for Sundar's career

Lex Fridman (2:05:27.340)

and it was good for Google,

Lex Fridman (2:05:28.180)

but it never came to pass that they had to defend.

Brendan Eich (2:05:30.780)

Microsoft was still slow.

Lex Fridman (2:05:32.940)

And by the time they saw Chrome come out,

Brendan Eich (2:05:35.000)

then they did what would have been IE9.

Lex Fridman (2:05:37.420)

And then they said,

Brendan Eich (2:05:38.260)

we're gonna have a fast JavaScript engine

Lex Fridman (2:05:39.500)

to Chakra, Chakra core.

Lex Fridman (2:05:41.380)

And they did okay.

Lex Fridman (2:05:43.140)

They were another process isolated,

Brendan Eich (2:05:46.220)

fast JavaScript browser, tab browser.

Lex Fridman (2:05:48.980)

So it sounds like there's a deep fundamental coupling

Brendan Eich (2:05:52.340)

of search engine and browser

Lex Fridman (2:05:53.640)

that's mixing this whole thing up.

Lex Fridman (2:05:55.780)

And obviously Firefox doesn't have a search engine.

Lex Fridman (2:05:59.980)

That's like, I mean, you're partnering with somebody

Brendan Eich (2:06:03.680)

with a search engine.

Lex Fridman (2:06:04.940)

With Yahoo or with Google or so on.

Brendan Eich (2:06:08.780)

They tried Yahoo, that was unfortunate

Lex Fridman (2:06:10.660)

because I think even though Marissa Mayer talked about it,

Brendan Eich (2:06:14.660)

she never pulled it off.

Lex Fridman (2:06:15.840)

They never restored the search team

Brendan Eich (2:06:18.620)

that had been laid off.

Lex Fridman (2:06:19.600)

I believe Carol Bartz was running Yahoo

Brendan Eich (2:06:21.460)

when Carol said, I've got to get rid

Lex Fridman (2:06:23.100)

of one of the three expensive things.

Brendan Eich (2:06:24.580)

I'm gonna get rid of search.

Lex Fridman (2:06:25.940)

And those researchers went to Google and Microsoft

Lex Fridman (2:06:29.800)

and there was no way to put Yahoo search back together.

Lex Fridman (2:06:32.500)

So when Firefox tried switching all their users

Brendan Eich (2:06:36.020)

who'd stuck with a default from Google to Yahoo,

Lex Fridman (2:06:38.700)

it was like mid December, 2014,

Lex Fridman (2:06:40.980)

a bunch of users said, what just happened to my Firefox?

Lex Fridman (2:06:44.060)

And others didn't notice right away,

Lex Fridman (2:06:45.300)

but over time they did.

Lex Fridman (2:06:46.220)

And so over the next year,

Brendan Eich (2:06:48.320)

the traffic just went away for Yahoo.

Lex Fridman (2:06:51.360)

And yet they were obliged, I understand it.

Brendan Eich (2:06:54.420)

I don't have inside knowledge, but this has leaked out

Lex Fridman (2:06:56.700)

and Danny Sullivan's written about it,

Brendan Eich (2:06:58.340)

search engine land.

Lex Fridman (2:06:59.260)

I think the deal was like fixed payments to Mozilla.

Lex Fridman (2:07:01.520)

So Mozilla was getting a bunch of money for traffic

Lex Fridman (2:07:03.500)

that wasn't staying because users

Brendan Eich (2:07:04.960)

were resetting their default.

Lex Fridman (2:07:06.580)

And this shows how defaults are important,

Lex Fridman (2:07:09.220)

but they have to be good enough

Lex Fridman (2:07:10.520)

that the user doesn't override them.

Lex Fridman (2:07:12.580)

And a lot of the commercial value in popular apps

Lex Fridman (2:07:16.620)

is what are the default settings?

Lex Fridman (2:07:18.040)

What is the default search?

Lex Fridman (2:07:20.120)

But oftentimes there's something just like you said,

Brendan Eich (2:07:21.860)

I mean, if there's something compelling

Lex Fridman (2:07:25.020)

that's also can beat out the default,

Brendan Eich (2:07:27.220)

like tab browsing and so on.

Lex Fridman (2:07:29.020)

And that's where, I mean, we'll talk about brave browser.

Brendan Eich (2:07:32.300)

It feels like now we're in this third stage

Lex Fridman (2:07:36.200)

where there's a Chrome, Firefox, Edge,

Brendan Eich (2:07:41.440)

I guess it's called and brave.

Lex Fridman (2:07:44.620)

And these are all seem like really exciting,

Brendan Eich (2:07:48.500)

I don't know, innovative browsers.

Lex Fridman (2:07:50.560)

They're all kind of copying off of each other,

Brendan Eich (2:07:52.260)

picking up the good stuff.

Lex Fridman (2:07:53.760)

There's evolution again, especially on tracking protection.

Lex Fridman (2:07:56.260)

So privacy is this sort of global wave that's rising.

Lex Fridman (2:08:02.500)

I like to call it a wave because it's a large,

Brendan Eich (2:08:04.820)

somewhat chaotic structure.

Lex Fridman (2:08:07.020)

It's not a unitary good.

Brendan Eich (2:08:08.600)

You can't say I'm buying privacy for $3,

Lex Fridman (2:08:11.780)

I'm paying $3 privacy.

Brendan Eich (2:08:12.740)

Some people think a VPN does this

Lex Fridman (2:08:14.300)

and are disappointed when it fails them.

Lex Fridman (2:08:15.980)

But often people use VPNs for region unlocking video

Lex Fridman (2:08:19.700)

or getting the US Netflix catalog.

Lex Fridman (2:08:22.860)

But privacy is not a unitary good, it's complex

Lex Fridman (2:08:25.380)

and people are understanding it only over time

Lex Fridman (2:08:28.060)

and as they get burned, but there's a genie

Lex Fridman (2:08:30.700)

that's not going back in the bottle there.

Brendan Eich (2:08:31.980)

People are fed up.

Lex Fridman (2:08:33.380)

Apple has responded to this.

Brendan Eich (2:08:34.700)

Apple was always making Safari, I think,

Lex Fridman (2:08:36.620)

more of a privacy branded browser from the very beginning.

Brendan Eich (2:08:40.340)

I think this was probably Steve Jobs.

Lex Fridman (2:08:42.300)

Safari had private windows, private tabs

Brendan Eich (2:08:45.260)

before Firefox did.

Lex Fridman (2:08:48.220)

And these are only private in the sense

Brendan Eich (2:08:50.060)

that they don't leave local traces,

Lex Fridman (2:08:53.020)

if you don't want them to.

Brendan Eich (2:08:53.980)

Turns out Safari does keep them around between shutdown.

Lex Fridman (2:08:57.180)

But the canonical model is no local traces

Brendan Eich (2:08:59.860)

after you close the private window.

Lex Fridman (2:09:01.780)

No leftover traces that you went to some site

Brendan Eich (2:09:04.900)

that you were embarrassed by

Lex Fridman (2:09:06.140)

or bought a gift for somebody you wanted to keep secret.

Lex Fridman (2:09:10.220)

But there's still some level of tracking.

Lex Fridman (2:09:11.780)

There's network tracking.

Brendan Eich (2:09:14.860)

Network privacy is not guaranteed at all

Lex Fridman (2:09:16.900)

because you're using the same internet and ISP

Brendan Eich (2:09:18.980)

as a public window, a non private window.

Lex Fridman (2:09:21.340)

But, Safari had that early on.

Brendan Eich (2:09:22.900)

They also had a cookie blocking policy

Lex Fridman (2:09:26.060)

that might take a little explaining.

Brendan Eich (2:09:27.860)

When you, if you know what a cookie is,

Lex Fridman (2:09:29.820)

it's a little bit of storage in the browser

Brendan Eich (2:09:31.300)

indexed by the name of the site.

Lex Fridman (2:09:33.420)

And it's really only the main name of the site,

Brendan Eich (2:09:35.060)

like bofa.com or, you know, something like npr.org.

Lex Fridman (2:09:43.700)

Every site can store some information in a cookie.

Brendan Eich (2:09:45.980)

Every time it's contacted by the browser,

Lex Fridman (2:09:48.620)

the previous version is sent back.

Lex Fridman (2:09:50.860)

And in the response from the server, the cookie's updated.

Lex Fridman (2:09:54.060)

So it's this little bit of storage in the browser

Brendan Eich (2:09:56.580)

that the site can keep updating

Lex Fridman (2:09:58.180)

and it can store an encrypted version

Brendan Eich (2:10:01.460)

of your login credentials with a timestamp

Lex Fridman (2:10:03.820)

so you can stay logged in

Brendan Eich (2:10:05.980)

without having to retype your password

Lex Fridman (2:10:07.700)

every time you navigate,

Brendan Eich (2:10:08.820)

which is how it would be if you didn't have cookies.

Lex Fridman (2:10:11.220)

The web protocols, especially in the 90s,

Brendan Eich (2:10:14.420)

are so called stateless protocols.

Lex Fridman (2:10:16.020)

So you go to your bank, you log in,

Brendan Eich (2:10:18.740)

you go from your login confirmed page

Lex Fridman (2:10:21.220)

to your account view.

Brendan Eich (2:10:22.460)

If you didn't have a cookie, you'd be logging in again.

Lex Fridman (2:10:24.980)

Every time you type in the source.

Lex Fridman (2:10:26.580)

So that was the great thing about cookies.

Lex Fridman (2:10:27.900)

Luhmann truly did it in a hurry in 1994

Brendan Eich (2:10:29.820)

before I joined Escape

Lex Fridman (2:10:30.860)

and he did it for really holding that kind of credential.

Lex Fridman (2:10:34.860)

But even then there was the image element

Lex Fridman (2:10:37.500)

embedded in the page

Lex Fridman (2:10:38.580)

and the image gets fetched possibly from a different server

Lex Fridman (2:10:41.980)

and that request carries the last cookie,

Brendan Eich (2:10:44.220)

which could be empty at first,

Lex Fridman (2:10:45.740)

and the response carries the updated cookie.

Lex Fridman (2:10:47.980)

So just by having images and cookies,

Lex Fridman (2:10:49.780)

you got tracking because that image server

Brendan Eich (2:10:51.420)

can be serving a little one by one pixel

Lex Fridman (2:10:54.220)

and they still use the word pixel in ad tech.

Lex Fridman (2:10:56.500)

And that pixel can be served from the same server,

Lex Fridman (2:10:59.380)

embedded differently with different URL spellings

Brendan Eich (2:11:02.140)

in the New York Times and ESPN.

Lex Fridman (2:11:03.740)

And as you go from one to the other,

Brendan Eich (2:11:05.860)

the image server can say,

Lex Fridman (2:11:06.980)

I haven't got a cookie for you.

Brendan Eich (2:11:08.180)

It's empty initially.

Lex Fridman (2:11:09.020)

I'm gonna assign you user number 1234.

Brendan Eich (2:11:11.260)

I'm gonna put a database entry in.

Lex Fridman (2:11:13.020)

And I see, by the way,

Brendan Eich (2:11:13.860)

I always fetch the name of the path part of the URL

Lex Fridman (2:11:15.980)

that I was in the New York Times.

Lex Fridman (2:11:17.460)

So you're a New York Times reader.

Lex Fridman (2:11:18.980)

And then you hit ESPN, same thing.

Lex Fridman (2:11:20.980)

And the database gets updated

Lex Fridman (2:11:22.260)

and the number user 1234 indexes in the database

Brendan Eich (2:11:25.580)

to a profile of you, you've been tracked.

Lex Fridman (2:11:28.140)

This was not intended.

Lex Fridman (2:11:29.420)

And it was too late to undo by the time I got the Netscape.

Lex Fridman (2:11:32.740)

I think Lou wanted to do Twinkies, he called them.

Lex Fridman (2:11:35.100)

And he was trying to solve several problems.

Lex Fridman (2:11:37.380)

He wanted them to be bigger

Brendan Eich (2:11:38.380)

because initially cookies had a short size limit.

Lex Fridman (2:11:40.540)

I think he wanted to solve the third party problem,

Lex Fridman (2:11:42.340)

but Tom Paquin, the engineering manager said,

Lex Fridman (2:11:45.100)

nope, no Twinkies, just cookies.

Brendan Eich (2:11:47.180)

We're done.

Lex Fridman (2:11:48.100)

You're done, son.

Lex Fridman (2:11:49.060)

And that's how a lot of that stuff was.

Lex Fridman (2:11:51.540)

That's how JavaScript got frozen

Brendan Eich (2:11:54.620)

like a flying Amber in some ways

Lex Fridman (2:11:55.860)

with that sloppy equality operator that I made

Brendan Eich (2:11:57.900)

because of the early adopters.

Lex Fridman (2:11:58.980)

And the cookie got stuck with this tracking hazard.

Lex Fridman (2:12:01.900)

And then because JavaScripts can be like images,

Lex Fridman (2:12:04.300)

they're embedded in the page.

Brendan Eich (2:12:05.940)

By the time Netscape 3, I made that work.

Lex Fridman (2:12:08.660)

You can get a request with the last cookie value

Lex Fridman (2:12:10.820)

and the response updates it.

Lex Fridman (2:12:11.940)

That's a tracking mechanism.

Lex Fridman (2:12:12.940)

And that's why you don't even need images to track.

Lex Fridman (2:12:14.940)

Now you just use scripts.

Lex Fridman (2:12:16.460)

So this whole tracking economy evolved

Lex Fridman (2:12:20.020)

and it depended on these accidents of the 90s,

Brendan Eich (2:12:23.860)

these unintended consequences.

Lex Fridman (2:12:25.300)

Well, it created some of the richest companies

Lex Fridman (2:12:27.200)

in the world, right?

Lex Fridman (2:12:28.040)

I mean, it's the social media.

Brendan Eich (2:12:28.860)

All I got was T shirts.

Lex Fridman (2:12:29.980)

All I got is this crappy T shirt.

Brendan Eich (2:12:33.180)

Yeah.

Lex Fridman (2:12:34.100)

I mean, so that's the fundamental problem

Brendan Eich (2:12:38.420)

the world is facing now.

Lex Fridman (2:12:39.500)

They're looking at what social media has created

Lex Fridman (2:12:42.020)

and they're looking at,

Lex Fridman (2:12:44.180)

and like a world is looking at itself in the mirror

Lex Fridman (2:12:47.020)

and seeing that privacy is actually something

Lex Fridman (2:12:50.980)

as opposed to like a nice thing to have.

Brendan Eich (2:12:53.540)

It's something that is actually should be fundamental

Lex Fridman (2:12:57.260)

to the way we interact with the world

Brendan Eich (2:12:59.220)

as part of our tooling.

Lex Fridman (2:13:00.600)

And that's where the Brave browser comes in.

Lex Fridman (2:13:03.540)

And I suppose others as well are playing with this idea,

Lex Fridman (2:13:06.060)

but Brave is at the forefront of that.

Lex Fridman (2:13:07.660)

So maybe can you like describe what Brave is

Lex Fridman (2:13:11.020)

and what are its key principles and what's broken

Lex Fridman (2:13:14.700)

and what is it Brave trying to fix?

Lex Fridman (2:13:16.140)

So when I realized that these accidents

Brendan Eich (2:13:18.380)

like the third party cookie,

Lex Fridman (2:13:19.900)

the image or script that's tracking you

Brendan Eich (2:13:21.840)

or the JavaScripts that can do an invisibly now,

Lex Fridman (2:13:26.020)

that all this stuff wasn't intended

Lex Fridman (2:13:28.700)

and that Firefox had supported extensions

Lex Fridman (2:13:31.660)

that block some of these things,

Brendan Eich (2:13:32.640)

I thought probably we should have browsers

Lex Fridman (2:13:34.740)

just block some of these things by default.

Brendan Eich (2:13:37.440)

These were not intended and they're now unsafe.

Lex Fridman (2:13:39.780)

They're tracking you.

Brendan Eich (2:13:40.620)

There could be data breaches, malware distribution,

Lex Fridman (2:13:44.940)

bullying and psyops and other attacks on people.

Brendan Eich (2:13:49.660)

Block that stuff, block that JavaScript.

Lex Fridman (2:13:51.500)

I'm Dr. Frankenstein, I've got to deal with a monster here.

Lex Fridman (2:13:54.700)

But obviously you go to Gmail,

Lex Fridman (2:13:56.540)

there's a bunch of script there

Brendan Eich (2:13:57.540)

to make that amazing web client.

Lex Fridman (2:13:59.320)

That's okay, that's first party JavaScript.

Lex Fridman (2:14:02.180)

So how do you tell the first from the third party?

Lex Fridman (2:14:04.020)

And it's not easy.

Brendan Eich (2:14:04.860)

It's not a matter of just what's embedded

Lex Fridman (2:14:06.700)

from a different server because a lot of publishers

Brendan Eich (2:14:09.860)

use benign scripts from unrelated domains

Lex Fridman (2:14:12.980)

or apparently unrelated domains.

Lex Fridman (2:14:14.360)

So you end up having to develop a sort of human

Lex Fridman (2:14:16.660)

and machine learning practice around blocking.

Lex Fridman (2:14:19.700)

And at Brave, we did that from the start

Lex Fridman (2:14:22.300)

and built a research team to help drive it and automate it.

Brendan Eich (2:14:25.740)

We realized that protecting people needed machine learning

Lex Fridman (2:14:28.900)

and around 2017 spring,

Brendan Eich (2:14:30.420)

I talked to my friends at Apple about this too

Lex Fridman (2:14:32.680)

and they were also doing

Lex Fridman (2:14:34.060)

what they call intelligent tracking prevention,

Lex Fridman (2:14:36.640)

which uses local machine learning in the browser.

Lex Fridman (2:14:39.320)

And the funny thing is, great minds think alike,

Lex Fridman (2:14:42.880)

they were taking their third party cookie blocker

Brendan Eich (2:14:44.780)

that was in Safari from the old days

Lex Fridman (2:14:46.460)

and making it not have a big loophole.

Brendan Eich (2:14:48.540)

Because what they did was in 2003, when Safari came out,

Lex Fridman (2:14:51.740)

they said, we're gonna block cookies

Brendan Eich (2:14:53.740)

that are from those third party embedded elements

Lex Fridman (2:14:57.220)

where you've never visited that site before.

Lex Fridman (2:14:59.120)

So I'm gonna pick an ad company that got sold to AT&T,

Lex Fridman (2:15:02.340)

so I'm not picking on anybody unfairly, appnexus.com.

Lex Fridman (2:15:05.140)

Have you ever been to appnexus.com?

Lex Fridman (2:15:06.660)

Nope.

Brendan Eich (2:15:07.500)

I've never been there, but I guarantee you 10 years ago,

Lex Fridman (2:15:09.020)

you probably had, if you were using Firefox,

Brendan Eich (2:15:11.180)

you had a cookie, third party cookie,

Lex Fridman (2:15:13.020)

because you were being tracked by them

Lex Fridman (2:15:14.980)

and they were using that cookie

Lex Fridman (2:15:15.920)

to build up a profile of you.

Brendan Eich (2:15:17.560)

In Safari, as long as the user never went to appnexus,

Lex Fridman (2:15:20.700)

that cookie would not be set.

Lex Fridman (2:15:22.380)

And that was a real move for privacy early on

Lex Fridman (2:15:25.460)

when jobs were still around in Safari.

Lex Fridman (2:15:27.660)

But it had this loophole that if you do go to appnexus,

Lex Fridman (2:15:30.620)

then why it's okay to be a third party cookie.

Lex Fridman (2:15:32.540)

And so appnexus did something very naughty.

Lex Fridman (2:15:35.180)

They took their ad partners

Brendan Eich (2:15:36.820)

to put the actual ad you click on.

