Richard Craib: WallStreetBets, Numerai, and the Future of Stock Trading

Lex Fridman (01:24.120)

are able to coordinate and collaborate

Richard Craib (01:26.520)

in taking on the elite centralized power structures,

Lex Fridman (01:32.760)

especially when those elites are misbehaving.

Richard Craib (01:36.240)

I believe that power in as many cases as possible

Lex Fridman (01:39.360)

should be distributed.

Lex Fridman (01:40.800)

And in this case, the internet, as it is for many cases,

Lex Fridman (01:44.860)

is the fundamental enabler of that power.

Lex Fridman (01:48.640)

And at the core, what the internet

Lex Fridman (01:50.740)

in its distributed nature represents is freedom.

Richard Craib (01:53.960)

Of course, the thing about freedom

Lex Fridman (01:55.960)

is it enables chaos or progress, or sometimes both.

Lex Fridman (02:02.480)

And that's kind of the point of the thing.

Lex Fridman (02:04.400)

Freedom is empowering, but ultimately unpredictable.

Lex Fridman (02:09.160)

And I think in the end, freedom wins.

Lex Fridman (02:12.520)

If you enjoy this podcast, subscribe on YouTube,

Richard Craib (02:15.560)

review it on Apple Podcasts, follow on Spotify,

Lex Fridman (02:18.680)

support on Patreon, or connect with me on Twitter

Richard Craib (02:21.800)

at Lex Friedman.

Lex Fridman (02:23.340)

And now, here's my conversation with Richard Crabe.

Richard Craib (02:28.080)

From your perspective, can you summarize

Lex Fridman (02:29.880)

the important events around this amazing saga

Richard Craib (02:32.720)

that we've been living through of Wall Street Bets,

Lex Fridman (02:34.960)

the subreddit and GameStop, and in general,

Richard Craib (02:38.160)

just what are your thoughts about it

Lex Fridman (02:39.720)

from a technical to the philosophical level?

Richard Craib (02:42.120)

I think it's amazing.

Lex Fridman (02:42.960)

It's like my favorite story ever.

Richard Craib (02:46.500)

Like when I was reading about it,

Lex Fridman (02:47.760)

I was like, this is the best.

Lex Fridman (02:49.380)

And it's also connected with my company,

Lex Fridman (02:53.780)

which we can talk about.

Lex Fridman (02:54.740)

But what I liked about it is like,

Lex Fridman (02:57.020)

I like decentralized coordination

Lex Fridman (02:59.420)

and looking at the mechanisms

Lex Fridman (03:01.500)

that these are Wall Street Bets users use

Richard Craib (03:04.860)

to hype each other up, to get excited,

Lex Fridman (03:08.180)

to prove that they bought the stock and they're holding.

Lex Fridman (03:12.100)

And then also to see that how big of an impact

Lex Fridman (03:15.940)

that that decentralized coordination had.

Lex Fridman (03:18.000)

So it really was a big deal.

Lex Fridman (03:21.240)

Were you impressed by the distributed coordination,

Richard Craib (03:24.420)

the collaboration amongst like,

Lex Fridman (03:26.400)

I don't know what the numbers are.

Richard Craib (03:27.480)

I know I'm numerized looking at the data.

Lex Fridman (03:30.480)

After all of this is over and done,

Richard Craib (03:32.640)

it'd be interesting to see like

Lex Fridman (03:34.440)

from a large scale distributed system perspective

Richard Craib (03:38.200)

to see how everything played out.

Lex Fridman (03:40.340)

But just from your current perspective, what we know,

Richard Craib (03:44.160)

is it obvious to you that such incredible level

Lex Fridman (03:47.640)

of coordination could happen

Richard Craib (03:49.920)

where a lot of people come together in a distributed sense,

Lex Fridman (03:52.780)

there's an emergent behavior that happens after that.

Richard Craib (03:54.960)

No, it's not at all obvious.