Lex Fridman (2:15:39.000)

And they said, hey, add a little script

Lex Fridman (2:15:40.300)

so that when somebody clicks on the ad,

Lex Fridman (2:15:42.620)

before it goes to your landing page,

Brendan Eich (2:15:44.500)

redirect to appnexus and we'll redirect

Lex Fridman (2:15:46.340)

to the landing page.

Lex Fridman (2:15:47.180)

And by doing that, they set a first party cookie

Lex Fridman (2:15:49.140)

and they got whitelisted.

Lex Fridman (2:15:49.980)

So it was a loophole they exploited.

Lex Fridman (2:15:51.660)

Intelligent tracking prevention in Safari

Brendan Eich (2:15:53.580)

was sophisticated enough to counteract this

Lex Fridman (2:15:56.620)

and it did other things and it's evolved since they did it.

Lex Fridman (2:15:59.100)

And we've evolved brave too.

Lex Fridman (2:16:00.720)

And so when I say machine and human learning,

Brendan Eich (2:16:02.560)

there's a real set of techniques here.

Lex Fridman (2:16:05.580)

They have to fight.

Brendan Eich (2:16:06.420)

This is a fascinating problem.

Lex Fridman (2:16:07.580)

Fingerprinting, right?

Brendan Eich (2:16:08.400)

Anytime you have a little bit of storage in the browser

Lex Fridman (2:16:10.720)

associated with a website,

Brendan Eich (2:16:13.580)

if the bad guy can get 32 websites,

Lex Fridman (2:16:15.540)

each one has a bit of storage, that's 32 bits.

Brendan Eich (2:16:17.660)

You can turn the bit on or off,

Lex Fridman (2:16:19.820)

you can make 4 billion numbers,

Brendan Eich (2:16:21.280)

you can make an identifier.

Lex Fridman (2:16:22.260)

It's called a super cookie sometimes.

Brendan Eich (2:16:26.660)

There are weaker ways that are statistical.

Lex Fridman (2:16:28.880)

They're called fingerprinting.

Brendan Eich (2:16:29.780)

You have to block all of them

Lex Fridman (2:16:30.780)

and you have to not only automate,

Brendan Eich (2:16:32.840)

you want to work in the web standards body

Lex Fridman (2:16:34.940)

to put privacy in by default, by design,

Brendan Eich (2:16:38.480)

from the get go, not add it as an afterthought

Lex Fridman (2:16:40.500)

or go hogwile with new web APIs

Brendan Eich (2:16:43.180)

to add a bunch more local storage or fingerprint surface area.

Lex Fridman (2:16:47.500)

And that's been a struggle too,

Brendan Eich (2:16:48.660)

because guess who's the new Microsoft

Lex Fridman (2:16:50.940)

in the standards body?

Brendan Eich (2:16:51.780)

It's Google.

Lex Fridman (2:16:52.700)

And they're not in favor of privacy first.

Brendan Eich (2:16:55.900)

They want to do privacy their way,

Lex Fridman (2:16:58.740)

only under, I would say, market pressure.

Lex Fridman (2:17:01.180)

But with Apple and with Brave leading the way,

Lex Fridman (2:17:03.660)

we block third party cookies almost without exception.

Lex Fridman (2:17:06.280)

So we've just blocked them.

Lex Fridman (2:17:07.940)

And that gives us a very strong privacy benefit,

Lex Fridman (2:17:11.540)

but it also means some sites just don't work right.

Lex Fridman (2:17:13.740)

Embedded YouTube videos might not work right.

Lex Fridman (2:17:15.500)

So we're adapting in a similar way to Apple's done with ITP

Lex Fridman (2:17:19.620)

to make third party cookies blocked,

Lex Fridman (2:17:22.740)

but to sort of simulate what looks like

Lex Fridman (2:17:26.460)

a working third party cookie for the site.

Brendan Eich (2:17:28.860)

It essentially tries to partition each site

Lex Fridman (2:17:32.320)

and its third parties into its own sort of cookie jar.

Brendan Eich (2:17:34.860)

Got it.

Lex Fridman (2:17:35.700)

And so, like you said,

Brendan Eich (2:17:37.060)

is this both like a human fine tuning issue

Lex Fridman (2:17:41.100)

and a machine learning problem?

Lex Fridman (2:17:42.580)

And as the humans learn,

Lex Fridman (2:17:44.540)

then they train the machine learning.

Brendan Eich (2:17:46.380)

But, you know, maybe Google aside or including Google,

Lex Fridman (2:17:50.580)

there's millions of dollars, if not B,

Brendan Eich (2:17:53.220)

billions of dollars to be made

Lex Fridman (2:17:54.900)

from fighting the ways of Brave.

Brendan Eich (2:17:57.860)

That's right.

Lex Fridman (2:17:58.700)

And it's been an interesting change

Brendan Eich (2:18:00.660)

from when we started in 2015.

Lex Fridman (2:18:01.980)

When we started, you know, ad blocking extensions,

Brendan Eich (2:18:04.420)

ad block plus was one of the big ones

Lex Fridman (2:18:06.140)

that started on Firefox in 2006, I believe,

Brendan Eich (2:18:08.660)

had gotten to a certain level of use around the world.

Lex Fridman (2:18:11.020)

And browsers like UC Web, UC Browser in Asia

Brendan Eich (2:18:14.380)

had some amount of ad blocking built in and on by default.

Lex Fridman (2:18:17.300)

So, a page fair was a startup

Lex Fridman (2:18:20.500)

and they measured ad blocking adoption.

Lex Fridman (2:18:23.580)

And they tried to say,

Brendan Eich (2:18:24.420)

hey publishers, you're, you know,

Lex Fridman (2:18:26.580)

30% of the visitors to Pitchfork or Wire

Brendan Eich (2:18:30.200)

to Conda NAS properties are using ad blockers.

Lex Fridman (2:18:32.900)

If we can somehow convince them

Brendan Eich (2:18:34.380)

to lower their ad blocking for your site,

Lex Fridman (2:18:37.340)

that could be like a 43% lift, right?

Brendan Eich (2:18:39.300)

And, you know, three sevenths.

Lex Fridman (2:18:42.020)

Well, that's easier said than done.

Lex Fridman (2:18:44.260)

And PageFair and others, SourcePoint,

Lex Fridman (2:18:46.140)

and many others tried to either smuggle ads through

Brendan Eich (2:18:48.060)

or cajole the user into letting, you know, ads appear.

Lex Fridman (2:18:52.260)

And it didn't really work.

Lex Fridman (2:18:53.100)

And meanwhile, the ad blocking adoption

Lex Fridman (2:18:54.460)

has just continued intelligent tracking prevention

Brendan Eich (2:18:56.860)

in Safari in 2017, Brave from 2016 on

Lex Fridman (2:19:01.620)

with very strong cookie blocking and other protections.

Lex Fridman (2:19:04.460)

And this is not going away.

Lex Fridman (2:19:06.660)

The publishers used to rage against it.

Brendan Eich (2:19:08.500)

Like we would try to say, we can help you.

Lex Fridman (2:19:10.460)

You're dealing with users

Brendan Eich (2:19:11.960)

who are already blocking all your ads.

Lex Fridman (2:19:14.100)

We can try to put back some economics

Brendan Eich (2:19:16.800)

that help the user and you

Lex Fridman (2:19:18.460)

that led to the basic attention token

Brendan Eich (2:19:19.980)

that we started with Bitcoin.

Lex Fridman (2:19:21.460)

We can be your friend.

Brendan Eich (2:19:22.460)

Don't just fingerprint us as an ad blocker

Lex Fridman (2:19:24.980)

and treat us as an enemy.

Lex Fridman (2:19:26.480)

But in 2015 or 16, it was like,

Lex Fridman (2:19:28.620)

nah, you're an ad blocker.

Brendan Eich (2:19:29.540)

Get out of here.

Lex Fridman (2:19:30.380)

I hate you.

Lex Fridman (2:19:31.200)

And by 2017 or 18, it's like something's happening.

Lex Fridman (2:19:33.740)

The ad blocking is not stopping

Lex Fridman (2:19:34.940)

and we're all getting sort of pulled

Lex Fridman (2:19:37.020)

on the Google's plantation through AMP, AMP.

Brendan Eich (2:19:41.580)

Or we're getting killed by the Google ad system we use

Lex Fridman (2:19:44.580)

because it's taking all the revenue

Brendan Eich (2:19:46.620)

or it's permitting or some other vendors we use

Lex Fridman (2:19:49.140)

are permitting ad fraud.

Lex Fridman (2:19:50.140)

And so a fake New York Times is getting paid

Lex Fridman (2:19:52.640)

by the marketer running an ad that a bot clicks on.

Lex Fridman (2:19:57.140)

And the real New York Times

Lex Fridman (2:19:58.240)

that's supposed to get the ad doesn't get it.

Lex Fridman (2:20:00.500)

And there's something really broken

Lex Fridman (2:20:01.540)

about that kind of system.

Lex Fridman (2:20:02.460)

And that fraud is mediated through Google's ad exchange,

Lex Fridman (2:20:06.700)

which is the biggest of them all.

Lex Fridman (2:20:07.860)

And Google takes a fee.

Lex Fridman (2:20:09.660)

There's a flip side of that,

Brendan Eich (2:20:10.700)

which is malware distribution, malvertising,

Lex Fridman (2:20:12.660)

where fake advertisers put malware payloads

Brendan Eich (2:20:16.300)

in or exploit hit loaders in JavaScript

Lex Fridman (2:20:19.340)

and they smuggle them in ads onto real publisher pages,

Brendan Eich (2:20:22.460)

the ad exchange takes the fee.

Lex Fridman (2:20:24.460)

Now, I'm not a lawyer.

Brendan Eich (2:20:25.580)

I'm not gonna say this is a RICO predicate,

Lex Fridman (2:20:27.100)

but why is the ad exchange facilitating fraud

Lex Fridman (2:20:30.220)

and malware distribution and taking a fee?

Lex Fridman (2:20:32.060)

It's not right.

Brendan Eich (2:20:32.880)

As opposed to just fighting,

Lex Fridman (2:20:33.940)

this is the really interesting thing about Brave

Brendan Eich (2:20:35.660)

is as opposed to just fighting

Lex Fridman (2:20:37.700)

and then being treated like an ad blocker,

Brendan Eich (2:20:39.980)

you're providing an alternate.

Lex Fridman (2:20:42.060)

There's a philosophical idea here

Brendan Eich (2:20:45.460)

that might change the nature of the internet

Lex Fridman (2:20:47.500)

with the basic attention token.

Brendan Eich (2:20:49.220)

Yes.

Lex Fridman (2:20:50.060)

Well, maybe what is basic attention token BAT

Lex Fridman (2:20:54.060)

and how does it work?

Lex Fridman (2:20:55.300)

Okay, I'll tell the story first by saying how I came to it.

Brendan Eich (2:20:58.200)

I realized for a long time at Firefox,

Lex Fridman (2:21:00.820)

we were dependent on this Google search deal.

Lex Fridman (2:21:02.940)

And I thought, now that Chrome's out,

Lex Fridman (2:21:05.700)

maybe that's gonna go away.

Lex Fridman (2:21:07.260)

And they just, at some point, Google will say,

Lex Fridman (2:21:10.740)

Firefox, like old yeller, you saved me from the rabid beast,

Brendan Eich (2:21:14.620)

now I have to shoot you in the head.

Lex Fridman (2:21:16.360)

Done your job, sad but true, goodbye.

Lex Fridman (2:21:19.660)

And what could we do?

Lex Fridman (2:21:21.780)

And I think Mozilla doesn't know what to do.

Brendan Eich (2:21:24.140)

This is something that I couldn't solve there

Lex Fridman (2:21:25.900)

and I don't think they can solve.

Lex Fridman (2:21:27.220)

But I thought, why is the browser

Lex Fridman (2:21:29.460)

the sort of passive servant of these big tech companies?

Lex Fridman (2:21:32.300)

Why is it a blind runtime for ad tech JavaScripts,

Lex Fridman (2:21:36.420)

including from Google?

Lex Fridman (2:21:38.100)

Why doesn't it block some?

Lex Fridman (2:21:39.220)

And if it blocks some, why can't it reconnect users,

Lex Fridman (2:21:43.940)

readers, fans with publishers, creators, websites?

Lex Fridman (2:21:48.480)

Why can't it help people make direct payments

Brendan Eich (2:21:51.220)

or even possibly get an ad revenue share for private ads

Lex Fridman (2:21:54.900)

that are placed in the browser?

Brendan Eich (2:21:56.300)

The ads are all placed in the browser.

Lex Fridman (2:21:57.580)

Some people have this sort of model

Brendan Eich (2:21:58.840)

that the server's painting the ad into some,

Lex Fridman (2:22:01.620)

flash combined package or into some giant image

Lex Fridman (2:22:06.500)

and then it all gets sent down, that's not how it works.

Lex Fridman (2:22:08.600)

All the ads you see on the web are placed in your browser

Brendan Eich (2:22:11.340)

by it calling out to various ad tech partners

Lex Fridman (2:22:13.980)

and Google's among them.

Lex Fridman (2:22:15.620)

And so if you block those scripts,

Lex Fridman (2:22:17.580)

you break the advertising flow of money

Brendan Eich (2:22:22.420)

from the brands and their agencies to the publishers.

Lex Fridman (2:22:26.300)

And if you want to reconnect it directly with the user,

Brendan Eich (2:22:29.460)

you have limited choices.

Lex Fridman (2:22:30.420)

The user generally isn't gonna sign up

Brendan Eich (2:22:31.940)

with a ACH bank connection or a credit card.

Lex Fridman (2:22:34.980)

The publisher isn't gonna sign up the user

Brendan Eich (2:22:37.260)

except as a subscriber and then they're gonna overcharge you

Lex Fridman (2:22:39.980)

because they want you to cross subsidize all the content

Lex Fridman (2:22:42.060)

and buy more than you read and all that stuff.

Lex Fridman (2:22:43.660)

And how many, people are doing great who are big names

Brendan Eich (2:22:47.220)

like New York Times and The Washington Post,

Lex Fridman (2:22:48.740)

but how many subscriptions are you as a user gonna pay for?

Brendan Eich (2:22:51.820)

This is why startups like Tony Hale Scroll

Lex Fridman (2:22:54.420)

are trying to do a portable subscription system.

Brendan Eich (2:22:57.160)

By the way, just a small tangent there,

Lex Fridman (2:22:58.860)

even the New York Times is really annoying

Lex Fridman (2:23:00.860)

how difficult it is to subscribe.

Lex Fridman (2:23:03.300)

There's way too many clicks.

Brendan Eich (2:23:04.620)

They don't make it easy.

Lex Fridman (2:23:05.460)

And I had friends a few years ago,

Brendan Eich (2:23:06.300)

I think they fixed this, who would pay for the paper

Lex Fridman (2:23:08.340)

and then they'd go online and they get upcharged

Brendan Eich (2:23:11.140)

for the digital and there was no break.

Lex Fridman (2:23:13.140)

There was no connection between them.

Lex Fridman (2:23:15.380)

But publishers are not that technical

Lex Fridman (2:23:17.260)

and they can't all get you to subscribe.

Brendan Eich (2:23:19.580)

You can't have a thousand subscriptions.

Lex Fridman (2:23:21.300)

So for a long time, people talked about micropayments.

Brendan Eich (2:23:23.460)

There was Blendle and other ones which came to the US,

Lex Fridman (2:23:25.860)

but it didn't grow.

Lex Fridman (2:23:27.580)

And I thought, if you have just a browser

Lex Fridman (2:23:29.820)

and it's protecting you by blocking

Brendan Eich (2:23:31.200)

all this ad tech tracking junk,

Lex Fridman (2:23:33.820)

it can provide you an option that uses cryptocurrency

Brendan Eich (2:23:36.620)

to let you support your favorite sites

Lex Fridman (2:23:39.420)

and even YouTube channels.

Lex Fridman (2:23:40.580)

And that we prototyped with Bitcoin.

Lex Fridman (2:23:42.740)

And that meant the user had to be of means to contribute

Lex Fridman (2:23:45.940)

and willing to contribute,

Lex Fridman (2:23:47.480)

but it could be done on the Bitcoin blockchain

Lex Fridman (2:23:49.220)

and it could be fairly efficient

Lex Fridman (2:23:50.560)

even though Bitcoin went through a period

Brendan Eich (2:23:52.260)

when we had this prototype running in 2016 into 2017

Lex Fridman (2:23:55.740)

where Bitcoin was very congested and very slow to confirm

Lex Fridman (2:23:58.780)

and the fees got very high.

Lex Fridman (2:24:01.020)

And a lot of users who were not Bitcoin maximalists

Brendan Eich (2:24:03.540)

or even experienced,

Lex Fridman (2:24:04.940)

we helped them out by embedding a Coinbase buy widget

Lex Fridman (2:24:07.620)

and they had the income to buy, but it was hard.

Lex Fridman (2:24:10.180)

It was like, do I buy $5 a month?

Lex Fridman (2:24:12.420)

But the fee is like 450.

Lex Fridman (2:24:14.320)

I better buy in larger batches, right?

Lex Fridman (2:24:16.420)

And they're like, I don't wanna own that much Bitcoin.

Lex Fridman (2:24:18.540)

So it became this painful thing.

Lex Fridman (2:24:21.180)

And the real idea that I had of private ads

Lex Fridman (2:24:23.800)

that pay the user a rev share

Brendan Eich (2:24:25.940)

couldn't be realized alone in that kind of system.

Lex Fridman (2:24:30.580)

In these cryptocurrency systems,

Brendan Eich (2:24:33.060)

especially with the blockchain we switched to Ethereum,

Lex Fridman (2:24:35.180)

you can have smart contracts.

Brendan Eich (2:24:36.920)

The Bitcoin system is not turned complete.

Lex Fridman (2:24:38.980)

So what you can do with the scripts is more limited,

Lex Fridman (2:24:41.100)

but you can still do sort of clever things

Lex Fridman (2:24:44.500)

even with Bitcoin script.

Lex Fridman (2:24:46.140)

What we wanted to do was sort of a three sided ecosystem.

Lex Fridman (2:24:49.280)

We wanted users, creators or publishers and advertisers.

Lex Fridman (2:24:53.740)

And we wanted the advertisers to put money in

Lex Fridman (2:24:55.740)

just like they do today,

Lex Fridman (2:24:57.140)

but without going through the Googles and the app nexuses

Lex Fridman (2:25:00.380)

and all these other ad tech companies,

Brendan Eich (2:25:01.980)

because those companies take out a huge cut.

Lex Fridman (2:25:04.420)

The Guardian in the UK once did an experiment for a month.

Brendan Eich (2:25:06.940)

They bought out their own ad space.

Lex Fridman (2:25:08.820)

They put in a pound and they were paid 30 pence.

Brendan Eich (2:25:11.180)

70% was coming out to the intermediary vendors

Lex Fridman (2:25:15.200)

they were using.

Lex Fridman (2:25:17.020)

And that's like the opposite of what the app store does.

Lex Fridman (2:25:19.260)

The app store takes 30% and gives the publisher 70%.

Lex Fridman (2:25:22.220)

So pretty broken, in the old days of the superstation TBS,

Lex Fridman (2:25:26.180)

the media owner would get 85%.

Lex Fridman (2:25:29.900)

So these splits have become really unbalanced

Lex Fridman (2:25:33.460)

and the middle players, the ad tech vendors

Brendan Eich (2:25:35.720)

are taking out way too much money.

Lex Fridman (2:25:37.420)

And they're doing something worse, which has been noticed.