Lex Fridman (03:57.400)

And one of the reasons is the lack of credibility.

Richard Craib (04:01.760)

To coordinate with someone,

Lex Fridman (04:02.740)

you need to make credible contracts or credible claims.

Lex Fridman (04:06.600)

So if you have a username on our Wall Street Bets,

Lex Fridman (04:11.620)

like some of them are, like deep fucking value

Richard Craib (04:14.700)

is one of them.

Lex Fridman (04:15.540)

That's an actual username.

Richard Craib (04:16.580)

By the way, we're talking about,

Lex Fridman (04:18.240)

there's a website called Reddit

Lex Fridman (04:19.600)

and there's subreddits on it.

Lex Fridman (04:21.120)

And a lot of people, mostly anonymous,

Richard Craib (04:24.120)

I think for the most part anonymous,

Lex Fridman (04:27.080)

can create user accounts

Lex Fridman (04:28.260)

and then can then just talk on forum like style boards.

Lex Fridman (04:31.240)

You should know what Reddit is.

Richard Craib (04:32.340)

If you don't know what Reddit is, check it out.

Lex Fridman (04:34.920)

If you don't know what Reddit is,

Richard Craib (04:36.040)

maybe go to the awesome subreddit first,

Lex Fridman (04:40.560)

aww with cute pictures of cats and dogs.

Richard Craib (04:43.120)

That's my recommendation.

Lex Fridman (04:44.080)

Anyway.

Richard Craib (04:44.920)

Okay, yeah, that would be a good start to Reddit.

Lex Fridman (04:46.140)

When you get into it more, go to our Wall Street Bets.

Richard Craib (04:48.960)

It gets dark quickly.

Lex Fridman (04:50.960)

We'll probably talk about that too.

Lex Fridman (04:53.060)

So yeah, so there's these users

Lex Fridman (04:56.240)

and there's no contracts, like you were saying.

Richard Craib (04:58.200)

There's no contracts, the users are anonymous,

Lex Fridman (05:01.480)

but there are little things that do help.

Lex Fridman (05:03.560)

So for example,

Lex Fridman (05:04.480)

if you've posted a really good investment idea in the past,

Richard Craib (05:07.720)

that exists on Reddit as well.

Lex Fridman (05:10.040)

And it might have lots of upvotes.

Lex Fridman (05:12.640)

And that's also kind of like giving credibility

Lex Fridman (05:14.840)

to your next thing.

Lex Fridman (05:17.880)

And then they are also putting up screenshots,

Lex Fridman (05:20.820)

like here's the trades I've made and here's a screenshot.

Richard Craib (05:26.160)

Now you could fake the screenshot,

Lex Fridman (05:28.160)

but still it seems like if you've got a lot of karma

Lex Fridman (05:32.200)

and you've had a good performance on the community,

Lex Fridman (05:35.280)

it somehow becomes credible enough

Lex Fridman (05:37.680)

for other people to be like, you know what?

Lex Fridman (05:39.760)

He actually probably did put a million dollars into this.

Lex Fridman (05:43.240)

And you know what, I can follow that trade easily.

Lex Fridman (05:46.360)

And there's a bunch of people like that.

Lex Fridman (05:47.680)

So you're kind of integrating all that information

Lex Fridman (05:50.260)

together yourself to see like, huh,

Richard Craib (05:52.600)

there's something happening here.

Lex Fridman (05:53.520)

And then you jump onto this little boat of like behavior,

Richard Craib (05:57.440)

like we should buy the stock or sell the stock.

Lex Fridman (06:00.400)

And then another person jumps on,

Richard Craib (06:02.440)

another person jumps on.

Lex Fridman (06:04.520)

And all of a sudden you have just a huge number of people

Richard Craib (06:07.280)

behaving in the same direction.

Lex Fridman (06:08.640)

It's like flock of whatever birds.

Richard Craib (06:10.640)

Exactly.

Lex Fridman (06:11.480)

What was strange with this one,

Richard Craib (06:12.480)

it wasn't just let's all buy Tesla.

Lex Fridman (06:15.320)

We love Elon, we love Tesla, let's all buy Tesla.

Lex Fridman (06:18.800)

Because that we've heard before, right?

Lex Fridman (06:21.040)

Everybody likes Tesla.

Richard Craib (06:23.960)

Well, now they do.

Lex Fridman (06:26.200)

So what they did with this in this case,

Richard Craib (06:29.560)

they're buying a stock that was bad.

Lex Fridman (06:31.360)

They're buying it because it was bad.

Lex Fridman (06:33.440)

And that's really weird because that's a little bit

Lex Fridman (06:37.240)

too galaxy brain for a decentralized community.

Lex Fridman (06:41.340)

How did they come up with it?

Lex Fridman (06:43.220)

How did they know that was the right one?

Lex Fridman (06:44.820)

And the reason they liked it

Lex Fridman (06:46.580)

is because it had really, really high short interest.

Richard Craib (06:49.860)

It had been shorted more than its own float, I believe.

Lex Fridman (06:54.780)

And so they figured out that if they all bought

Richard Craib (06:57.500)

this bad stock, they could short squeeze some hedge funds.

Lex Fridman (07:03.260)

And those hedge funds would have to capitulate

Lex Fridman (07:05.220)

and buy the stock at really, really high prices.

Lex Fridman (07:08.380)

And we should say that shorted means that

Richard Craib (07:10.540)

these are a bunch of people, when you short a stock,

Lex Fridman (07:13.180)

you're betting on the, you're predicting

Richard Craib (07:16.140)

that the stock's going to go down

Lex Fridman (07:17.500)

and then you will make money if it does.

Lex Fridman (07:19.580)

And then what's a short squeeze?

Lex Fridman (07:22.340)

It's really that if you are a hedge fund

Lex Fridman (07:24.580)

and you take a big short position in a company,

Lex Fridman (07:28.920)

there's a certain level at which

Richard Craib (07:31.100)

you can't sustain holding that position.

Lex Fridman (07:34.780)

There's no limit to how high a stock can go,

Lex Fridman (07:37.320)

but there is a limit to how low it can go, right?

Lex Fridman (07:39.700)

So if you short something, you have infinite loss potential.

Lex Fridman (07:43.180)

And if the stock doubles overnight, like GameStop did,

Lex Fridman (07:48.580)

you're putting a lot of stress on that hedge fund.

Lex Fridman (07:51.900)

And that hedge fund manager might have to say,

Lex Fridman (07:53.780)

you know what, I have to get out of the trade.

Lex Fridman (07:56.340)

And the only way to get out is to buy the bad stock

Lex Fridman (07:59.580)

that they don't want, like they believe will go down.

Lex Fridman (08:02.580)

So it's an interesting situation,

Lex Fridman (08:05.380)

particularly because it's not zero sum.

Richard Craib (08:09.140)

If you say, let's all get together

Lex Fridman (08:11.460)

and make a bubble in watermelons,

Richard Craib (08:13.620)

you buy a bunch of watermelons,

Lex Fridman (08:14.860)

the price goes up, it comes down again,

Richard Craib (08:18.260)

it's a zero sum game.

Lex Fridman (08:20.860)

If someone's already shorted a stock

Lex Fridman (08:22.460)

and you can make them short squeeze,

Lex Fridman (08:24.040)

it's actually a positive sum game.

Lex Fridman (08:25.900)

So yes, some Redditors will make a lot of money,

Lex Fridman (08:28.160)

some will lose a lot,

Lex Fridman (08:29.740)

but actually the whole group will make money.

Lex Fridman (08:32.500)

And that's really why it was such a clever thing

Richard Craib (08:36.980)

for them to do.