Brendan Eich (2:25:39.700)

They're letting not just the malware vendors,

Lex Fridman (2:25:44.060)

but also the ad fraud side, which fakes the publishers

Lex Fridman (2:25:47.220)

and clickbait merchants come in and steal traffic

Lex Fridman (2:25:52.120)

from good sites.

Brendan Eich (2:25:53.220)

Because once you have a certain audience identified

Lex Fridman (2:25:55.640)

at one site, Jason Calconas told me this

Brendan Eich (2:25:57.500)

about his experience with, I guess it was in Gadget,

Lex Fridman (2:25:59.900)

I forget which site he was running.

Lex Fridman (2:26:01.620)

But once he started using an ad partner

Lex Fridman (2:26:04.180)

that was sharing his audience information

Brendan Eich (2:26:06.200)

across multiple sites, he saw his competitors

Lex Fridman (2:26:08.420)

stealing all his traffic.

Lex Fridman (2:26:09.540)

And then what's worse is the clickbait sites

Lex Fridman (2:26:11.540)

that just have much cheaper rates steal all that traffic.

Lex Fridman (2:26:15.500)

And that facilitates fraud, facilitates fake news,

Lex Fridman (2:26:20.300)

all sorts of problems.

Lex Fridman (2:26:21.340)

So Gray blocks it and then we give users the ability

Lex Fridman (2:26:24.580)

to give back and because we invented

Brendan Eich (2:26:27.460)

the basic attention token on Ethereum,

Lex Fridman (2:26:28.940)

we can do this three way split.

Lex Fridman (2:26:31.020)

And we can give users a share of the revenue.

Lex Fridman (2:26:33.620)

And if they want to take it out, they can.

Brendan Eich (2:26:35.500)

Now, unfortunately for us and for all blockchain,

Lex Fridman (2:26:38.240)

the regulators are saying,

Brendan Eich (2:26:39.800)

we're gonna have to know who you are.

Lex Fridman (2:26:41.300)

There's the Treasury Department's FinCEN agency.

Brendan Eich (2:26:47.320)

There's the Office of Foreign Asset Controls, OFAC.

Lex Fridman (2:26:50.780)

There's the other regulators in the federal government

Brendan Eich (2:26:55.260)

that take a very dark look at things like money laundering

Lex Fridman (2:26:58.260)

and sending money to someone named Osama bin Laden.

Lex Fridman (2:27:01.260)

So compliance starts to come in.

Lex Fridman (2:27:03.540)

And even now they're threatening for pure Bitcoin

Brendan Eich (2:27:06.100)

sending to some address.

Lex Fridman (2:27:07.820)

If you're a Coinbase, you're gonna have to know

Brendan Eich (2:27:10.260)

who's at that address.

Lex Fridman (2:27:11.340)

You're gonna have to start.

Brendan Eich (2:27:12.180)

Like the actual identities of people involved.

Lex Fridman (2:27:13.580)

Yeah, now with Coinbase members,

Brendan Eich (2:27:14.900)

you sign up and they know you

Lex Fridman (2:27:15.900)

and they comply with the regulations.

Brendan Eich (2:27:16.980)

They're a regulated money services business.

Lex Fridman (2:27:20.220)

And, but if somebody's using their own self custody,

Lex Fridman (2:27:25.600)

so called self custodial wallet

Lex Fridman (2:27:27.240)

where they have the hardware private key

Lex Fridman (2:27:29.340)

and they're not named and they want to send to that address,

Lex Fridman (2:27:33.620)

our friends in the federal government are talking about

Brendan Eich (2:27:35.080)

requiring at some threshold and knowing who that is.

Lex Fridman (2:27:37.560)

So.

Brendan Eich (2:27:38.400)

Some threshold that's unreasonable.

Lex Fridman (2:27:40.460)

It's not that big.

Brendan Eich (2:27:42.220)

Yeah.

Lex Fridman (2:27:43.060)

Yeah, I don't know how this will play out.

Brendan Eich (2:27:43.880)

I think crypto is here to stay.

Lex Fridman (2:27:44.820)

I think the beauty of being able to send peer to peer

Brendan Eich (2:27:47.260)

without any bank in the middle,

Lex Fridman (2:27:48.500)

without any huge wire charge and two day delay

Lex Fridman (2:27:51.620)

and all that nonsense.

Lex Fridman (2:27:52.460)

It's beautiful and I've used it and I love it.

Lex Fridman (2:27:54.340)

But we're pragmatists are brave about crypto.

Lex Fridman (2:27:56.400)

And we realized that anything like a revenue split,

Brendan Eich (2:27:59.580)

we can't facilitate without being licensed in a certain way

Lex Fridman (2:28:02.180)

and it requires knowing who the user is.

Lex Fridman (2:28:03.880)

So our default mode doesn't know who the user is.

Lex Fridman (2:28:06.460)

It instead imputes to the user's browser,

Brendan Eich (2:28:09.540)

some of the revenue and allows that browser

Lex Fridman (2:28:11.820)

to steer it back to the creators.

Lex Fridman (2:28:14.940)

And we do have to identify the creators.

Lex Fridman (2:28:16.680)

But as things improve and who knows how it'll play out,

Brendan Eich (2:28:20.460)

there should become a day when this full vision

Lex Fridman (2:28:22.820)

can be done more fully on a blockchain.

Lex Fridman (2:28:25.740)

But regulations and the practicalities of today's blockchains,

Lex Fridman (2:28:29.260)

which are not that fast and not anonymous over time,

Brendan Eich (2:28:32.700)

you fingerprint yourself over time.

Lex Fridman (2:28:34.780)

We do some of this with the browser.

Lex Fridman (2:28:36.260)

So one of the ideas of the basic attention token

Lex Fridman (2:28:39.780)

is to make a hybrid system

Brendan Eich (2:28:41.880)

that's stronger than blockchain alone.

Lex Fridman (2:28:43.420)

It's the browser and the blockchain.

Lex Fridman (2:28:45.900)

And the browser is this trusted endpoint software.

Lex Fridman (2:28:48.420)

It's this universal app.

Brendan Eich (2:28:49.980)

Everyone uses browsers.

Lex Fridman (2:28:51.120)

The bigger the screen, the more you're in the browser

Lex Fridman (2:28:53.020)

and the less you install fat clients for things.

Lex Fridman (2:28:56.140)

I use Slack on Mac OS and it's like a browser.

Brendan Eich (2:29:00.140)

It's based on an electron framework we used to use.

Lex Fridman (2:29:02.140)

And it's just, it's not that great.

Brendan Eich (2:29:04.740)

Some of the people at Brave use Slack in Brave as a...

Lex Fridman (2:29:07.780)

In the browser, yeah.

Brendan Eich (2:29:08.620)

In the browser, yeah.

Lex Fridman (2:29:09.440)

I use that often, yeah.

Lex Fridman (2:29:10.340)

And I noticed on the iPad, I use apps less.

Lex Fridman (2:29:13.140)

The smaller the screen, the browser got handicapped

Brendan Eich (2:29:16.460)

by Apple and Android both.

Lex Fridman (2:29:17.960)

And it also can be slower or not have the right affordances,

Brendan Eich (2:29:22.700)

the interface with the security limited APIs.

Lex Fridman (2:29:26.780)

But in principle, with the right permissioning,

Brendan Eich (2:29:28.740)

you can make the web browser just as good as any app.

Lex Fridman (2:29:31.680)

You make it be a super app.

Lex Fridman (2:29:33.380)

And that's part of our mission at Brave.

Lex Fridman (2:29:35.120)

So we want to have the economics that got captured

Brendan Eich (2:29:37.820)

by these big tech companies through tracking

Lex Fridman (2:29:40.060)

and through social networks.

Brendan Eich (2:29:41.380)

We want to block that for your own safety

Lex Fridman (2:29:43.300)

and then let you opt into a cleaner world

Brendan Eich (2:29:45.480)

where you keep your data defended in your browser

Lex Fridman (2:29:47.980)

and you can actually realize value from it.

Lex Fridman (2:29:49.840)

So the way our ad system works,

Lex Fridman (2:29:51.300)

I mentioned it being private, but how does that work?

Brendan Eich (2:29:53.340)

We don't see your data at all.

Lex Fridman (2:29:54.980)

All browsers are sort of the mother of all data feeds,

Brendan Eich (2:29:57.740)

your history, all your searches at all engines.

Lex Fridman (2:30:00.700)

Each engine sees the queries you send to it,

Lex Fridman (2:30:02.560)

but it doesn't see the others,

Lex Fridman (2:30:03.500)

but the browser sees them all.

Brendan Eich (2:30:05.220)

Machine learning in the browser that you can opt into

Lex Fridman (2:30:07.540)

can study all that in a very complete way

Lex Fridman (2:30:10.140)

and do a better job than Google does.

Lex Fridman (2:30:11.980)

Google has cookie and scripts across the web

Brendan Eich (2:30:15.180)

from acquiring DoubleClick, they have YouTube,

Lex Fridman (2:30:17.860)

they have Android, they have search,

Brendan Eich (2:30:19.700)

which is still their big revenue lane,

Lex Fridman (2:30:21.580)

but they don't see everything.

Brendan Eich (2:30:22.620)

The browser sees everything.

Lex Fridman (2:30:23.780)

And if it can do a good job locally,

Lex Fridman (2:30:25.180)

and this is not advanced machine learning,

Lex Fridman (2:30:26.660)

this is not TensorFlow, this is like SVMs now,

Brendan Eich (2:30:29.620)

naive Bayes, then you can match intense signals,

Lex Fridman (2:30:35.020)

intense signals from those data feeds,

Brendan Eich (2:30:37.180)

searches, the queries, the history, how much you're

Lex Fridman (2:30:40.220)

scrolling down a page, how much you redid a search.

Brendan Eich (2:30:43.860)

It's all blind browser algorithm, we don't see that data.

Lex Fridman (2:30:47.020)

And then pick the best ad from a fixed catalog per day.

Lex Fridman (2:30:50.820)

And the catalog is fixed across a large population per day,

Lex Fridman (2:30:53.460)

and it only updates once a day,

Brendan Eich (2:30:55.100)

because new offers come in and old ones expire,

Lex Fridman (2:30:57.620)

sometimes every week or every month.

Lex Fridman (2:30:59.620)

And that catalog, and there can be many such catalogs,

Lex Fridman (2:31:02.740)

is sold by our direct sales team.

Lex Fridman (2:31:06.220)

And so we're making an anonymous audience available

Lex Fridman (2:31:09.420)

to advertisers without the advertisers tracking them.

Brendan Eich (2:31:12.020)

Instead, each browser is a little machine learning system

Lex Fridman (2:31:15.300)

that's picking the best catalog entry.

Lex Fridman (2:31:17.180)

Now, the catalog is not the ads, those are big, right?

Lex Fridman (2:31:19.340)

It's a video or a webpage, it's just the link

Brendan Eich (2:31:21.420)

to an edge cache, and there are many such edge caches.

Lex Fridman (2:31:23.820)

We're not trying to protect them from seeing

Brendan Eich (2:31:25.620)

your IP address, it's not really feasible.

Lex Fridman (2:31:27.780)

We could use Tor, but we don't yet.

Lex Fridman (2:31:30.260)

And then some keywords about the ad.

Lex Fridman (2:31:31.980)

So it's basically like metadata and a link.

Lex Fridman (2:31:34.900)

And that's what the catalog consists of,

Lex Fridman (2:31:36.340)

and that's what the machine learning picks.

Lex Fridman (2:31:37.460)

And the machine learning is learning

Lex Fridman (2:31:39.020)

about the use specifically locally

Brendan Eich (2:31:41.820)

in order to choose from the catalog of different ads.

Lex Fridman (2:31:44.180)

Couldn't this possibly be like a multi billion dollar,

Lex Fridman (2:31:47.780)

isn't this taken on the Google ad?

Lex Fridman (2:31:50.980)

Could be.

Lex Fridman (2:31:51.820)

So like what, I mean, one question to ask,

Lex Fridman (2:31:54.580)

there seems to be some really profound ideas here

Brendan Eich (2:31:57.380)

that are different than what the internet has grown up to be.

Lex Fridman (2:32:01.900)

If Brave or something like Brave,

Brendan Eich (2:32:05.340)

the ideas, the fundamental philosophical ideas

Lex Fridman (2:32:07.580)

underlying Brave went out and runs 95% of the internet,

Lex Fridman (2:32:13.100)

how does that change the, what are the major things

Lex Fridman (2:32:18.500)

these changes about the internet?

Lex Fridman (2:32:19.620)

So social networks and then the creatives,

Lex Fridman (2:32:21.620)

like YouTube creators and all that kind of stuff.

Lex Fridman (2:32:23.660)

So let's talk about that.

Lex Fridman (2:32:24.580)

First of all, if Brave gets 95%,

Brendan Eich (2:32:26.260)

I'm gonna demand a recount, because I won't believe it.

Lex Fridman (2:32:29.100)

I don't know, I think we're trying to put things

Brendan Eich (2:32:31.980)

into web standards that can be standardized across browsers.

Lex Fridman (2:32:34.980)

So the main value of Brave will be the trust users have in us

Lex Fridman (2:32:38.940)

and our ability to give the best deal to users.

Lex Fridman (2:32:41.340)

So 70% of the gross ad revenue we give to the user.

Lex Fridman (2:32:45.140)

And if they go through that KYC process I mentioned,

Lex Fridman (2:32:47.940)

they can take it out.

Brendan Eich (2:32:48.940)

They can also give it back, they can take some out,

Lex Fridman (2:32:50.620)

give the rest back.

Brendan Eich (2:32:51.460)

They can add basic attention tokens to give back.

Lex Fridman (2:32:54.660)

Some of them turn off the ads,

Brendan Eich (2:32:56.140)

cause they just don't like ads,

Lex Fridman (2:32:57.220)

but they put in $20 a month.

Lex Fridman (2:32:58.980)

But I believe Zuko of Zcashframe does that.

Lex Fridman (2:33:01.140)

And that's very generous,

Brendan Eich (2:33:02.060)

because the browser is just anonymously

Lex Fridman (2:33:04.100)

based on his browsing, sort of keeping score

Brendan Eich (2:33:06.260)

on how much time he spent on this video, on that website.

Lex Fridman (2:33:09.540)

And if those sites verify in sort of a,

Brendan Eich (2:33:13.260)

like getting a domain certificate fashion,

Lex Fridman (2:33:15.240)

they can get paid, they can get part of his $20 a month.

Lex Fridman (2:33:19.260)

So that vision could go big.

Lex Fridman (2:33:21.140)

And if it does, I hope it's across multiple browsers.

Brendan Eich (2:33:23.400)

I don't know that they'll all compete well

Lex Fridman (2:33:25.980)

on the quality of the ads, the quality of the ad blocking

Lex Fridman (2:33:29.020)

and tracking protection.

Lex Fridman (2:33:30.340)

Those are subject to competition.

Brendan Eich (2:33:31.860)

It'll take a while to standardize them.

Lex Fridman (2:33:33.380)

But I think that would be a better world.

Brendan Eich (2:33:35.380)

It would have less counterparty risk,

Lex Fridman (2:33:38.220)

fewer fee takers in the middle, really just the browser.

Brendan Eich (2:33:40.740)

We're taking 30%, sort of the app store split.

Lex Fridman (2:33:44.580)

And if we get bigger, maybe we can take even less.

Brendan Eich (2:33:47.420)

Social networks, creators.

Lex Fridman (2:33:49.540)

If you look at YouTubers, a lot of them are the indies

Brendan Eich (2:33:52.220)

that are getting some size are getting sponsorship deals.

Lex Fridman (2:33:56.700)

They're using Patreon.

Brendan Eich (2:33:58.840)

They're encouraging people to subscribe

Lex Fridman (2:34:01.420)

and give them regular money through Patreon.

Lex Fridman (2:34:04.300)

But that's centralized through Patreon.

Lex Fridman (2:34:06.080)

So there's censorship hazards, there's a 5% fee.

Lex Fridman (2:34:09.840)

What if that were a web standard?

Lex Fridman (2:34:10.960)

What if Brave pioneered it first and we took 3%?

Lex Fridman (2:34:14.500)

And we did it in a way that was through your browser

Lex Fridman (2:34:17.460)

so we couldn't censor it.

Brendan Eich (2:34:19.180)

That's brilliant.

Lex Fridman (2:34:20.220)

Do you think it could be standardized across browsers?

Brendan Eich (2:34:22.780)

Can Internet Explorer come in again and...

Lex Fridman (2:34:25.580)

Yeah, protocols are easy to copy

Brendan Eich (2:34:27.900)

in that they're meant to be interoperable.

Lex Fridman (2:34:29.500)

So there's a risk there.

Lex Fridman (2:34:31.100)

And the loyal users might be tricked into leaving you.

Lex Fridman (2:34:33.980)

Or they might, because of that distribution power,

Brendan Eich (2:34:36.180)

you might end up getting stomped.

Lex Fridman (2:34:38.260)

I don't know, I can't predict the future.

Brendan Eich (2:34:39.640)

I think antitrust is back on the case finally in the US.

Lex Fridman (2:34:42.300)

And certainly in Europe, G Comp is doing its thing.

Lex Fridman (2:34:45.420)

So I'm hopeful that we'll have a period of innovation.

Lex Fridman (2:34:48.820)

People were talking, like Elizabeth Warren was talking

Brendan Eich (2:34:50.620)

about breaking up the tech companies very clearly.

Lex Fridman (2:34:53.700)

Now she didn't win, and I suspect that won't happen,

Lex Fridman (2:34:56.460)

but I also suspect that Google might be smart enough

Lex Fridman (2:34:58.400)

to see they should do something more

Brendan Eich (2:35:01.040)

than just put privacy perfume on Chrome.

Lex Fridman (2:35:03.380)

They should maybe get rid of DoubleClick or something,

Brendan Eich (2:35:05.300)

divest something.

Lex Fridman (2:35:06.700)

I don't know, it might happen.

Lex Fridman (2:35:07.780)

So Brave might inspire Google to completely change

Lex Fridman (2:35:10.380)

the way they're doing things in the browser?

Brendan Eich (2:35:11.420)

They're already doing something,

Lex Fridman (2:35:12.780)

you may have read about called the Privacy Sandbox

Brendan Eich (2:35:15.460)

or Flock, which they have this bird metaphor going,

Lex Fridman (2:35:20.020)

Turtle Dove, Fledge.

Lex Fridman (2:35:22.780)

But these systems have been very Googley,

Lex Fridman (2:35:25.860)

kind of overengineered, and yet,

Brendan Eich (2:35:27.820)

depending on differential privacy,

Lex Fridman (2:35:29.260)

which has weakness over time, if you know how that works,

Brendan Eich (2:35:31.420)

it's kind of injecting noise to hide you in a crowd,

Lex Fridman (2:35:34.260)

but over time an adversary can pull you out of the crowd.

Brendan Eich (2:35:37.660)

This doesn't look like it's gonna become a standard.

Lex Fridman (2:35:39.380)

Like Apple, Brave, Mozilla, we're not gonna just say,

Brendan Eich (2:35:42.740)

oh, Google, you saved us.

Lex Fridman (2:35:44.700)

You've invented the Privacy Sandbox,

Lex Fridman (2:35:46.060)

so we'll all just adopt it.

Lex Fridman (2:35:47.660)

Not gonna be that easy.

Brendan Eich (2:35:48.580)

It's gonna be more like pieces of what we do in Brave,

Lex Fridman (2:35:51.540)

the synonymous ad matching or the blind signature

Brendan Eich (2:35:54.420)

cryptography we use to confirm the ad impressions.

Lex Fridman (2:35:57.020)

That's David Chow's invention.

Brendan Eich (2:35:59.260)

That could get standardized.

Lex Fridman (2:36:00.340)

In fact, some of that is being standardized.