Lex Fridman (08:38.140)

And coupled to the fact that shorting,

Richard Craib (08:40.660)

I mean, maybe you can push back,

Lex Fridman (08:42.740)

but to me always from an outsider's perspective,

Richard Craib (08:45.840)

seemed, I hope I'm not using too strong of a word,

Lex Fridman (08:48.860)

but it seemed almost unethical.

Richard Craib (08:51.180)

Maybe not unethical, maybe it's just a asshole thing to do.

Lex Fridman (08:55.620)

Okay, I'm speaking not from an economics

Richard Craib (08:57.760)

or financial perspective,

Lex Fridman (08:58.780)

I'm speaking from just somebody who loves,

Richard Craib (09:02.360)

I'm a fan of a lot of people,

Lex Fridman (09:03.540)

I love celebrating the success of a lot of people.

Lex Fridman (09:07.000)

And this is like the stock market equivalent of like haters.

Lex Fridman (09:13.060)

I know that's not what it is.

Richard Craib (09:14.420)

I know that there's efficient,

Lex Fridman (09:15.740)

you wanna have an economy efficient mechanism

Richard Craib (09:18.240)

for punishing sort of overhyped, overvalued things.

Lex Fridman (09:23.860)

That's what shorthand guess is designed for.

Lex Fridman (09:26.100)

But it just always felt like these people are just,

Lex Fridman (09:29.540)

because they're not just betting on the loss of the company.

Richard Craib (09:33.500)

It feels like they're also using their leverage and power

Lex Fridman (09:37.340)

to manipulate media or just to write articles

Richard Craib (09:40.780)

or just to hate on you on social media.

Lex Fridman (09:43.440)

Then you get to see that with Elon Musk and so on.

Lex Fridman (09:46.500)

So this is like the man,

Lex Fridman (09:50.560)

so people like hedge funds that were shorting

Richard Craib (09:53.780)

are like the sort of embodiment of the evil

Lex Fridman (09:58.740)

or just the bad guy, the overpowerful

Richard Craib (10:01.320)

that's misusing their power.

Lex Fridman (10:02.860)

And here's the crowd,

Richard Craib (10:04.140)

the people that are standing up and rising up.

Lex Fridman (10:06.380)

So it's not just that they were able to collaborate

Richard Craib (10:10.500)

on Wall Street bets to sort of effectively

Lex Fridman (10:13.340)

make money for themselves.

Richard Craib (10:14.820)

It's also that this is like a symbol

Lex Fridman (10:17.380)

of the people getting together

Lex Fridman (10:19.420)

and fighting the centralized elites, the powerful.

Lex Fridman (10:23.380)

And that, I don't know what your thoughts are

Richard Craib (10:27.140)

about that in general.

Lex Fridman (10:28.580)

At this stage, it feels like that's really exciting

Richard Craib (10:32.140)

that people have power,

Lex Fridman (10:35.340)

just like regular people have power.

Richard Craib (10:38.200)

At the same time, it's scary a little bit

Lex Fridman (10:40.780)

because just studying history,

Richard Craib (10:44.320)

people could be manipulated by charismatic leaders.

Lex Fridman (10:49.340)

And so just like Elon right now is manipulating,

Richard Craib (10:54.220)

encouraging people to buy Dogecoin or whatever,

Lex Fridman (10:58.820)

there can be good charismatic leaders

Lex Fridman (11:00.780)

and there can be bad charismatic leaders.

Lex Fridman (11:02.540)

And so it's nerve wracking.

Richard Craib (11:04.580)

It's a little bit scary how much power

Lex Fridman (11:06.860)

a subreddit can have to destroy somebody

Richard Craib (11:11.220)

because right now we're celebrating

Lex Fridman (11:12.980)

they might be attacking or destroying somebody

Richard Craib (11:15.020)

that everybody doesn't like,

Lex Fridman (11:17.140)

but what if they attack somebody

Lex Fridman (11:19.140)

that is actually good for this world?

Lex Fridman (11:21.060)

So that, and that's kind of the awesomeness

Lex Fridman (11:25.940)

and the price of freedom.