Brendan Eich (2:36:02.220)

Even Google's in favor of so called trust tokens,

Lex Fridman (2:36:04.460)

which are Chowmian blind signature certs.

Lex Fridman (2:36:07.460)

But they're not using them for ad confirmations

Lex Fridman (2:36:09.300)

because they don't wanna blow up their own business.

Lex Fridman (2:36:11.460)

And they need to let some of the publishers they serve

Lex Fridman (2:36:14.740)

have other ad tech scripts on the page.

Lex Fridman (2:36:17.340)

And so they're kind of caught.

Lex Fridman (2:36:18.620)

And this is something I realized doing Brave.

Brendan Eich (2:36:19.860)

I thought, what's Google's innovators dilemma

Lex Fridman (2:36:22.900)

apart from just being mature and having trouble innovating?

Brendan Eich (2:36:26.300)

It's that they have come to depend on this ad tech system

Lex Fridman (2:36:30.620)

that has all these vendors that publishers rely on

Brendan Eich (2:36:34.980)

because publishers aren't technical enough.

Lex Fridman (2:36:36.380)

And I feel for the publishers,

Lex Fridman (2:36:37.980)

but I realized the users have to come first.

Lex Fridman (2:36:39.900)

And if you give the users a better browser that's faster,

Brendan Eich (2:36:43.700)

then you'll get enough users to give back

Lex Fridman (2:36:46.220)

or support publishers.

Brendan Eich (2:36:47.700)

The speed and the battery savings

Lex Fridman (2:36:49.420)

and the data plan savings are significant.

Brendan Eich (2:36:51.140)

There's so much bad JavaScript involved in ad tech

Lex Fridman (2:36:53.340)

that if you block it, you sort of chop off

Brendan Eich (2:36:55.740)

what's called the programmatic waterfall,

Lex Fridman (2:36:57.500)

which chains a bunch of requests.

Brendan Eich (2:36:59.140)

Yeah, that's one of the incredible things about Brave.

Lex Fridman (2:37:01.220)

I guess you're saying you should attribute it

Brendan Eich (2:37:03.220)

to the fact that the messy JavaScript, no offense.

Lex Fridman (2:37:06.620)

No offense.

Brendan Eich (2:37:07.460)

Not my solution.

Lex Fridman (2:37:08.300)

I mean, Brave just feels faster.

Brendan Eich (2:37:13.900)

Even then, I mean, Chrome was fast.

Lex Fridman (2:37:16.420)

One of the things that it was like impressive

Brendan Eich (2:37:18.420)

is it showed that browsers can be really fast

Lex Fridman (2:37:21.620)

and Brave is even faster than that, which is incredible.

Lex Fridman (2:37:24.260)

And it saves the network, which means data plan.

Lex Fridman (2:37:26.940)

It saves battery because the radio consumes your battery

Brendan Eich (2:37:29.420)

when it's running more to do those requests.

Lex Fridman (2:37:31.140)

And it's just stunning how many there are.

Brendan Eich (2:37:32.940)

Like some of my Google friends were like,

Lex Fridman (2:37:34.600)

oh, that's just that bad site.

Brendan Eich (2:37:35.860)

They'll fix it.

Lex Fridman (2:37:36.700)

And you actually do a survey of web pages

Brendan Eich (2:37:38.580)

that they're like mostly like that.

Lex Fridman (2:37:40.340)

I know Google engineers could make everything

Brendan Eich (2:37:42.460)

super efficient, but they can't,

Lex Fridman (2:37:44.300)

especially in antitrust court, do it.

Brendan Eich (2:37:45.780)

They cannot take over all the publishers and do that.

Lex Fridman (2:37:48.460)

They're trying with accelerated mobile profile, AMP.

Brendan Eich (2:37:51.780)

They're trying to pull publishers.

Lex Fridman (2:37:53.460)

They're like, oh, you poor publishers

Brendan Eich (2:37:54.620)

don't know how to make your pages fast.

Lex Fridman (2:37:56.900)

Put them on our AMP system.

Brendan Eich (2:37:58.180)

We'll give you extra placement in the search carousel.

Lex Fridman (2:38:00.660)

That's an antitrust problem for one.

Lex Fridman (2:38:02.420)

But it's also publishers we talk to hate it

Lex Fridman (2:38:04.340)

because it degrades their brand.

Brendan Eich (2:38:06.020)

Now they look like a gig writer wrote a piece

Lex Fridman (2:38:08.020)

that's got Google's framing an AMP URL on top of it.

Lex Fridman (2:38:12.140)

And they're trying to fix that too.

Lex Fridman (2:38:13.340)

But it just looks like Google's Borgifying

Brendan Eich (2:38:16.340)

all these publishers and they don't want to be plugged

Lex Fridman (2:38:18.160)

into the Borg cube.

Brendan Eich (2:38:19.020)

They want to build up their own brand and have loyal readers.

Lex Fridman (2:38:22.040)

So, you know, I'm in favor of giving the users power

Brendan Eich (2:38:25.580)

to help all the publishers and the little platoons

Lex Fridman (2:38:28.740)

and the creators.

Lex Fridman (2:38:29.580)

And so we talked about Patreon.

Lex Fridman (2:38:30.860)

What about social networks?

Brendan Eich (2:38:31.980)

Well, they're inherently like search, a global algorithm.

Lex Fridman (2:38:35.420)

You're trying to find friends of friends.

Brendan Eich (2:38:37.100)

You're doing the transitive closure of a graph

Lex Fridman (2:38:39.340)

induced by this friend of relation.

Lex Fridman (2:38:41.700)

But you should own your friend relation.

Lex Fridman (2:38:45.320)

You should own your posts.

Brendan Eich (2:38:46.980)

They shouldn't be owned by somebody else

Lex Fridman (2:38:48.660)

who can take them down or censor them.

Lex Fridman (2:38:51.060)

And your friend relations, you should be able

Lex Fridman (2:38:52.540)

to find those friends on other networks.

Lex Fridman (2:38:54.220)

And that's why I've tweeted about this.

Lex Fridman (2:38:55.420)

I haven't built it yet.

Lex Fridman (2:38:56.700)

What if the browser could keep track of those for you?

Lex Fridman (2:38:59.740)

What if the browser could maybe combine Facebook and Twitter

Lex Fridman (2:39:03.500)

and you could find your friends on both

Lex Fridman (2:39:05.300)

and you could have a sort of multi.

Lex Fridman (2:39:06.780)

So that relationship is not owned by Facebook or Twitter.

Lex Fridman (2:39:09.900)

It's owned by you through the browser.

Brendan Eich (2:39:12.020)

They'll have terms of use and they'll say they own it.

Lex Fridman (2:39:14.020)

But if they zap you on one and you're still on the other,

Brendan Eich (2:39:17.000)

your friends find you and the browser

Lex Fridman (2:39:18.500)

could preserve a combined view.

Brendan Eich (2:39:20.820)

You could resurrect almost across networks.

Lex Fridman (2:39:23.800)

It's something I wanna maybe quickly ask you about.

Brendan Eich (2:39:26.600)

On that front, there's been quite a lot of centralized,

Lex Fridman (2:39:31.340)

we talked about Wall Street Bets and then Robinhood.

Brendan Eich (2:39:35.260)

There's been centralized banning of different accounts

Lex Fridman (2:39:39.300)

and removing like Parler, for example, from AWS

Lex Fridman (2:39:42.620)

and this kind of overreach of centralized control.

Lex Fridman (2:39:46.740)

Is your hope that it's possible to,

Lex Fridman (2:39:49.360)

like what are your thoughts about that in general?

Lex Fridman (2:39:51.300)

And then is it possible to create tools

Brendan Eich (2:39:54.480)

that give individual people the power

Lex Fridman (2:39:56.700)

to fight back against overreach of such control?

Lex Fridman (2:40:00.060)

So we're talking about oligarchy, I do think.

Lex Fridman (2:40:02.660)

And if it controls a nation state, that's formidable.

Brendan Eich (2:40:05.260)

It's the tax and the police power, the military power.

Lex Fridman (2:40:08.220)

It means that you may have the great firewall of China.

Brendan Eich (2:40:10.860)

You may have people in China who are jailed

Lex Fridman (2:40:13.380)

because of their tweets, right?

Brendan Eich (2:40:14.880)

This is a serious threat.

Lex Fridman (2:40:15.740)

I can't minimize it or say that we'll win.

Brendan Eich (2:40:18.240)

I don't know how it's gonna go.

Lex Fridman (2:40:19.980)

But I do think, like I said earlier

Brendan Eich (2:40:21.660)

about the kind of reason people find ways around things.

Lex Fridman (2:40:24.000)

The internet routes around censorship.

Lex Fridman (2:40:25.820)

And this is not to endorse any particular bad faction.

Lex Fridman (2:40:28.820)

One of the things that happens

Brendan Eich (2:40:29.900)

when you try to wave the free speech flag too much,

Lex Fridman (2:40:33.020)

you say, I'm not gonna censor anything

Lex Fridman (2:40:34.620)

and you get colonized by terrible, terrible people.

Lex Fridman (2:40:37.900)

I don't care if you call them neo Nazis,

Brendan Eich (2:40:39.820)

some of them could be doing illegal things.

Lex Fridman (2:40:41.840)

And you don't want them colonizing

Brendan Eich (2:40:44.020)

because it'll ruin your reputation

Lex Fridman (2:40:45.220)

and destroy your business.

Lex Fridman (2:40:46.380)

So what you really want is that kind of

Lex Fridman (2:40:48.740)

user first subsidiarity, that subjectivity.

Brendan Eich (2:40:51.580)

I want my social networks to be composited

Lex Fridman (2:40:55.320)

in some multi social user interface

Brendan Eich (2:40:57.100)

where I don't lose track of people across networks.

Lex Fridman (2:40:59.420)

And if they leave one or they get banned from one,

Brendan Eich (2:41:01.620)

I can find them on another,

Lex Fridman (2:41:02.520)

I can still sort of thread them together.

Brendan Eich (2:41:04.360)

That's brilliant.

Lex Fridman (2:41:05.200)

And this didn't happen because browsers

Brendan Eich (2:41:07.740)

got captured by the central powers.

Lex Fridman (2:41:10.060)

Why did they get captured?

Brendan Eich (2:41:10.900)

Mostly because of search.

Lex Fridman (2:41:11.900)

And search is a central algorithm.

Lex Fridman (2:41:13.740)

So Larry Page said this too many years ago.

Lex Fridman (2:41:15.620)

He said, with search, you're giving up a little privacy

Brendan Eich (2:41:17.960)

by handing the query over to us.

Lex Fridman (2:41:19.820)

And we'll error correct it.

Brendan Eich (2:41:21.460)

Alan used to be a Google executive.

Lex Fridman (2:41:24.340)

He said, oh yeah, we used to laugh.

Brendan Eich (2:41:25.600)

They'd all be doing typos and they'd be typing

Lex Fridman (2:41:27.480)

the wrong word.

Lex Fridman (2:41:28.320)

And we'd be like, no dummy, type that query.

Lex Fridman (2:41:29.760)

And it's like, okay Google,

Brendan Eich (2:41:31.340)

might want to dial back that ego a little bit.

Lex Fridman (2:41:33.060)

But yes, you do see all the queries

Lex Fridman (2:41:35.260)

and you can improve them and you can find the best results.

Lex Fridman (2:41:37.340)

And that was Google's forte.

Brendan Eich (2:41:38.900)

When we did the Firefox deal in 2004,

Lex Fridman (2:41:40.580)

Google was really good.

Lex Fridman (2:41:42.380)

And over time, SEO, which is an adversarial game,

Lex Fridman (2:41:46.260)

and Google itself buying all these companies

Lex Fridman (2:41:48.580)

and crowding its own results page

Lex Fridman (2:41:50.100)

with its own tied in stuff.

Brendan Eich (2:41:53.160)

The YouTube.

Lex Fridman (2:41:54.000)

It's a slippery slope that happens

Brendan Eich (2:41:55.260)

when you have control over these kinds

Lex Fridman (2:41:58.380)

of really important mechanisms.

Brendan Eich (2:42:00.020)

Yeah, monopoly capitalism or cartel.

Lex Fridman (2:42:02.140)

You get this with the Robin Hoods and the hedge funds.

Brendan Eich (2:42:05.300)

You get sort of the money interests take over

Lex Fridman (2:42:07.940)

and kind of abuse their power and wear out their welcome.

Lex Fridman (2:42:10.100)

So how do you get around that?

Lex Fridman (2:42:11.940)

You have to have either new land to go to,

Brendan Eich (2:42:15.380)

which some people's ancestors, not mine,

Lex Fridman (2:42:18.220)

did to found the country.

Brendan Eich (2:42:20.820)

I'm mostly Irish, German.

Lex Fridman (2:42:22.780)

You have new virtual space people go to

Lex Fridman (2:42:26.020)

and that requires an ISP or a colo center

Lex Fridman (2:42:30.300)

or Amazon to host you.

Brendan Eich (2:42:32.060)

It requires domain name registrar who will not strike you.

Lex Fridman (2:42:35.980)

And so when Parler was taken down,

Brendan Eich (2:42:39.580)

I thought that was egregious.

Lex Fridman (2:42:40.860)

Parler, it was not well designed

Lex Fridman (2:42:42.980)

and I tried it out because I tried all these things,

Lex Fridman (2:42:45.900)

but I didn't use it.

Lex Fridman (2:42:46.740)

And I also felt they were being unfairly scored

Lex Fridman (2:42:49.380)

for not moderating because you can find tweets to this day

Brendan Eich (2:42:51.780)

that are horrendous and threaten all sorts of violence.

Lex Fridman (2:42:54.640)

Whereas Twitter, why isn't Twitter being taken down?

Lex Fridman (2:42:56.700)

But so it was very selective.

Lex Fridman (2:42:57.640)

It was the insiders who have the power

Brendan Eich (2:42:59.900)

are gonna take out the newcomer.

Lex Fridman (2:43:01.660)

And it looked bad, sort of like the hedge funds,

Brendan Eich (2:43:05.380)

shorting GameStop, looked bad.

Lex Fridman (2:43:08.060)

You're seeing a piece in Time Magazine this week

Brendan Eich (2:43:10.060)

that's like basically saying,

Lex Fridman (2:43:11.340)

yeah, we interfere with the election,

Lex Fridman (2:43:12.660)

but it was great, aren't we good?

Lex Fridman (2:43:14.580)

I don't know if you've seen this piece yet.

Brendan Eich (2:43:16.620)

If you tried to say that as a Trump supporter

Lex Fridman (2:43:18.980)

in November after the election,

Brendan Eich (2:43:21.500)

you'd get banned from Twitter.

Lex Fridman (2:43:22.900)

But now Time in its Twitter account is saying,

Brendan Eich (2:43:25.980)

we saved the day, it's AFL, CIO and big business,

Lex Fridman (2:43:30.260)

the Better Business Bureau got together

Lex Fridman (2:43:32.060)

and kept Trump from spreading fake news.

Lex Fridman (2:43:34.980)

So the country's kind of broken.

Brendan Eich (2:43:36.860)

I don't know how to fix that.

Lex Fridman (2:43:37.820)

The oligarchs have run wild in my opinion.

Lex Fridman (2:43:40.180)

And big tech is in the antitrust dock.

Lex Fridman (2:43:42.780)

What's gonna happen?

Brendan Eich (2:43:43.740)

I don't think they get out.

Lex Fridman (2:43:44.780)

I think some of the DOJ and certainly the state cases,

Brendan Eich (2:43:47.900)

because they're separate cases,

Lex Fridman (2:43:49.060)

are not gonna go away just

Brendan Eich (2:43:50.100)

because somebody got elected differently.

Lex Fridman (2:43:52.420)

And these are career prosecutors

Lex Fridman (2:43:53.900)

and they have a strong case.

Lex Fridman (2:43:55.940)

And Google's smart.

Lex Fridman (2:43:56.780)

And Microsoft almost got split up, right?

Lex Fridman (2:43:59.620)

The judge, Thomas Penfield Jackson, he overreached.

Brendan Eich (2:44:02.780)

He didn't hold a hearing about the remedy.

Lex Fridman (2:44:05.020)

He just said, I'm gonna break you up.

Lex Fridman (2:44:06.820)

And Microsoft appealed and the higher level court said,

Lex Fridman (2:44:10.580)

go back and figure this out.

Brendan Eich (2:44:11.660)

You're not breaking them up.

Lex Fridman (2:44:12.480)

You didn't even hold a hearing.

Lex Fridman (2:44:13.700)

And when they got back, Microsoft said,

Lex Fridman (2:44:15.460)

let's settle, let's settle, we don't wanna get broken up.

Brendan Eich (2:44:17.660)

Because Jackson was gonna make the Opsco,

Lex Fridman (2:44:20.500)

the operating system company,

Lex Fridman (2:44:21.620)

and the Appsco office, you know, Word and Excel.

Lex Fridman (2:44:24.780)

And that would have been a huge blow to Microsoft, so.

Lex Fridman (2:44:28.500)

But ultimately, I don't know if you're optimistic

Lex Fridman (2:44:31.140)

or cynical about the possibility of breaking up big tech.

Brendan Eich (2:44:34.540)

To me, I'm optimistic that tools like Brave,

Lex Fridman (2:44:40.060)

I love the idea of owning your friendships.

Brendan Eich (2:44:42.100)

Like users more and more owning the stuff

Lex Fridman (2:44:44.660)

is the only real way.

Brendan Eich (2:44:46.460)

Unfortunately, it's like the WallStreetBets subreddit

Lex Fridman (2:44:48.540)

is the only real way to fight decentralized power.

Brendan Eich (2:44:51.100)

You can't break them up with the regulation.

Lex Fridman (2:44:52.700)

It's very difficult.

Brendan Eich (2:44:53.540)

Certainly don't wanna wait for the law.

Lex Fridman (2:44:55.220)

Netscape was long dead or acquired by AOL

Lex Fridman (2:44:57.660)

and effectively dead.

Lex Fridman (2:44:58.500)

It was only Mozilla that returned Firefox to the market

Brendan Eich (2:45:01.980)

by the time that the US v. Microsoft case

Lex Fridman (2:45:04.140)

was finally settled and the penalties were put in place.

Lex Fridman (2:45:07.180)

And yet, antitrust has a role to play.

Lex Fridman (2:45:11.020)

Those penalties caused Microsoft

Brendan Eich (2:45:12.660)

to kind of turn away from the web.

Lex Fridman (2:45:14.220)

They did Windows Vista and they thought,

Brendan Eich (2:45:16.060)

the web's too painful.

Lex Fridman (2:45:16.900)

We got punished in court and we had to standardize things

Brendan Eich (2:45:19.700)

with those icky standards people.

Lex Fridman (2:45:21.220)

So they ran back to proprietary lock in

Lex Fridman (2:45:23.460)

and Windows Vista flopped.

Lex Fridman (2:45:24.660)

It was late, it was bloated.

Lex Fridman (2:45:26.620)

Longhorn, remember?

Lex Fridman (2:45:27.700)

Now, what I was gonna say, but Google's smart enough,

Brendan Eich (2:45:30.340)

they won't get split up.

Lex Fridman (2:45:31.620)

They'll split something out to get off the hook, I think.

Brendan Eich (2:45:35.780)

This is a complicated subject, but I myself was so,

Lex Fridman (2:45:39.660)

I decided to journey out from the world

Brendan Eich (2:45:43.220)

of being a researcher at MIT

Lex Fridman (2:45:45.260)

and potentially doing a startup myself.

Lex Fridman (2:45:48.340)

And I've been thinking of, you know,

Lex Fridman (2:45:51.420)

I wanted to come to Silicon Valley to do so.

Brendan Eich (2:45:53.660)

It's the land of the entrepreneur.