Lex Fridman (11:28.540)

It's like it could destroy the world

Richard Craib (11:31.200)

or it can save the world.

Lex Fridman (11:32.640)

But at this stage, it feels like, I don't know,

Richard Craib (11:35.220)

overall, when you sit back,

Lex Fridman (11:36.700)

do you think this was just a positive wave

Lex Fridman (11:39.620)

of emergent behavior?

Lex Fridman (11:41.060)

Is there something negative about what happened?

Richard Craib (11:43.280)

Well, yeah, the cool thing is they weren't doing anything,

Lex Fridman (11:47.020)

the Reddit people weren't doing anything exotic.

Richard Craib (11:50.720)

It was a creative trade, but it wasn't exotic.

Lex Fridman (11:55.140)

It wasn't, it was just buying the stock.

Richard Craib (11:57.700)

Okay, maybe they bought some options too,

Lex Fridman (12:00.300)

but it was the hedge fund that was doing the exotic thing.

Lex Fridman (12:04.780)

So I liked that.

Lex Fridman (12:06.380)

It was, it's hard to say, well, we've got together

Lex Fridman (12:10.580)

and we've pulled all our money together

Lex Fridman (12:12.980)

and now there's a company out there that's worth more.

Lex Fridman (12:16.480)

What's wrong with that?

Lex Fridman (12:17.620)

Yeah. Right?

Lex Fridman (12:18.580)

But it doesn't talk about the motivations, which is,

Lex Fridman (12:21.820)

and then we destroyed some hedge funds in the process.

Richard Craib (12:25.020)

Is there something to be said about the humor

Lex Fridman (12:28.180)

and the, I don't know, the edginess,

Lex Fridman (12:31.780)

sometimes viciousness of that subreddit?

Lex Fridman (12:34.220)

I haven't looked at it too much,

Lex Fridman (12:36.420)

but it feels like people can be quite aggressive on there.

Lex Fridman (12:40.900)

So is there, what is that?

Lex Fridman (12:45.180)

Is that what Freedom looks like?

Lex Fridman (12:49.020)

I think it does, yeah.

Richard Craib (12:50.080)

You definitely need to let people,

Lex Fridman (12:52.380)

one of the things that people have compared it to

Richard Craib (12:54.740)

is the Occupy Wall Street, which is, let's say,

Lex Fridman (12:58.180)

some very sincere liberals, like 23 years old, whatever,

Lex Fridman (13:03.540)

and they go out with signs

Lex Fridman (13:04.980)

and they have some kind of case to make.

Lex Fridman (13:08.540)

But this isn't sincere, really.

Lex Fridman (13:11.740)

It's like a little bit more nihilistic,

Richard Craib (13:14.700)

a little bit more YOLO, and therefore a little bit

Lex Fridman (13:18.020)

more scary because who's scared of the Occupy Wall Street

Lex Fridman (13:22.500)

people with the signs?

Lex Fridman (13:23.660)

Nobody.

Lex Fridman (13:24.500)

But these hedge funds really are scared.

Lex Fridman (13:26.020)

I was scared of the Wall Street bats people.

Richard Craib (13:29.820)

I'm still scared of them.

Lex Fridman (13:31.900)

Yeah, the anonymity is a bit terrifying and exciting.

Richard Craib (13:36.900)

Yeah.

Lex Fridman (13:37.860)

I mean, yeah, I don't know what to do with this.

Richard Craib (13:40.500)

I've been following events in Russia, for example.

Lex Fridman (13:43.500)

It's like there's a struggle between centralized power

Lex Fridman (13:46.140)

and the distributed.

Lex Fridman (13:47.460)

I mean, that's the struggle of the history

Lex Fridman (13:50.260)

of human civilization, right?

Lex Fridman (13:51.740)

But this on the internet, just that you can multiply people.

Richard Craib (13:57.700)

Like some of them don't have to be real.