Lex Fridman (2:45:55.780)

And there's a lot of my friends,

Brendan Eich (2:45:57.980)

a lot of them are successfully,

Lex Fridman (2:45:59.580)

have been entrepreneurs themselves,

Brendan Eich (2:46:01.380)

have said, do not come to Silicon Valley.

Lex Fridman (2:46:04.540)

It'd be, you've started, you ran amazing teams of engineers.

Brendan Eich (2:46:09.100)

You started a lot of successful businesses.

Lex Fridman (2:46:12.100)

I wondered if you could comment on why a lot of people

Brendan Eich (2:46:15.860)

are leaving California.

Lex Fridman (2:46:17.180)

Is there something that could be fixed about California?

Brendan Eich (2:46:19.780)

If you were starting a business today,

Lex Fridman (2:46:22.700)

would you consider somewhere else,

Lex Fridman (2:46:24.780)

like Austin or some other place?

Lex Fridman (2:46:27.300)

Or is Silicon Valley still, is it just a little lull,

Brendan Eich (2:46:30.660)

everybody's being overdramatic during this particular year

Lex Fridman (2:46:34.100)

of the coronavirus and so on?

Brendan Eich (2:46:35.820)

I think, you know, even Austin's getting overheated, I hear.

Lex Fridman (2:46:39.220)

And I've had relatives and friends move to Texas

Brendan Eich (2:46:42.420)

within the last few months.

Lex Fridman (2:46:43.260)

So Texas as a whole is a big place.

Brendan Eich (2:46:45.860)

And, you know, people are moving to Florida.

Lex Fridman (2:46:48.460)

There's a big movement toward Miami,

Brendan Eich (2:46:50.700)

Peter Thielke, these people.

Lex Fridman (2:46:53.260)

The mayor has been very business friendly about it,

Brendan Eich (2:46:56.260)

which I think is just good politics.

Lex Fridman (2:46:58.460)

America is fundamentally a commercial republic.

Lex Fridman (2:47:00.660)

So you would think this would be what's happening.

Lex Fridman (2:47:02.540)

For a long time, California was the golden state.

Brendan Eich (2:47:04.420)

I came here in late 76 when I was a teenager.

Lex Fridman (2:47:06.820)

So it's in crushing debt due to the lockdowns.

Brendan Eich (2:47:11.380)

It's got the highest taxes.

Lex Fridman (2:47:13.460)

That's got to matter.

Brendan Eich (2:47:14.500)

People will do high taxes.

Lex Fridman (2:47:17.640)

It's got likely fires every year because of the deadfall.

Brendan Eich (2:47:21.580)

It's not global warming.

Lex Fridman (2:47:22.780)

It's because the forests weren't managed

Brendan Eich (2:47:24.220)

like they had been in the first part of the 20th century.

Lex Fridman (2:47:28.140)

Just, I would say corruption at all levels,

Brendan Eich (2:47:31.100)

especially up to the governor,

Lex Fridman (2:47:32.260)

who famously was eating at the French Laundry

Lex Fridman (2:47:34.460)

and claimed that the outside was inside.

Lex Fridman (2:47:36.740)

And they wore masks off and it was great.

Brendan Eich (2:47:39.540)

Do what I say, not what I do.

Lex Fridman (2:47:42.300)

Rules for thee, but not for me.

Brendan Eich (2:47:45.380)

When you see that in leadership,

Lex Fridman (2:47:47.580)

people either run or they get rid of the leadership.

Lex Fridman (2:47:50.400)

So there's a recall drive,

Lex Fridman (2:47:51.420)

which is about to reach the threshold.

Lex Fridman (2:47:53.500)

Or in the old days, they get their guns, right?

Lex Fridman (2:47:56.140)

You don't put up with this junk.

Lex Fridman (2:47:57.580)

But ultimately, the thing that made Silicon Valley

Lex Fridman (2:48:00.700)

a special place, it gave freedom to young kids,

Brendan Eich (2:48:04.600)

entrepreneurs, young minds, brave minds

Lex Fridman (2:48:08.100)

to think bold, to try different stuff.

Brendan Eich (2:48:10.820)

I mean, even if the taxes are high,

Lex Fridman (2:48:12.940)

so outside of financial stuff, outside of all of that.

Brendan Eich (2:48:15.540)

Housing's super expensive.

Lex Fridman (2:48:16.980)

Housing's super.

Lex Fridman (2:48:17.820)

So it's hard.

Lex Fridman (2:48:18.820)

Okay, everything about startups is hard.

Lex Fridman (2:48:19.660)

Peninsula was narrow and they didn't plan the roads, right?

Lex Fridman (2:48:22.820)

Yeah.

Brendan Eich (2:48:23.660)

They got rid of public transportation in LA,

Lex Fridman (2:48:26.020)

like the Who Framed Roger Rabbit cartoon show.

Brendan Eich (2:48:28.260)

They used to have trolley cars in Portland too.

Lex Fridman (2:48:30.860)

The oil companies and the DOD conspired to build highways

Lex Fridman (2:48:34.300)

and make cars dominant.

Lex Fridman (2:48:35.660)

And the rights of way are long gone.

Brendan Eich (2:48:37.420)

Like Elon's gonna go underground.

Lex Fridman (2:48:39.740)

And I wish him well.

Brendan Eich (2:48:40.740)

That's probably the only way to do it now.

Lex Fridman (2:48:42.780)

But is it still a place,

Lex Fridman (2:48:44.820)

do you think it's possible that Silicon Valley

Lex Fridman (2:48:46.500)

is still a place where magic happens?

Lex Fridman (2:48:48.180)

Where the next Google's built?

Lex Fridman (2:48:49.500)

Where the next, I mean, Brave is built where?

Brendan Eich (2:48:53.580)

I think all good things come to an end.

Lex Fridman (2:48:54.900)

I think the problem is Silicon Valley

Brendan Eich (2:48:56.660)

had strong network effects through Stanford,

Lex Fridman (2:48:58.820)

through the angel investor networks and the wealth effect.

Lex Fridman (2:49:02.300)

And originally you have to give the federal government credit

Lex Fridman (2:49:05.220)

like the ARPANET was a government project.

Brendan Eich (2:49:07.300)

Let's not kid ourselves.

Lex Fridman (2:49:08.140)

This wasn't wild free market, libertarian capitalism.

Brendan Eich (2:49:11.180)

This was all Cold War stuff.

Lex Fridman (2:49:14.020)

You had out of the academia, you had Shockley

Lex Fridman (2:49:17.060)

and then the Traders Aid and Fairchild and Intel.

Lex Fridman (2:49:21.060)

But now, when's the last fab that was built in the Valley?

Brendan Eich (2:49:25.040)

MicroUnity might've been the last, I don't know.

Lex Fridman (2:49:26.880)

I haven't followed.

Brendan Eich (2:49:27.720)

We built a fab in Sunnyvale and MicroUnity

Lex Fridman (2:49:31.340)

in starting early 90s.

Lex Fridman (2:49:33.300)

And now the fabs are overseas.

Lex Fridman (2:49:36.600)

And the one thing that I would say

Brendan Eich (2:49:38.820)

that the oligarchs have intentionally done

Lex Fridman (2:49:40.640)

in both parties is sort of labor

Lex Fridman (2:49:44.220)

and environmental protection law arbitrage

Lex Fridman (2:49:46.180)

by going where the labor is cheaper

Lex Fridman (2:49:47.580)

and the environmental laws aren't as strict.

Lex Fridman (2:49:49.460)

And that's polluted the hell out of parts of China,

Lex Fridman (2:49:52.020)

but it's made things, you can make cheaper junk.

Lex Fridman (2:49:54.940)

And this is not a story that's over yet.

Lex Fridman (2:49:58.900)

So what is Silicon Valley for now?

Lex Fridman (2:50:00.740)

It's for the network effect, the brain trust

Brendan Eich (2:50:03.300)

of who you know, the parties, the Stanford sort of network.

Lex Fridman (2:50:09.500)

That's fragile too over time, I'm afraid.

Brendan Eich (2:50:11.860)

Stanford, a lot of good professors are like,

Lex Fridman (2:50:16.100)

they still filter mainly based on socioeconomic status,

Lex Fridman (2:50:19.900)

but it's kind of a skate school.

Lex Fridman (2:50:22.180)

I had a friend hired out of Harvard 20 years ago at Netscape

Lex Fridman (2:50:26.980)

and we talked about Harvard and he said,

Lex Fridman (2:50:28.860)

yeah, there's still professors who do great on the curve.

Lex Fridman (2:50:30.700)

And I said, oh yeah, I don't think they're any doing that

Lex Fridman (2:50:32.340)

at Stanford anymore.

Lex Fridman (2:50:33.180)

And he said, yo, it was shocking.

Lex Fridman (2:50:35.100)

Some of the students got Cs and Ds and they were crying.

Brendan Eich (2:50:37.700)

It's like, yes, that's right, the precious deers

Lex Fridman (2:50:39.620)

can't take that at Stanford, so they get As and Bs.

Brendan Eich (2:50:42.420)

Now you look at China and people say what you all about China,

Lex Fridman (2:50:46.260)

they prove Russia too, a lot of math science training,

Brendan Eich (2:50:49.700)

a lot of engineering, a lot of people who are doing

Lex Fridman (2:50:52.540)

their coursework to get the As and the Bs.

Lex Fridman (2:50:55.220)

So I'm an American, I'm born on the 4th of July.

Lex Fridman (2:50:59.940)

American, yeah.

Brendan Eich (2:51:00.780)

4th of July, wow.

Lex Fridman (2:51:02.100)

And America, as I say, fundamentally

Brendan Eich (2:51:03.940)

is a commercial republic.

Lex Fridman (2:51:04.860)

You can try to make it something else.

Brendan Eich (2:51:06.340)

You can say it's the new Atlantis and mystify it.

Lex Fridman (2:51:09.020)

You could talk about it in a more, I think, correct way,

Brendan Eich (2:51:11.260)

which is 13 colonies that grew and then there's a lot

Lex Fridman (2:51:13.780)

of local or original design anyway,

Brendan Eich (2:51:16.740)

the federalist papers talk about this,

Lex Fridman (2:51:18.820)

there's a lot of subsidiarity,

Lex Fridman (2:51:20.660)

but that's been eroded over time.

Lex Fridman (2:51:23.940)

And like I say, a lot of the offshoring is hurt.

Lex Fridman (2:51:26.660)

So what happened with coronavirus?

Lex Fridman (2:51:28.860)

People working from home, and at first it was funny

Brendan Eich (2:51:30.940)

because I have friends at Google who used to grumble

Lex Fridman (2:51:33.540)

that not only did they have to come into the office,

Brendan Eich (2:51:35.420)

if they joined a different team that was centered

Lex Fridman (2:51:37.340)

in a different office, they had to move.

Brendan Eich (2:51:39.100)

Or if the VA team was reconstituted in Munich,

Lex Fridman (2:51:43.060)

which it was after Lars Bock just got tired of JavaScript,

Brendan Eich (2:51:46.700)

that they hired in Munich or they hired PhDs

Lex Fridman (2:51:48.780)

in Germany and moved them to Munich.

Brendan Eich (2:51:50.780)

With coronavirus, everyone's working from home

Lex Fridman (2:51:52.300)

and it's like, what a relief, I can work for Google

Brendan Eich (2:51:53.940)

from home.

Lex Fridman (2:51:54.820)

But then the next shoe dropped and people started asking

Brendan Eich (2:51:57.220)

Mark Zuckerberg, hey, can I move to my hometown

Lex Fridman (2:52:00.460)

in the Midwest?

Lex Fridman (2:52:01.580)

And he said, okay.

Lex Fridman (2:52:02.420)

And they said, oh, can I keep getting my Silicon Valley pay?

Brendan Eich (2:52:05.100)

No.

Lex Fridman (2:52:06.660)

We're gonna figure out what your cost of living there is

Lex Fridman (2:52:08.500)

and we're gonna adjust your pay accordingly.

Lex Fridman (2:52:10.300)

And these colonies and these little mini experiments

Brendan Eich (2:52:13.020)

that all combine to the big giant experiment,

Lex Fridman (2:52:15.140)

I have a, I don't know, I have this vision of America,

Brendan Eich (2:52:20.060)

which the country, I was born in Russia, like I said here,

Lex Fridman (2:52:22.180)

and this is truly a wonderful country.

Brendan Eich (2:52:24.980)

I wasn't born on the 4th of July, but I might as well be.

Lex Fridman (2:52:27.500)

People still flee here.

Brendan Eich (2:52:28.340)

I still, and I'm a red blooded American at this point.

Lex Fridman (2:52:31.740)

And I have a sense that we figured it out somehow.

Brendan Eich (2:52:35.420)

If Silicon Valley burns, another place will come up

Lex Fridman (2:52:37.940)

in this place that even more innovation and people will move

Lex Fridman (2:52:42.020)

and the remote work might change fundamentally how we work

Lex Fridman (2:52:46.460)

or it might not.

Brendan Eich (2:52:47.300)

It might just give you the freedom to then create

Lex Fridman (2:52:49.540)

many other small Silicon Valleys throughout the place,

Brendan Eich (2:52:52.220)

like Austin included, but other places as well.

Lex Fridman (2:52:54.940)

And we somehow figured it out and...

Brendan Eich (2:52:58.540)

I think that's true that there will be more mobility

Lex Fridman (2:53:01.660)

and maybe new places that come up.

Brendan Eich (2:53:03.460)

I don't know if Silicon Valley has passed some sell by date

Lex Fridman (2:53:07.940)

because it did hurt.

Brendan Eich (2:53:10.820)

The coronavirus hurt, the lockdowns hurt in the sense

Lex Fridman (2:53:13.420)

that part of what keeps things going is social.

Lex Fridman (2:53:15.860)

And so a lot of young people, even before coronavirus,

Lex Fridman (2:53:18.300)

moved to San Francisco.

Brendan Eich (2:53:19.660)

It was very strange to watch because in the 80s,

Lex Fridman (2:53:21.660)

we all lived in the Valley and it was less populated

Lex Fridman (2:53:24.860)

and San Francisco was grungier.

Lex Fridman (2:53:26.060)

It was more like Dirty Harry in the 70s.

Lex Fridman (2:53:28.620)

But by the 90s, and Jamie runs a nightclub there

Lex Fridman (2:53:31.660)

and he's talked about this, you had sort of wealthy

Brendan Eich (2:53:33.780)

tech people moving in south of market,

Lex Fridman (2:53:36.100)

fancy townhouses being built.

Lex Fridman (2:53:37.780)

And that's continued in such a point that it's almost like,

Lex Fridman (2:53:41.020)

what's the movie by the South African director Nils Jody

Lex Fridman (2:53:43.540)

Foster up in the space colony?

Lex Fridman (2:53:46.020)

Matt Damon is the guy on the earth who has to go

Brendan Eich (2:53:48.180)

up and anyway, it's about the stratification.

Lex Fridman (2:53:50.540)

It's about the great inequality that people

Brendan Eich (2:53:52.060)

in the space station have like amazing medical auto docs

Lex Fridman (2:53:55.540)

that can extend their life or save them cure cancer.

Brendan Eich (2:53:57.940)

People on earth are all suffering ground down in poverty.

Lex Fridman (2:54:02.820)

And that sort of happened while I was here.

Brendan Eich (2:54:06.260)

You saw a lot of money drive prices up

Lex Fridman (2:54:10.660)

along the narrow peninsula and the single people

Brendan Eich (2:54:12.540)

wanted nightlife so they were in the city

Lex Fridman (2:54:14.540)

and the condos in the city got super expensive.

Lex Fridman (2:54:16.700)

And I know even Google friends who are socially responsible

Lex Fridman (2:54:20.660)

say we should have more housing built.

Brendan Eich (2:54:22.820)

We should have yes in my backyard, not in my backyard.

Lex Fridman (2:54:25.860)

But that's not happening as far as I can tell.

Lex Fridman (2:54:28.460)

And from the government to the incumbent landowners

Lex Fridman (2:54:32.540)

and renters, it's just not happening.

Lex Fridman (2:54:34.500)

And that has to drive people away.

Lex Fridman (2:54:37.500)

I appreciate that people come here and you should wait

Brendan Eich (2:54:40.500)

for the prices to moderate, they will.

Lex Fridman (2:54:42.620)

But a lot of people are gonna go where the prices are lower.

Brendan Eich (2:54:47.360)

You, and sorry for silly questions here,

Lex Fridman (2:54:50.340)

but just looking back, you have created things,

Brendan Eich (2:54:54.380)

have been part of creating things

Lex Fridman (2:54:56.540)

that have transformed this world, the world of technology,

Brendan Eich (2:55:00.740)

perhaps more than almost anything else.

Lex Fridman (2:55:04.820)

But you're still a human being

Lex Fridman (2:55:07.580)

and unfortunately this ride ends.

Lex Fridman (2:55:10.900)

Do you ever think about your own mortality?

Brendan Eich (2:55:13.780)

Not too much, I mean, I'm a Roman Catholic

Lex Fridman (2:55:17.460)

so I am not afraid of death.

Brendan Eich (2:55:19.220)

I think a lot of people who have problems with death

Lex Fridman (2:55:25.140)

are suffering from some lack of either faith

Brendan Eich (2:55:30.700)

in their transcending death or maybe they don't have

Lex Fridman (2:55:33.380)

children or they feel like they get later in life

Lex Fridman (2:55:37.260)

and they feel like they've missed opportunities

Lex Fridman (2:55:39.220)

to do something that endures.

Lex Fridman (2:55:40.620)

And I sympathize with a lot because I'm old,

Lex Fridman (2:55:43.540)

I got married fairly old, so I understand all that.

Brendan Eich (2:55:46.440)

Nothing human is alien to me as Terrence said.

Lex Fridman (2:55:50.820)

But I don't fear it, no.

Lex Fridman (2:55:54.100)

What do you hope your legacy is?

Lex Fridman (2:55:56.300)

Yeah, it's gonna be JavaScript.

Brendan Eich (2:55:57.740)

I think, no, I think my legacy has more to do

Lex Fridman (2:56:01.100)

with my children and their children.

Brendan Eich (2:56:03.540)

I think it also has to do with web standards,

Lex Fridman (2:56:06.460)

it has to do with things like Brave,

Brendan Eich (2:56:08.340)

the things we did with Firefox.

Lex Fridman (2:56:10.060)

When we did, I'm not gonna oversell Brave

Lex Fridman (2:56:13.100)

but I think Brave is important and we'll continue

Lex Fridman (2:56:15.140)

to prove this in a way that counts

Brendan Eich (2:56:16.780)

for many decades to come.

Lex Fridman (2:56:18.120)

But even Firefox, whatever its future fortune,

Brendan Eich (2:56:21.420)

showed you can restart the browser market.

Lex Fridman (2:56:23.300)

This thing you said about people opting out

Lex Fridman (2:56:26.180)

and routing around, you don't need everybody to do that.

Lex Fridman (2:56:28.600)

It's more like Taleb's stubborn minorities that do that.

Brendan Eich (2:56:31.220)

It's the lead users, Erfron Hipple's lead users.

Lex Fridman (2:56:33.640)

You can be a few percent, you can tilt the market.

Lex Fridman (2:56:36.100)

And that can be done in spite of the incumbents,

Lex Fridman (2:56:39.380)

of moneyed interests not being in favor of what you're doing.

Lex Fridman (2:56:41.540)

So I think what we do with Firefox won't be forgotten

Lex Fridman (2:56:45.260)

and it needs to be done more and we're doing it with Brave.

Lex Fridman (2:56:47.140)

And you could argue that other projects are doing it.

Lex Fridman (2:56:48.780)

In some ways, blockchain is doing it.