Lex Fridman (13:59.700)

Like you can probably create bots.

Richard Craib (14:01.460)

Like it starts getting me, me as a programmer,

Lex Fridman (14:05.500)

I start to think like, hmm, me as one person,

Lex Fridman (14:08.780)

how much chaos can I create by writing some bots?

Lex Fridman (14:12.260)

Yeah.

Lex Fridman (14:13.380)

And I'm sure I'm not the only one thinking that.

Lex Fridman (14:17.700)

There's, I'm sure there's hundreds, thousands

Richard Craib (14:19.820)

of good developers out there listening to this,

Lex Fridman (14:22.420)

thinking the same thing.

Lex Fridman (14:23.860)

And then as that develops further and further

Lex Fridman (14:26.540)

in the next like decade or two,

Lex Fridman (14:28.300)

what impact does that have on financial markets,

Lex Fridman (14:30.580)

on just destruction of reputations of just,

Richard Craib (14:37.140)

or politics, the bickering of left and right

Lex Fridman (14:41.500)

political discourse, the dynamics of that

Richard Craib (14:44.020)

being manipulated by, you know,

Lex Fridman (14:46.860)

people talk about like Russian bots or whatever.

Lex Fridman (14:49.580)

We're probably in a very early stage of that, right?

Lex Fridman (14:52.660)

Yeah, exactly.

Lex Fridman (14:53.980)

And this is a good example.

Lex Fridman (14:56.580)

So do you have a sense that most of WallStreetBets folks

Lex Fridman (15:00.540)

are actually individual people, right?

Lex Fridman (15:02.780)

That's the feeling I have is they're just individual,

Richard Craib (15:05.340)

maybe young investors, just doing a little bit

Lex Fridman (15:07.900)

of an investment, but just on a large scale.

Richard Craib (15:10.140)

Yeah, exactly.

Lex Fridman (15:11.180)

The reason I found out, I've known about WallStreetBets

Richard Craib (15:13.780)

for a while, but the reason I found out about GameStop

Lex Fridman (15:15.500)

was this, just I met somebody at a party

Richard Craib (15:17.860)

who told me about it and he was like 21 years old

Lex Fridman (15:20.300)

and he's like, it's gonna go up 100% in the next one day.

Lex Fridman (15:23.860)

That we're talking about in last year?

Lex Fridman (15:26.020)

This was probably, no, this was, yeah, a few days ago.

Richard Craib (15:29.220)

Yeah, it was like maybe two weeks ago or something.

Lex Fridman (15:34.260)

So it was already high GameStop,

Lex Fridman (15:37.420)

but it was just strange to me that there was someone

Lex Fridman (15:39.300)

telling me at a party how to trade stocks

Richard Craib (15:42.740)

who was like 21 years old.

Lex Fridman (15:44.340)

And I started to, yeah, I started to look into it.

Lex Fridman (15:49.100)

And yeah, and he did make, he made it, yeah,

Lex Fridman (15:52.100)

he made 140% in one day, he was right.

Lex Fridman (15:55.500)

And now he's supercharged, he's a little bit wealthier

Lex Fridman (15:58.980)

and now he's gonna wait for the next thing.

Lex Fridman (16:01.180)

And this decentralized entity

Lex Fridman (16:03.100)

is just gonna get bigger and bigger.

Lex Fridman (16:04.580)

And they're gonna together search for the next thing.

Lex Fridman (16:06.580)

So there's thousands of folks like him

Lex Fridman (16:08.820)

and they're going to probably search

Lex Fridman (16:10.020)

for the next thing to attack.

Richard Craib (16:11.580)

People that have power in this world that sit there

Lex Fridman (16:14.460)

with power right now in government and in finance

Lex Fridman (16:18.180)

and any kind of position

Lex Fridman (16:20.260)

are probably a little bit scared right now.

Lex Fridman (16:22.940)

And honestly, that's probably a little bit good.