Brendan Eich (2:56:50.780)

The Robinhood, take down the use of Robinhood

Lex Fridman (2:56:56.100)

by the Wall Street Bets kids, similar.

Lex Fridman (2:56:59.100)

So yeah, that kind of spirit endures.

Lex Fridman (2:57:01.300)

And I think it, in some ways it's American, right?

Brendan Eich (2:57:04.180)

It's not hard revolutionary.

Lex Fridman (2:57:06.540)

It's not trying to burn the past and destroy everything.

Brendan Eich (2:57:08.780)

It's more like we have these certain, let's say rights.

Lex Fridman (2:57:15.140)

We have duties too.

Lex Fridman (2:57:16.180)

So there's some debate about which comes first

Lex Fridman (2:57:17.780)

in American jurisprudence and the founding documents.

Lex Fridman (2:57:20.360)

But as long as things are working,

Lex Fridman (2:57:23.020)

we'll be like pragmatic Americans,

Brendan Eich (2:57:25.900)

like de Tocqueville described in his writings.

Lex Fridman (2:57:29.300)

But if things get too out of whack for one reason or another,

Brendan Eich (2:57:32.620)

too unequal, too oligarchic and abusive,

Lex Fridman (2:57:35.540)

we're gonna start our rights.

Lex Fridman (2:57:37.060)

And even a few of us can do it.

Lex Fridman (2:57:38.860)

And even in the American Revolution,

Brendan Eich (2:57:40.460)

it was the minority who fought and put their lives,

Lex Fridman (2:57:44.100)

treasure and sacred honor at stake.

Brendan Eich (2:57:45.620)

It was a bunch of people went to upper Canada,

Lex Fridman (2:57:48.860)

I think it was called, Ontario.

Brendan Eich (2:57:51.140)

Yeah, that's the beautiful, I mean,

Lex Fridman (2:57:52.860)

that is at the core where your work stands for

Brendan Eich (2:57:55.100)

is that a few people can have the power

Lex Fridman (2:57:58.180)

to transform society with just a few radical ideas,

Brendan Eich (2:58:01.580)

with just a little bit of code, change the world.

Lex Fridman (2:58:04.580)

Gotta do it.

Lex Fridman (2:58:05.420)

And that's empowering.

Lex Fridman (2:58:06.240)

That is the American way.

Brendan Eich (2:58:07.220)

That's why this country is, I believe,

Lex Fridman (2:58:09.380)

the greatest country on Earth.

Brendan Eich (2:58:10.460)

That's not over, or Matt says it too much,

Lex Fridman (2:58:13.040)

but I think some special things

Brendan Eich (2:58:15.900)

have already happened in this country

Lex Fridman (2:58:17.580)

and will continue to happen.

Lex Fridman (2:58:18.800)

And that's really exciting.

Lex Fridman (2:58:19.640)

And that spirit can continue no matter who comes here.

Brendan Eich (2:58:22.520)

They can adopt those folkways and that spirit.

Lex Fridman (2:58:25.520)

Brandon, I can't tell you how much,

Brendan Eich (2:58:27.580)

I was freaking out how much of an honor it is

Lex Fridman (2:58:29.540)

to talk to you.

Brendan Eich (2:58:30.380)

You're an incredible human being.

Lex Fridman (2:58:31.420)

It's one of my favorite conversations ever.

Brendan Eich (2:58:33.580)

Thank you so much for wasting all this time with me.

Lex Fridman (2:58:35.660)

I really appreciate it.

Brendan Eich (2:58:36.580)

Oh, it seems like a breeze.

Lex Fridman (2:58:38.060)

My pleasure.

Brendan Eich (2:58:39.740)

Thank you for listening to this conversation

Lex Fridman (2:58:41.380)

with Brandon and Ike.

Lex Fridman (2:58:42.580)

And thank you to our sponsors,

Lex Fridman (2:58:44.440)

Jordan Harbert's Show, Sun Basket Meal Delivery Service,

Brendan Eich (2:58:48.460)

BetterHelp Online Therapy,

Lex Fridman (2:58:50.260)

and Eight Sleep Self Cooling Mattress.

Brendan Eich (2:58:52.820)

Click the sponsor links to get a discount

Lex Fridman (2:58:54.860)

and to support this podcast.

Lex Fridman (2:58:57.060)

And now, let me leave you with some words from Jeff Atwood.

Lex Fridman (2:59:00.720)

Any app that can be written in JavaScript

Brendan Eich (2:59:03.700)

will eventually be written in JavaScript.

Lex Fridman (2:59:06.980)

Thank you for listening and hope to see you next time.

Brendan Eich (30:03.200)

the early, what became the CSS box model.

Lex Fridman (30:06.080)

And run scripts to make the thing livelier,

Brendan Eich (30:09.600)

respond to user input.

Lex Fridman (30:11.160)

And all that event driven programming

Brendan Eich (30:14.240)

was in part like HyperCard

Lex Fridman (30:16.080)

because HyperCard had this on event name syntax.

Lex Fridman (30:19.440)

And so that's why you have in JavaScript on click

Lex Fridman (30:21.680)

run together as the name of the event handler.

Lex Fridman (30:24.120)

And there's some funny ones on mouse over

Lex Fridman (30:26.060)

and on mouse out, people still complain about those.

Lex Fridman (30:28.200)

But there were many more events now over the years

Lex Fridman (30:31.060)

standardized, but it was a mix of event driven

Brendan Eich (30:33.980)

single threaded programming because it had to run

Lex Fridman (30:36.240)

in the main thread of the browser where the action is

Lex Fridman (30:38.960)

and Java never got there.

Lex Fridman (30:40.680)

Which meant Java could not interact easily

Brendan Eich (30:42.640)

or quickly or in a nested way with the document,

Lex Fridman (30:46.720)

with the objects reflected from the HTML document,

Brendan Eich (30:49.200)

with the tables and forms and so on.

Lex Fridman (30:50.960)

And that is one of the reasons I think JavaScript

Brendan Eich (30:53.960)

survived and Java kind of died.

Lex Fridman (30:55.320)

Java was in this plugin prison.

Brendan Eich (30:56.960)

It essentially was confined to a rectangle,

Lex Fridman (30:58.760)

the applet rectangle.

Lex Fridman (30:59.960)

And while we even built next year,

Lex Fridman (31:02.720)

Nick Thompson, a friend from SGI

Brendan Eich (31:04.840)

who was an intern grad student at CMU at the time,

Lex Fridman (31:08.640)

built the first version of Live Connect

Brendan Eich (31:10.280)

to glue Java and JavaScript together

Lex Fridman (31:12.640)

to deliver on that vision where you do have commerce

Brendan Eich (31:14.680)

between the data types in the heap.

Lex Fridman (31:18.040)

Did it work?

Brendan Eich (31:18.880)

It worked, but Java was in charge.

Lex Fridman (31:21.280)

JavaScript was in charge and Java was just these components,

Brendan Eich (31:23.640)

these helper objects.

Lex Fridman (31:24.480)

You might as well do everything in JavaScript.

Lex Fridman (31:25.920)

What happened over time, it's like an evolutionary filter.

Lex Fridman (31:27.960)

It just kind of, who needs the plugin?

Lex Fridman (31:29.600)

And in fact, Sun mismanaged Java as a plugin.

Lex Fridman (31:33.000)

They thought, oh, Netscape is giving us

Brendan Eich (31:35.520)

the distribution vehicle and we don't care about the browser.

Lex Fridman (31:38.400)

It's just about getting Java out there.

Lex Fridman (31:39.680)

And that was a big miscalculation.

Lex Fridman (31:41.220)

They then tried, because Microsoft's killing Netscape

Brendan Eich (31:43.840)

after years, they tried getting into Microsoft.

Lex Fridman (31:46.120)

And you may remember there was a Sun Microsoft deal

Brendan Eich (31:48.520)

which famously blew up.

Lex Fridman (31:50.640)

And Microsoft kicked Java out of Windows.

Lex Fridman (31:54.800)

And that's when they really pulled the trigger.

Lex Fridman (31:56.400)

I think they'd already evaluated it and liked it

Brendan Eich (31:58.120)

on Anders Helzberg's.NET and C Sharp

Lex Fridman (32:01.740)

and decided we're gonna just not have Java.

Brendan Eich (32:04.160)

We don't want any of that Sun stuff.

Lex Fridman (32:06.440)

We don't want the patent risk.

Brendan Eich (32:07.560)

We don't want, I'm not sure what all fights were about.

Lex Fridman (32:10.080)

There was some patent angle to it, I think.

Lex Fridman (32:12.040)

And up till then, Microsoft had been using Java components

Lex Fridman (32:16.920)

like in Outlook Web Access,

Brendan Eich (32:19.240)

which had a lot of JavaScript to be a webmail

Lex Fridman (32:21.480)

like Hotmail, like user interface.

Brendan Eich (32:23.120)

They had to call the mail server through HTTP

Lex Fridman (32:27.760)

and they used a Java object to do this.

Lex Fridman (32:31.360)

And when they gave the boot to Sun,

Lex Fridman (32:34.280)

suddenly the left hand gave the boot

Lex Fridman (32:37.240)

and the right hand said, we better do something else

Lex Fridman (32:39.160)

in Outlook Web Access.

Lex Fridman (32:40.080)

What are we gonna do?

Lex Fridman (32:40.920)

And they said, let's just add an ActiveX component,

Brendan Eich (32:43.640)

which is their own native way of embedding things

Lex Fridman (32:46.300)

in languages and it'll be what became XML HTTP request,

Brendan Eich (32:51.960)

which is now a web standard for calling asynchronously.

Lex Fridman (32:54.760)

And it's been replaced by the fetch API

Brendan Eich (32:57.040)

in HTML5 or HTML living document.

Lex Fridman (32:59.480)

But this whole lineage goes back to Java

Brendan Eich (33:02.400)

being successfully the loser and getting kicked out.

Lex Fridman (33:05.100)

And after Microsoft kicked it out, it was a plugin

Lex Fridman (33:07.680)

and you would find it required for like smart card banking

Lex Fridman (33:11.040)

in the Nordic countries where that was mandated by law,

Lex Fridman (33:14.700)

but really didn't get used much.

Lex Fridman (33:16.360)

Or there were pilots who used it for flight information,

Lex Fridman (33:21.560)

but Flash, which Netscape could have bought,

Lex Fridman (33:24.160)

but fortunately didn't.

Brendan Eich (33:25.560)

The early days.

Lex Fridman (33:26.400)

Yeah, we would have screwed it up.

Lex Fridman (33:28.080)

What year are we talking about with Flash?

Lex Fridman (33:29.520)

I think after the IPO, so it was probably late 95.

Brendan Eich (33:32.160)

Oh, Flash was around.

Lex Fridman (33:33.320)

Was it Adobe?

Brendan Eich (33:34.160)

No, it wasn't.

Lex Fridman (33:34.980)

No, it was called Future Splash

Lex Fridman (33:36.360)

and it was these brothers, Jonathan Gay,

Lex Fridman (33:38.720)

I think his name was, and he came knocking

Lex Fridman (33:41.720)

and the marketing guy at Netscape,

Lex Fridman (33:43.520)

who was screening the technology partners

Brendan Eich (33:45.840)

or wannabe acquisitions was brutal

Lex Fridman (33:47.800)

and just everybody wanted to get in on the Netscape,

Brendan Eich (33:50.320)

stock gravy train and he sent them packing

Lex Fridman (33:52.480)

and they ended up selling to Macromedia

Lex Fridman (33:55.120)

and Macromedia was where Flash was created.

Lex Fridman (33:57.320)

And the good thing about Macromedia was

Brendan Eich (33:58.960)

it was a tool company.

Lex Fridman (34:02.000)

So it invested in the best ideas, I think,

Brendan Eich (34:05.800)

which are still somewhat lost to us of Flash,

Lex Fridman (34:08.120)

the timeline, animation has sort of been a mutable function

Brendan Eich (34:12.500)

over time.

Lex Fridman (34:13.800)

They had the tooling around that too,

Brendan Eich (34:14.920)

like they had Dreamweaver, there's a Flash.

Lex Fridman (34:17.240)

Flash Director, there were a bunch of them.

Brendan Eich (34:18.840)

Yeah, I mean, yeah, that was a great.

Lex Fridman (34:20.840)

Flash Builder was one of the last ones.

Brendan Eich (34:22.800)

These tools were used by real artists

Lex Fridman (34:25.420)

and special effects people and designers.

Brendan Eich (34:27.440)

All the restaurant websites around 2005

Lex Fridman (34:29.760)

were done in Flash, which was,

Brendan Eich (34:32.280)

we were trying to do HTML5 at the same time.

Lex Fridman (34:34.080)

That was the Firefox era.

Brendan Eich (34:35.080)

We were trying to make the web capable enough

Lex Fridman (34:37.640)

you didn't need Flash, but if you recall,

Brendan Eich (34:39.400)

you go to a restaurant and it's like,

Lex Fridman (34:41.080)

this is kind of like a game or something.

Brendan Eich (34:42.400)

It's like a Flash, all the font looks small.

Lex Fridman (34:44.600)

So you didn't like Flash from the beginning.

Brendan Eich (34:45.880)

You're like, this doesn't feel right.

Lex Fridman (34:47.640)

Not really.

Brendan Eich (34:48.480)

I actually admire Flash's technology

Lex Fridman (34:50.000)

and I'm pretty pragmatic about these things.

Lex Fridman (34:51.700)

And I realized that it doesn't matter

Lex Fridman (34:54.840)

if your Delta bad hand like JavaScript was a rush job,

Brendan Eich (34:58.280)

or if you have Flash as a plugin

Lex Fridman (35:00.240)

and you can invest in the tools and make it pretty good.

Brendan Eich (35:03.620)

You should make it better for your users

Lex Fridman (35:05.720)

and grow it as best you can.

Lex Fridman (35:07.000)

And what happened with the browser

Lex Fridman (35:08.800)

due to Microsoft's monopoly abuse

Brendan Eich (35:11.040)

for which they were convicted.

Lex Fridman (35:12.600)

And even after that, until I think Firefox and then Chrome

Brendan Eich (35:17.440)

was people kept saying, oh, the web can't do X, can't do Y.

Lex Fridman (35:21.240)

We'll have to have a plugin.

Brendan Eich (35:22.240)

We'll have to have a new approach.

Lex Fridman (35:23.680)

We'll clean the slate and have a new web.

Lex Fridman (35:26.860)

And everyone who said that failed.

Lex Fridman (35:28.960)

And the reason they failed is because

Brendan Eich (35:30.180)

there's too much value in the web, this huge network.

Lex Fridman (35:34.080)

And the worse is better principle means

Brendan Eich (35:36.040)

that you can not only start bad, which they all sneer at,

Lex Fridman (35:39.460)

but get on first and get wide distribution,

Brendan Eich (35:41.920)

get sort of evolutionary advantage in priority of place,

Lex Fridman (35:47.080)

but you can also improve it over time.

Lex Fridman (35:49.320)

And so if you're gonna improve Flash,

Lex Fridman (35:51.080)

and for some reason Flash is now out of favor.

Brendan Eich (35:53.000)

Steve Jobs said you can't have Flash on the iPhone.

Lex Fridman (35:55.000)

That was probably the death knell.

Brendan Eich (35:56.740)

Put your energy into JavaScript.

Lex Fridman (35:58.100)

And that happened, right?

Lex Fridman (35:58.940)

So we did things at Mozilla with Adobe to improve,

Lex Fridman (36:02.400)

which bought my Macromedia, to improve Flash

Lex Fridman (36:04.860)

and to improve the version of JavaScript that was in Flash.

Lex Fridman (36:07.560)

We tried to standardize that.

Brendan Eich (36:09.000)

Oh, that's right.

Lex Fridman (36:09.840)

I'm getting ahead of myself with it.

Brendan Eich (36:10.660)

It was ES4.

Lex Fridman (36:11.500)

Yeah.

Brendan Eich (36:12.340)

That's right, that's right.

Lex Fridman (36:13.180)

Can we just rewind to the magical, like, you know.

Brendan Eich (36:17.000)

It's a special moment in the history of all of computing.

Lex Fridman (36:20.160)

We'll talk about it later, but it's arguable.

Brendan Eich (36:22.840)

It's possible that the entirety of the world

Lex Fridman (36:25.280)

will run in JavaScript at some point.

Lex Fridman (36:27.280)

So like, it's like those days,

Lex Fridman (36:30.040)

it would be interesting if you could just describe,

Brendan Eich (36:33.320)

actually zooming in on how the cake was baked

Lex Fridman (36:38.320)

from the several days that you were working on it.

Lex Fridman (36:41.580)

What was on your mind?

Lex Fridman (36:43.320)

How much coffee were you drinking?

Lex Fridman (36:45.380)

Were you nervous?

Lex Fridman (36:46.220)

Why, freaking out?

Brendan Eich (36:47.360)

I'll try to remember it.

Lex Fridman (36:48.360)

I mean, you're right.

Brendan Eich (36:49.200)

There are these pregnant moments you see in hindsight,

Lex Fridman (36:50.940)

maybe they're overrated,

Lex Fridman (36:52.200)

but like Hegel sees Napoleon on horseback at Jena

Lex Fridman (36:55.760)

and says, there's the world spirit on horse.

Lex Fridman (37:00.440)

And I knew that there was a chance to do it.

Lex Fridman (37:04.120)

Mark knew, and he was my executive sponsor

Lex Fridman (37:07.960)

and he was the one sort of brainstorming

Lex Fridman (37:10.800)

how the JavaScript should be right there in the page.

Brendan Eich (37:13.360)

That was important for him to say that.

Lex Fridman (37:14.760)

Cause I thought so too, but a lot of people were like,

Brendan Eich (37:17.360)

well, you can't write programming language

Lex Fridman (37:18.680)

in the middle of the markup.

Lex Fridman (37:20.120)

And indeed there are problems.

Lex Fridman (37:21.040)

If you did it naively, you'd see the code laid out

Brendan Eich (37:23.600)

as like random gibberish.

Lex Fridman (37:25.300)

So I had to figure out how to hide that.

Brendan Eich (37:26.600)

That was a challenge.

Lex Fridman (37:28.120)

Is that a breakthrough idea?

Brendan Eich (37:29.800)

I mean, so you and Mark thinking about this idea

Lex Fridman (37:32.580)

that you just inject code in the middle of the markup.

Brendan Eich (37:36.080)

Of the webpage.

Lex Fridman (37:36.920)

I consider kind of heretical.

Brendan Eich (37:38.000)

There was an SGML guru, I forget his name,

Lex Fridman (37:40.120)

but he corresponded with me and at first he was angry.

Brendan Eich (37:42.120)

He's like, you should have used a marked section.

Lex Fridman (37:44.760)

Why didn't you use a marked section?

Lex Fridman (37:45.960)

And I said, well, SGML marked sections are not part of HTML

Lex Fridman (37:48.600)

by the way, and they're not supported in the browser.

Lex Fridman (37:50.600)

And so I did some hack that was equivalent

Lex Fridman (37:52.540)

and over time you could do the proper SGML thing.

Lex Fridman (37:55.280)

But eventually he came around and it was again,

Lex Fridman (37:58.200)

sort of evolutionary necessity.

Brendan Eich (37:59.640)

It was almost like introgression, like the idea

Lex Fridman (38:02.920)

which Lynn Margulies, I think helped get across

Brendan Eich (38:06.980)

that we have to consider mutualism biology

Lex Fridman (38:10.640)

that maybe mitochondria were ancient prokaryotes

Brendan Eich (38:14.080)

that got into the cell and became beneficial.

Lex Fridman (38:17.420)

Somehow the same sort of thinking applies.

Brendan Eich (38:21.200)

You have to embed JavaScript in HTML.

Lex Fridman (38:23.040)

It's gonna be a good virus.

Brendan Eich (38:24.760)

It won't hurt you.

Lex Fridman (38:25.720)

The code becomes data in the sense

Brendan Eich (38:27.400)

it just gets carried along.

Lex Fridman (38:30.040)

But is there the side of the, so you were focusing

Brendan Eich (38:33.680)

on Netscape at that time.

Lex Fridman (38:35.080)

Doesn't the browser have to support interpret correctly

Lex Fridman (38:38.360)

this mix of HTML and whatever code?

Lex Fridman (38:41.440)

I had to hide it from old browsers,

Brendan Eich (38:42.720)

including Netscape 1.1, which was predominant then.

Lex Fridman (38:45.140)

So I used an HTML comment, but the inside the container

Brendan Eich (38:49.560)

that comment lived in the script tag,

Lex Fridman (38:51.080)

which is a new element, I could make different semantics

Brendan Eich (38:53.480)

in Netscape 2 where those HTML comment delimiters

Lex Fridman (38:57.040)

instead of being multi line brackets became one line

Brendan Eich (38:59.760)

or essentially one line.

Lex Fridman (39:00.600)

So you wrote, so JavaScript was written,

Brendan Eich (39:02.600)

the programming language was written as a comment.

Lex Fridman (39:05.080)

A comment for old browsers and a set of brackets

Brendan Eich (39:08.120)

that were ignored with real code for new.

Lex Fridman (39:10.160)

And it was this two way comment hiding hack, as I called it,

Brendan Eich (39:13.480)

that was absolutely necessary for us to get off the ground.

Lex Fridman (39:15.700)

We couldn't have bootstrapped JavaScript without it.

Brendan Eich (39:17.680)

We didn't have scripts that were loaded from a separate file.

Lex Fridman (39:20.240)

The only scripts in Netscape 2 were inline in the document.

Lex Fridman (39:22.900)

What were the challenges here?

Lex Fridman (39:24.280)

What, like what, you know, typing,

Lex Fridman (39:27.440)

what were the choices you were thinking about?

Lex Fridman (39:30.120)

What was the design for this garbage collection?

Brendan Eich (39:31.560)

I didn't have time to write a garbage collector.

Lex Fridman (39:32.960)

So I just, I didn't at first.

Brendan Eich (39:34.680)

The thing was using essentially arenas

Lex Fridman (39:36.920)

or what GNU calls object pools

Lex Fridman (39:39.160)

and just would run out of memory eventually.

Lex Fridman (39:41.680)

And I added reference counting in a hurry

Brendan Eich (39:43.560)

after the 10 days in which I hacked.

Lex Fridman (39:45.340)

So after I was in the server team doing HTTP 1.1

Lex Fridman (39:48.120)

and thinking about the language,

Lex Fridman (39:49.640)

I finally got transferred to the client team in early May.

Lex Fridman (39:52.300)

And that's when I, you know, I got the go sign from Mark

Lex Fridman (39:55.260)

and it was like, we can't wait

Brendan Eich (39:57.280)

because people inside Netscape are doubting.

Lex Fridman (39:59.360)

Even people inside Sun are definitely doubting.

Brendan Eich (40:01.880)

Bill Joy was the champion, but he was like alone in that

Lex Fridman (40:04.960)

in seeing that there was a role for JavaScript

Brendan Eich (40:07.260)

as the, as I call it, the sidekick language,

Lex Fridman (40:10.560)

robbing the boy hostage.

Brendan Eich (40:12.720)

Frank Miller put it in the Dark Knight Returns

Lex Fridman (40:16.360)

that there was this silly little language

Brendan Eich (40:18.640)

that would be the glue language

Lex Fridman (40:19.520)

and it could become important over time.

Lex Fridman (40:20.960)

And you were better off having that complementarity,

Lex Fridman (40:23.880)

that pairing of languages,

Brendan Eich (40:25.120)

just like Microsoft stacked it with visual C plus plus

Lex Fridman (40:28.080)

and visual pacing.

Lex Fridman (40:28.900)

So what was the big moment of I'm done?

Lex Fridman (40:33.120)

So I had to do a demo.

Brendan Eich (40:34.320)

I forget the dates.

Lex Fridman (40:35.800)

I think I, for a history of programming languages paper

Brendan Eich (40:38.680)

that Alan Wiersbrock did with my help,

Lex Fridman (40:40.840)

he did a lot of the writing.

Brendan Eich (40:42.760)

I think it was the 10 days from like Thursday evening

Lex Fridman (40:47.440)

through to the following weeks, you know,

Brendan Eich (40:50.800)

the whole of that week and then into the Monday.

Lex Fridman (40:52.520)

You get sleep?

Brendan Eich (40:53.360)

Not enough.

Lex Fridman (40:54.400)

And I was really going fast

Brendan Eich (40:56.280)

because I had already used a lot of C compiler

Lex Fridman (41:00.080)

and front end compiler knowledge

Brendan Eich (41:02.320)

that I'd gained from undergraduate school.

Lex Fridman (41:04.700)

When I started getting into computing

Brendan Eich (41:07.560)

as a renegade physics major,

Lex Fridman (41:10.040)

people were formalizing more efficient bottom up grammars,

Brendan Eich (41:14.640)

parsers for bottom up languages,

Lex Fridman (41:17.040)

really LALR one was the big thing.

Lex Fridman (41:19.520)

And I studied all this and learned how to parse them.

Lex Fridman (41:22.440)

And in the end, if you're doing C languages,

Brendan Eich (41:25.100)

you often do what Dennis Ritchie did anyway,

Lex Fridman (41:28.640)

which is the recursive descent parser.

Brendan Eich (41:31.280)

You can hand code it.

Lex Fridman (41:32.200)

And I did that for JavaScript in a blazing hurry,

Brendan Eich (41:36.280)

mostly got it right.

Lex Fridman (41:37.280)

I didn't have precedence inversion problems or other bugs,

Lex Fridman (41:40.160)

but I copied a lot from Java and C.

Lex Fridman (41:43.160)

And I tried to keep things simple,

Brendan Eich (41:45.240)

like the equality operator in those 10 days sprint

Lex Fridman (41:49.400)

between two objects of different dynamic type said,

Brendan Eich (41:52.360)

no, they're not equal, their types are different.

Lex Fridman (41:54.320)

And then after that, I had internal early adopters

Lex Fridman (41:57.680)

and they were using JavaScript to match a number

Lex Fridman (42:01.240)

against a database field that had been stringized.

Lex Fridman (42:03.920)

And they said, oh, can we just have implicit conversion?

Lex Fridman (42:06.440)

Like an idiot, I agreed.

Brendan Eich (42:07.720)

I gave them what they wanted.

Lex Fridman (42:08.640)

I was trying to please them and get adoption.

Lex Fridman (42:10.680)

And that broke what equivalence relation

Lex Fridman (42:16.560)

nature there was to the double equal.

Brendan Eich (42:19.600)

There's some edge cases with not a number

Lex Fridman (42:22.360)

that break that too, but it really broke it.

Brendan Eich (42:25.040)

Having an implicit conversions in the operator

Lex Fridman (42:26.960)

is something that people still roast me over.

Lex Fridman (42:29.360)

So let's talk about two things.

Lex Fridman (42:31.720)

One, it sounds like the comparison operator,

Brendan Eich (42:34.280)

the equality operator is the thing that you regret.

Lex Fridman (42:38.760)

So maybe you can.

Brendan Eich (42:39.600)

Making it sloppy.

Lex Fridman (42:40.600)

Making it sloppy.

Lex Fridman (42:41.440)

So what is the biggest thing you regret in those 10 days?

Lex Fridman (42:45.200)

And what is the biggest thing you're proud of?

Lex Fridman (42:47.480)

So that making it sloppy came after the 10 days.

Lex Fridman (42:49.680)

And my lesson there, which I've tweeted is

Brendan Eich (42:51.880)

when people come to you saying,

Lex Fridman (42:53.160)

can you please make it sloppy or add this cute feature?

Brendan Eich (42:56.080)

The answer should be no.

Lex Fridman (42:57.640)

And I should have known that

Brendan Eich (42:58.480)

because I think Nicholas Viert, one of my heroes said,

Lex Fridman (43:01.200)

the essence of design is leaving things out.

Lex Fridman (43:04.560)

But during the 10 days, I also, like I said,

Lex Fridman (43:06.200)

I was in such a hurry, I left out garbage collection,

Brendan Eich (43:08.560)

came back to haunt me, but I got reference counting in

Lex Fridman (43:10.760)

in time that people weren't running out of memory

Brendan Eich (43:13.960)

right away on long lived JavaScript.

Lex Fridman (43:16.120)

What happens when you don't have garbage collection

Lex Fridman (43:17.760)

and you have objects?

Lex Fridman (43:18.600)

Well, you just run out of memory.

Lex Fridman (43:20.240)

And you know, at first you write a short script

Lex Fridman (43:24.120)

and the page doesn't last long or it doesn't do a lot.

Brendan Eich (43:25.920)

It's okay.

Lex Fridman (43:26.760)

Oh, I see, yeah, yeah.

Lex Fridman (43:27.600)

But if you're writing a game or something

Lex Fridman (43:28.600)

and you're doing event based allocation,

Brendan Eich (43:30.600)

you run out of memory.

Lex Fridman (43:31.680)

And this was noticed in the summer of 1985

Lex Fridman (43:34.320)

and people were like, what's going on?

Lex Fridman (43:35.720)

Oh yeah, I better go back and do reference counting.

Lex Fridman (43:38.600)

And then the problem with reference counting is

Lex Fridman (43:40.200)

you're writing the language in the runtime in C,

Brendan Eich (43:43.160)

an unsafe language, and if you're reference counting

Lex Fridman (43:45.760)

and you overflow the counter, you mismanage it

Lex Fridman (43:48.080)

so it goes high, it gets stuck high,

Lex Fridman (43:49.640)

you leak memory again and you run out.

Brendan Eich (43:51.560)

If you underflow it, you pre memory that's still in use.

Lex Fridman (43:56.200)

And even then we knew what all the security hackers

Brendan Eich (44:00.120)

came to know that you therefore have potentially

Lex Fridman (44:03.000)

a remote code execution vulnerability.

Brendan Eich (44:05.840)

Cause this was before things like non executable

Lex Fridman (44:09.800)

heap memory and staff defenses against taking over memory.

Lex Fridman (44:14.800)

So if you can, from the remote side,

Lex Fridman (44:17.780)

write some HTML and JavaScript that just happens

Brendan Eich (44:19.920)

to exploit a bug in memory safety,

Lex Fridman (44:21.820)

like it causes JavaScript to underflow reference counter.

Lex Fridman (44:24.960)

And the script still has its hands on that object

Lex Fridman (44:27.560)

and it's trying to call a method on it

Lex Fridman (44:29.040)

and there's some kind of lookup function table

Lex Fridman (44:30.680)

in the object, but you've managed to stuff the heap

Brendan Eich (44:34.600)

with strings that forwards their own lookalike

Lex Fridman (44:37.160)

for the function table.

Brendan Eich (44:39.720)

You can call some other code.

Lex Fridman (44:41.360)

And this was a problem right away.

Lex Fridman (44:43.400)

So security, JavaScript upped the ante.

Lex Fridman (44:46.280)

Java had this problem too, but in its own VM.

Lex Fridman (44:48.840)

And it just was a separate headache for Sun to worry about.

Lex Fridman (44:53.440)

We had this problem in Netscape right away.

Lex Fridman (44:55.280)

So Netscape 2 came out after my 10 days

Lex Fridman (44:58.360)

and after these follow on work to embed JavaScript

Brendan Eich (45:02.640)

better in the browser and to add garbage

Lex Fridman (45:04.480)

or collection through reference counting,

Brendan Eich (45:05.960)

really I call it reference counting and get it shipped.

Lex Fridman (45:09.640)

We had a bunch of dot releases where we fixed

Brendan Eich (45:11.400)

security bugs like maniacs.

Lex Fridman (45:13.560)

But what is the thing you're, you know,

Brendan Eich (45:16.200)

when you sit back on a porch and just look out

Lex Fridman (45:19.120)

into the sunset, what are you most proud of

Lex Fridman (45:21.000)

from those 10 days?

Lex Fridman (45:21.960)

I think the first class functions shines.

Brendan Eich (45:23.560)

I think especially since Java didn't have it

Lex Fridman (45:25.600)

and it was somewhat unusual.

Brendan Eich (45:27.600)

Scheme made it in somehow at the end of the day.

Lex Fridman (45:30.640)

In spirit, I mean, people complain because Scheme has,

Brendan Eich (45:33.440)

you know, minimalism.

Lex Fridman (45:34.840)

It has, you know, six or seven special forms.

Brendan Eich (45:37.000)

It has hygienic macros.

Lex Fridman (45:38.720)

It has call CC.

Brendan Eich (45:39.640)

It has sort of a beautiful complete set of forms

Lex Fridman (45:44.880)

to make the land of calculus pleasant to use in practice.

Lex Fridman (45:49.440)

And JavaScript is, you know, kind of a multi paradigm

Lex Fridman (45:54.120)

or shambolic.

Brendan Eich (45:55.200)

Just in a small tangent, you mentioned Mark Andreessen.

Lex Fridman (45:59.400)

It sounds like, and Bill Joy, but staying on Mark,

Brendan Eich (46:03.400)

it sounds like he had an impact on you

Lex Fridman (46:06.520)

in that he sort of believed in what you were doing there.

Lex Fridman (46:09.120)

Can you talk about like what role Mark had in your life?

Lex Fridman (46:11.920)

Yeah, we would meet at the Peninsula Creamery

Brendan Eich (46:15.880)

in downtown Palo Alto.

Lex Fridman (46:17.720)

And Mark was just fresh out of, you know, grad school

Brendan Eich (46:19.960)

or whatever he was doing and he was big dude

Lex Fridman (46:22.720)

and he got fitter later.

Brendan Eich (46:25.280)

He had hair, he would order giant milkshakes and burgers

Lex Fridman (46:28.800)

and we would meet there and brainstorm about what to do.

Lex Fridman (46:31.040)

And it was very direct because we didn't have much time.

Lex Fridman (46:34.280)

The sort of, we didn't talk about it.

Brendan Eich (46:36.400)

The implication was Microsoft was coming after us.

Lex Fridman (46:39.080)

Mark was saying things boldly pre IPO

Lex Fridman (46:41.760)

like Netscape plus Java kills Windows, right?

Lex Fridman (46:44.320)

This is, make a browser programmable.

Brendan Eich (46:46.840)

It becomes the new runtime for programs.

Lex Fridman (46:49.440)

It's the meta OS or it's the replacement OS.

Lex Fridman (46:52.000)

But he still saw value in JavaScript.

Lex Fridman (46:54.280)

Yes, even though he was saying that

Lex Fridman (46:56.000)

and Java was the big name, hence the trademark license,

Lex Fridman (46:59.800)

he saw JavaScript as important.

Lex Fridman (47:01.360)

And he even thought, what if we got,

Lex Fridman (47:03.720)

I told this in other interviews, I can say it.

Brendan Eich (47:05.040)

He thought, what if we had my friend Kip Hickman

Lex Fridman (47:07.680)

who'd been at Netscape from the beginning

Lex Fridman (47:09.560)

and who was a kernel hacker at SGI when I joined,

Lex Fridman (47:13.080)

he'd started writing his own JVM

Brendan Eich (47:14.920)

before we consummated the Sun deal

Lex Fridman (47:16.440)

and got our hands on their code.

Lex Fridman (47:17.960)

And the Java compiler, Java C,

Lex Fridman (47:20.480)

which Arthur Van Hoff had written very nice code,

Brendan Eich (47:23.200)

was all written in Java.

Lex Fridman (47:24.160)

It was self hosted or so called bootstrap.

Lex Fridman (47:26.480)

And so we could use that as soon as Kip's Java VM

Lex Fridman (47:29.320)

could run the bytecode from the Sun JVM

Brendan Eich (47:34.400)

running the self hosted compiler to emit the bytecode.

Lex Fridman (47:37.520)

So once we could bootstrap into Kip's VM,

Brendan Eich (47:39.400)

we wouldn't need Sun.

Lex Fridman (47:40.320)

And Mark was like, well, maybe we can just ditch Sun.

Brendan Eich (47:44.400)

Well, if Kip's Java VM, or if you're a JavaScript VM,

Lex Fridman (47:47.360)

now we need graphics.

Lex Fridman (47:48.240)

So Mark was thinking far ahead

Lex Fridman (47:49.600)

because he knew you could do things with HTML and images,

Lex Fridman (47:53.380)

but at some point you really want.

Lex Fridman (47:54.920)

Like dynamic graphics or three dimensional?

Brendan Eich (47:57.600)

Even SGI had already started its downfall

Lex Fridman (47:59.920)

because the first floor VLSI team there

Brendan Eich (48:02.960)

had gone off to do 3D effects

Lex Fridman (48:04.440)

and all these other companies

Lex Fridman (48:05.360)

that made the graphics card on your PC, right?

Lex Fridman (48:07.520)

Doom was big and Quake.

Lex Fridman (48:09.360)

And so you were, we were all playing Quake.

Lex Fridman (48:10.680)

I was old, so I was terrible.

Lex Fridman (48:12.880)

But why not put that graphics capability on the web?

Lex Fridman (48:16.400)

And in fact, it finally happened at Mozilla

Brendan Eich (48:18.720)

with Firefox era with Vlad Vukicovic

Lex Fridman (48:21.520)

taking OpenGL ES and reflecting it as WebGL.

Lex Fridman (48:24.420)

But OpenGL ES is the mobile version of OpenGL,

Lex Fridman (48:28.000)

which is a standard based on SGI GL.

Lex Fridman (48:30.320)

So this whole lineage of graphics libraries

Lex Fridman (48:32.800)

or really graphics languages for what became the GPU.

Lex Fridman (48:36.280)

And Mark was thinking ahead.

Lex Fridman (48:37.320)

He's like, we need graphics too.

Lex Fridman (48:38.560)

And I thought, okay, I can try to get somebody I knew at SGI,

Lex Fridman (48:40.960)

but he's a grad student at MIT.

Brendan Eich (48:42.880)

He was studying under Barbara Liskov.

Lex Fridman (48:44.440)

He laughed when he heard about this later, Andrew Myers.

Brendan Eich (48:46.780)

He's at Cornell, long time, I think he's a full professor.

Lex Fridman (48:50.040)

And Mark said, great, we'll get him.

Brendan Eich (48:52.200)

I'm not sure he's gonna come.

Lex Fridman (48:53.200)

We'll throw money, we'll stock options.

Brendan Eich (48:55.320)

We never did it.

Lex Fridman (48:56.160)

And they did the Sun deal.

Lex Fridman (48:57.640)

So Kip Nobly put aside his own JVM

Lex Fridman (49:00.960)

and we used the Sun JVM.

Lex Fridman (49:02.880)

So that was an ambitious period.

Lex Fridman (49:04.000)

And Mark was very generative because he was pushing hard.

Brendan Eich (49:06.640)

He was ambitious and he wanted to have Netscape

Lex Fridman (49:11.200)

possibly be in control of the ball.

Brendan Eich (49:13.680)

Maybe you can speak to this dance

Lex Fridman (49:16.840)

of Netscape versus Internet Explorer.

Brendan Eich (49:20.440)

You've thrown some loving words towards Microsoft

Lex Fridman (49:25.600)

throughout this conversation,

Lex Fridman (49:26.720)

but that's a theme with, I mean, Steve Jobs

Lex Fridman (49:29.080)

has a similar sort of commentary.

Brendan Eich (49:31.280)

From a big sort of philosophical principle perspective,

Lex Fridman (49:34.520)

can you comment on like the approach

Brendan Eich (49:36.600)

that Microsoft has taken with Internet Explorer

Lex Fridman (49:39.280)

from IE1 to Edge today?

Brendan Eich (49:42.120)

Is there something that you see as valuable

Lex Fridman (49:45.400)

that they're doing in the occasional copying

Lex Fridman (49:48.420)

and that kind of stuff?

Lex Fridman (49:49.600)

Or is it, is the world worse off

Lex Fridman (49:53.360)

because Internet Explorer exists?

Lex Fridman (49:55.240)

So I'm gonna segment this into historical areas

Brendan Eich (49:57.640)

because I think Microsoft today with Satya

Lex Fridman (49:59.560)

is quite a different company

Lex Fridman (50:00.640)

and what they're doing with Edge is different.

Lex Fridman (50:02.360)

But back then, Gates, aggressive character,

Brendan Eich (50:06.160)

not really original in my view, not an originator.

Lex Fridman (50:09.680)

Steve Jobs famously said once,

Brendan Eich (50:11.280)

he doesn't have any taste

Lex Fridman (50:12.440)

and I don't mean this in a small way, he has no taste.

Brendan Eich (50:15.080)

You can see this, Apple at the time had beautiful typography

Lex Fridman (50:19.360)

and ligatures and kerning and the fonts looked great.

Lex Fridman (50:22.600)

And Windows had this sort of ugly system font

Lex Fridman (50:25.240)

that was carefully aligned with Pixel

Lex Fridman (50:26.920)

so it didn't get anything.

Lex Fridman (50:28.640)

What is it?

Brendan Eich (50:29.480)

I'm sorry to keep interrupting,

Lex Fridman (50:30.300)

but why was Internet Explorer winning

Lex Fridman (50:32.360)

throughout the history of these competitions?

Lex Fridman (50:34.320)

Distribution.

Brendan Eich (50:35.240)

Distribution matters more than anything.

Lex Fridman (50:36.600)

And this is why even now we're seeing

Brendan Eich (50:39.840)

in the browser wars Edge doing better

Lex Fridman (50:41.640)

because it's being foisted on people of Windows.

Brendan Eich (50:43.440)

We have Windows 10 boxes at home.

Lex Fridman (50:45.200)

We have some Windows 7 boxes or laptops we keep running to

Brendan Eich (50:48.240)

because we don't connect them to the internet generally.

Lex Fridman (50:50.840)

But once you have that operating system to hold,

Brendan Eich (50:55.840)

you can force Edge.

Lex Fridman (50:57.800)

And Apple did it with Safari too.

Brendan Eich (50:59.480)

It's not unique to Microsoft.

Lex Fridman (51:01.340)

That's sad.

Lex Fridman (51:02.180)

But distribution matters.

Lex Fridman (51:03.680)

And that's why I think IE was going to win.

Brendan Eich (51:07.080)

That's why everybody at Netscape felt we're doomed.

Lex Fridman (51:09.560)

This was something Michael Toy and Jamie Woodson were doomed.

Lex Fridman (51:12.900)

But for a while there we had a chance

Lex Fridman (51:14.720)

and we innovated in Netscape too.

Brendan Eich (51:16.040)

We did a big platform push, Java and JavaScript

Lex Fridman (51:18.860)

and plugins, more plugins and more HTML table features.

Lex Fridman (51:23.860)

And really started making a programmable stack

Lex Fridman (51:28.700)

out of what were pretty static web languages.

Lex Fridman (51:30.820)

And even in the beta releases of Netscape,

Lex Fridman (51:33.220)

two people were using JavaScript

Brendan Eich (51:34.460)

to build what you would call single page applications

Lex Fridman (51:36.620)

like Gmail.

Lex Fridman (51:37.620)

And they were using JavaScript locally to compute things

Lex Fridman (51:40.380)

and to call the server on a hidden frame in the background.

Lex Fridman (51:43.660)

So it was prefiguring a lot of what came later as AJAX

Lex Fridman (51:46.860)

or dynamic JavaScript, dynamic HTML.

Lex Fridman (51:48.580)

So people saw that, I mean.

Lex Fridman (51:49.740)

Even then they saw it, yeah.

Brendan Eich (51:50.920)

That's kind of, I don't know.

Lex Fridman (51:52.900)

But from my perspective, that seems quite brilliant.

Brendan Eich (51:55.180)

It seems like really innovative

Lex Fridman (51:56.580)

that you would have code run in the browser.

Brendan Eich (51:59.220)

It did impress me with something

Lex Fridman (52:01.380)

which I learned later about from Eric Von Hippel of MIT,

Brendan Eich (52:04.020)

which is user innovation networks, lead user effects.

Lex Fridman (52:07.260)

That throwing out JavaScript,

Brendan Eich (52:09.200)

even though we weren't doing open source,

Lex Fridman (52:10.460)

we were doing beta releases early

Lex Fridman (52:12.320)

and permissively with Netscape.

Lex Fridman (52:14.360)

Getting early developer feedback, absolutely critical.

Brendan Eich (52:17.180)

I loved it.

Lex Fridman (52:18.020)

I did some of that with SGI

Brendan Eich (52:19.140)

with some of the products I worked on,

Lex Fridman (52:20.260)

but it really came to the fore in Netscape.

Lex Fridman (52:21.860)

And that culminated in Mozilla

Lex Fridman (52:23.780)

where you're dealing with developers all the time

Lex Fridman (52:25.740)

and early adopters, lead users.

Lex Fridman (52:27.240)

But the lead users helped improve JavaScript,

Brendan Eich (52:29.840)

even in those last few betas

Lex Fridman (52:31.580)

where I could hardly change things.

Brendan Eich (52:32.660)

I was under pretty rigid change control.

Lex Fridman (52:34.580)

So we're talking about just a small collection

Brendan Eich (52:36.260)

of individuals that are just like upfront.

Lex Fridman (52:38.340)

A guy named Bill Dorch.

Brendan Eich (52:39.420)

You can find his work in the web archives,

Lex Fridman (52:40.940)

still from 1996.

Brendan Eich (52:41.960)

It's a single page application.

Lex Fridman (52:43.300)

It's an artist gallery of mountain art.

Lex Fridman (52:46.340)

He used JavaScript?

Lex Fridman (52:47.180)

It doesn't quite work.

Brendan Eich (52:48.000)

He uses JavaScript locally.

Lex Fridman (52:48.840)

He uses a local database.

Lex Fridman (52:50.580)

What you would think of now is JSON,

Lex Fridman (52:52.180)

but it's all pure JavaScript code,

Brendan Eich (52:54.260)

a bunch of objects being constructed.

Lex Fridman (52:55.820)

That's so cool.

Lex Fridman (52:57.460)

So how is, if you can do sort of a big sweeping progress

Lex Fridman (53:02.620)

of JavaScript, how has JavaScript changed over the years?

Brendan Eich (53:05.380)

Any of you from those early 10 days

Lex Fridman (53:07.300)

with a quick addition of garbage collection

Lex Fridman (53:09.460)

and fixes around security,

Lex Fridman (53:11.140)

how has this evolution that now it's taken over the world?

Brendan Eich (53:14.500)

In this, it's been a bumpy ride

Lex Fridman (53:16.340)

because the standards body got shut down after Microsoft,

Brendan Eich (53:19.900)

I think, took over the web and then felt punished

Lex Fridman (53:22.580)

by the USB Microsoft antitrust case.

Lex Fridman (53:25.360)

Can you speak to the standard body?

Lex Fridman (53:27.380)

That was a fun ride too

Brendan Eich (53:28.260)

because Netscape had taken the lead

Lex Fridman (53:30.940)

with the web and HTML innovations

Brendan Eich (53:34.340)

like frames and framesets tables.

Lex Fridman (53:36.900)

And the W3C was sort of off even then,

Brendan Eich (53:39.220)

sort of in SGML land heading toward XML, la la land.

Lex Fridman (53:43.100)

I'm gonna be a little harsh on it.

Lex Fridman (53:44.340)

What's SGML?

Lex Fridman (53:45.180)

I'm sorry.

Brendan Eich (53:46.000)

SGML was the precursor markup language to HTML.

Lex Fridman (53:48.900)

It was sort of the more extensible standards,

Brendan Eich (53:52.740)

generalized markup language.

Lex Fridman (53:54.140)

It was a...

Brendan Eich (53:54.980)

XML like...

Lex Fridman (53:55.820)

Pointy brackets, but it had all sorts of elaborate syntax

Brendan Eich (53:58.540)

for doing different semantics.

Lex Fridman (54:00.940)

And this is why I think TBL and others

Brendan Eich (54:03.780)

who wanted to do the semantic web then took XML forward,

Lex Fridman (54:08.320)

but they had this, or some of them anyway,

Brendan Eich (54:10.400)

had this strange idea they could replace the web with XML

Lex Fridman (54:12.980)

or that they would upgrade the web to be XML.

Lex Fridman (54:15.720)

And it couldn't be done.

Lex Fridman (54:17.240)

Worse is better had concrete meaning.

Brendan Eich (54:19.500)

The web was very forgiving of HTML,

Lex Fridman (54:22.660)

including sort of minor syntax errors

Brendan Eich (54:25.180)

that could be error corrected.

Lex Fridman (54:26.500)

Like error correction isn't generally done

Brendan Eich (54:28.140)

in programming languages because...

Lex Fridman (54:29.980)

That's another amazing thing about HTML is like,

Brendan Eich (54:32.540)

it's more like biology than programming.

Lex Fridman (54:35.140)

Exactly.

Lex Fridman (54:35.980)

And so XML was in its standard form super strict

Lex Fridman (54:40.700)

and could never have admitted the kind of users

Brendan Eich (54:43.220)

who were committing these errors.

Lex Fridman (54:44.420)

And the funniest part was Microsoft said,

Brendan Eich (54:46.500)

hey, we're doing XML,

Lex Fridman (54:47.620)

but the way they put it in Internet Explorer

Brendan Eich (54:49.540)

under the default media type,

Lex Fridman (54:51.500)

put it through the HTML error corrector.

Brendan Eich (54:53.940)

Oh, wow.

Lex Fridman (54:54.940)

So they kind of bastardized it to make it popular

Lex Fridman (54:57.020)

and usable and accessible.

Lex Fridman (54:58.580)

And so XML as a pure thing was never gonna take over.

Lex Fridman (55:04.220)

And what W3C was kind of not fully functional

Lex Fridman (55:07.140)

because Netscape wasn't cooperating with them.

Brendan Eich (55:09.700)

We thought about where to take JavaScript

Lex Fridman (55:11.100)

and we realized our standards,

Brendan Eich (55:12.660)

Guru Kargal realized there was a European standards body

Lex Fridman (55:17.140)

that had already given Microsoft fits

Brendan Eich (55:18.800)

by standardizing parts of the Windows 3.1 API,

Lex Fridman (55:22.580)

which European governments insisted on.

Brendan Eich (55:24.020)

They said, Microsoft, we can't use your operating system

Lex Fridman (55:26.260)

without some standards.

Lex Fridman (55:27.580)

And Microsoft said, here's our docs.

Lex Fridman (55:29.660)

And the government said, no, we need a European standard.

Lex Fridman (55:32.460)

So this body called

Lex Fridman (55:33.420)

the European Computer Manufacturers Association, ECMA,

Brendan Eich (55:37.500)

which eventually became global

Lex Fridman (55:38.620)

and became a proper noun instead of an acronym.

Brendan Eich (55:40.980)

Right, it's just one capital E now with a lowercase CMA.

Lex Fridman (55:44.660)

Right, and as one of the early Microsoft guys I met

Brendan Eich (55:47.580)

when we first convened a working group

Lex Fridman (55:50.260)

to talk about JavaScript said,

Brendan Eich (55:51.420)

it sounds like a skin disease.

Lex Fridman (55:53.420)

But it gave, I mean, maybe you'll speak to that,

Lex Fridman (55:56.140)

but it gave the name to JavaScript of ECMA script.

Lex Fridman (55:58.980)

That was the standard name

Brendan Eich (55:59.820)

because Java was a trademark of Suns.

Lex Fridman (56:02.820)

They were so aggressive,

Brendan Eich (56:03.720)

they were sending cease and desist letters

Lex Fridman (56:05.840)

to people whose middle European heritage

Brendan Eich (56:08.500)

meant their surname was Javanko

Lex Fridman (56:10.820)

and they called their website javanko.com

Lex Fridman (56:12.900)

and Sunwood sent them a letter saying,

Lex Fridman (56:14.680)

you're using J A V A at the start of your domain name,

Brendan Eich (56:16.820)

you must cease and desist.

Lex Fridman (56:18.940)

I love marketing more than anything else in this world.

Lex Fridman (56:22.020)

So ECMA script and now is popularly named as ES plus version.

Lex Fridman (56:27.420)

I would say people use JS more than anything.

Brendan Eich (56:29.900)

People still say JavaScript.

Lex Fridman (56:31.140)

JavaScript is in all the books.

Lex Fridman (56:32.460)

So I mean, when you're referring to it,

Lex Fridman (56:33.740)

it's usually JavaScript.

Lex Fridman (56:34.900)

And when you wanna refer to a version of JavaScript,

Lex Fridman (56:37.340)

you'll say ES6, ES5.

Brendan Eich (56:39.180)

Yes, or now they've gone to years,

Lex Fridman (56:40.860)

which is kind of confusing

Brendan Eich (56:41.840)

because it's an offset of 2009, ES6, ES 2016.

Lex Fridman (56:50.740)

Yeah, it doesn't match the years perfectly.

Brendan Eich (56:52.700)

Yeah, so what were the choices made

Lex Fridman (56:55.680)

and how did JavaScript evolve here?

Lex Fridman (56:58.020)

So we took this new standards body,

Lex Fridman (57:01.100)

which we thought sort of a proven record

Brendan Eich (57:04.340)

of standing up to Microsoft,

Lex Fridman (57:05.420)

but Microsoft sent a lot of people.

Brendan Eich (57:07.540)

They sent some people who were pretty good.

Lex Fridman (57:10.540)

And then when they realized that I was there

Lex Fridman (57:12.940)

and Netscape was not gonna just bend over

Lex Fridman (57:15.220)

and do whatever they wanted,

Brendan Eich (57:16.380)

they sent somebody really good.

Lex Fridman (57:17.220)

And he was a smart guy.

Brendan Eich (57:18.060)

He did a lot of the work on the first draft of the spec.

Lex Fridman (57:21.000)

Sean Katzenberger, he's left Microsoft.

Brendan Eich (57:22.780)

He even did what I sort of did.

Lex Fridman (57:25.280)

He told his bosses, stop bugging me to do other things.

Brendan Eich (57:28.000)

I'm focused on this.

Lex Fridman (57:28.960)

Cause it took a lot of focused work

Brendan Eich (57:30.580)

to create the first draft of the spec.

Lex Fridman (57:31.940)

And I was still holding,

Brendan Eich (57:33.560)

I was spending almost all the plates.

Lex Fridman (57:34.900)

I had like part time help in certain areas.

Lex Fridman (57:37.220)

And on the front end integrations,

Lex Fridman (57:38.740)

I had the front end guys.

Lex Fridman (57:40.080)

But I couldn't take as much time as Sean was

Lex Fridman (57:42.620)

to write the draft spec,

Lex Fridman (57:43.680)

but I had to participate

Lex Fridman (57:44.660)

because I was essentially helping write down

Lex Fridman (57:46.880)

what the language did

Lex Fridman (57:48.020)

and in areas where we didn't like what it did

Lex Fridman (57:50.180)

and Microsoft didn't agree,

Lex Fridman (57:51.620)

we sometimes got away with slight changes.

Lex Fridman (57:54.580)

And that's the story of standards.

Lex Fridman (57:55.800)

You have different implementations

Lex Fridman (57:58.020)

and depending on their market power,

Lex Fridman (57:59.840)

they interoperate where you have agreement

Lex Fridman (58:01.840)

and where they don't,

Lex Fridman (58:02.780)

the dominant one usually sets the de facto standard.

Lex Fridman (58:05.900)

And then you should probably reflect that

Lex Fridman (58:07.500)

into the de jure standard.

Lex Fridman (58:09.340)

And this happened with JavaScript.

Lex Fridman (58:11.060)

Over time, as Netscape went down and Microsoft went up,

Brendan Eich (58:14.120)

we did the first edition of the standard codified in 1997

Lex Fridman (58:18.180)

in France, we had a trip to Nice, which was very memorable.

Lex Fridman (58:22.140)

For any interesting reason or just because it's Nice?

Lex Fridman (58:24.540)

And ECMA's European and IBM and others were there.

Brendan Eich (58:28.980)

Mike Kalashaw, an IBM fellow was a British.

Lex Fridman (58:31.340)

And the guy who ran ECMA at the time, Jan van der Bell,

Brendan Eich (58:35.260)

was quite a raconteur and a very fun guy.

Lex Fridman (58:39.060)

And he had us out for, you know, the great,

Brendan Eich (58:41.220)

you know, Fui de mer, the bouillabaisse and the...

Lex Fridman (58:43.480)

Was the standardization process beautiful or painful

Lex Fridman (58:46.140)

that those early days, you as a designer of the language?

Lex Fridman (58:48.460)

It was painful because it was rushed.

Brendan Eich (58:50.140)

Now, Guy Steele was contributed by Sun.

Lex Fridman (58:52.400)

So even more than Sean,

Brendan Eich (58:53.620)

you had this giant brain Guy Steele helping,

Lex Fridman (58:55.960)

bringing some of that scheme magic.

Lex Fridman (58:57.220)

And he even brought Richard Gabriel for funding.

Lex Fridman (58:59.460)

Richard wrote the fourth clause of the ECMA standard,

Brendan Eich (59:02.900)

which was kind of an intro to what JavaScript's all about.

Lex Fridman (59:05.300)

So we had some really good people

Lex Fridman (59:07.260)

and we didn't fight too much.

Lex Fridman (59:09.140)

There was some tension where I was fixing bugs

Lex Fridman (59:11.480)

and I was late to a meeting and Sean Katzenberger,

Lex Fridman (59:13.860)

Microsoft, was actually mad.

Lex Fridman (59:15.020)

Like, where is he?

Lex Fridman (59:15.840)

We need him.

Lex Fridman (59:16.680)

And when I got there,

Lex Fridman (59:18.660)

I saw that only he saw this sort of off by one bug

Brendan Eich (59:21.940)

in somewhere in the spec.

Lex Fridman (59:22.960)

And then I saw it too.

Lex Fridman (59:24.020)

And I said, there's a fence post bug there.

Lex Fridman (59:25.300)

And then we kind of locked eyes

Lex Fridman (59:27.020)

and we realized we were on the same page.

Lex Fridman (59:28.480)

And we kind of, he wasn't mad anymore.

Lex Fridman (59:30.300)

What were the features that are being struggled over

Lex Fridman (59:33.020)

and debated and thought about?

Brendan Eich (59:35.000)

It was mainly writing down what worked

Lex Fridman (59:36.480)

and what we thought should work in the edge cases

Brendan Eich (59:38.420)

that didn't interoperate or that seemed wrong.

Lex Fridman (59:41.540)

But we were already laying the groundwork

Brendan Eich (59:43.080)

for the future additions that I was already implementing.

Lex Fridman (59:46.220)

I was still trying to lead the standard

Brendan Eich (59:48.120)

by using the dominant market power

Lex Fridman (59:50.100)

to write the code that actually shipped.

Lex Fridman (59:52.500)

So the de facto standard would lead the de jure standard.

Lex Fridman (59:54.740)

And I was putting in the missing function forms

Brendan Eich (59:58.220)

that I didn't have time for in the 10 days.