Lex Fridman (16:26.260)

It's dangerous, but it's good.

Richard Craib (16:28.100)

Yeah, it certainly makes you think twice about shorting.

Lex Fridman (16:31.420)

It certainly makes you think twice

Richard Craib (16:32.740)

about putting a lot of money into a short.

Lex Fridman (16:35.020)

Like these funds put a lot into one or two names.

Lex Fridman (16:38.620)

And so it was very, very badly risk managed.

Lex Fridman (16:41.700)

Do you think shorting is, can you speak at a high level

Lex Fridman (16:46.020)

just for your own as a person, is it good for the world?

Lex Fridman (16:50.060)

Is it good for markets?

Richard Craib (16:52.140)

I do think that the two kinds of shorting,

Lex Fridman (16:55.180)

evil shorting and chill shorting.

Richard Craib (17:00.860)

Evil shorting is what Melvin Capital was doing.

Lex Fridman (17:04.500)

And it's like, you put a huge position down,

Richard Craib (17:08.780)

you get all your buddies to also short it

Lex Fridman (17:10.780)

and you start making press

Lex Fridman (17:12.420)

and trying to bring this company down.

Lex Fridman (17:16.660)

And I don't think in some cases,

Richard Craib (17:20.420)

you go out after like fraudulent companies say,

Lex Fridman (17:22.380)

this company is a fraud.

Richard Craib (17:24.100)

Maybe that's okay.

Lex Fridman (17:24.940)

Like some, but they weren't even saying,

Richard Craib (17:27.020)

they're just saying it's a bad company

Lex Fridman (17:29.140)

and we're going to bring it to the ground,

Richard Craib (17:30.940)

bring it to its knees.

Lex Fridman (17:32.020)

So a quant fund like Numerai,

Richard Craib (17:37.100)

we always have lots of positions

Lex Fridman (17:39.100)

and we never have a position

Richard Craib (17:40.260)

that's like more than 1% of our fund.

Lex Fridman (17:43.420)

So we actually have right now, 250 shorts.

Richard Craib (17:48.260)

I don't know any of them except for one

Lex Fridman (17:52.060)

because it was one of the meme stocks.

Lex Fridman (17:53.940)

But we shorting them not to make them go,

Lex Fridman (17:58.940)

we don't even want them to go down necessarily.

Richard Craib (18:01.460)

That doesn't sound a bit strange that I say that,

Lex Fridman (18:03.180)

but we just want them to not go up as much as our longs.

Lex Fridman (18:09.420)

So by shorting a little bit,

Lex Fridman (18:13.420)

we can actually go long more

Richard Craib (18:15.060)

in the things we do believe in.

Lex Fridman (18:16.780)

So when we were going long in Tesla,

Richard Craib (18:19.500)

we could do it with more money than we had

Lex Fridman (18:22.020)

because we would borrow from banks

Richard Craib (18:24.860)

who would lend us money to go down.

Lex Fridman (18:27.380)

Who would lend us money because we had longs and shorts,

Richard Craib (18:32.220)

because we didn't have market exposure,

Lex Fridman (18:34.660)

we didn't have market risk.

Lex Fridman (18:35.780)

And so I think that's a good thing

Lex Fridman (18:37.380)

because that means we can short the oil companies

Lex Fridman (18:41.060)

and go long Tesla and make the future come forward faster.

Lex Fridman (18:44.180)

And I do think that's not a bad thing.

Lex Fridman (18:47.180)

So we've talked about this incredible distributed system

Lex Fridman (18:51.060)

created by Wall Street Bets.

Lex Fridman (18:52.580)

And then there's a platform which is Robinhood,

Lex Fridman (18:56.940)

which allows investors to efficiently as far,

Richard Craib (18:59.340)

you can correct me if I'm wrong,

Lex Fridman (19:00.500)

but there's those and there's others in this new MRI

Richard Craib (19:04.220)

that allow you to make it accessible for people to invest.

Lex Fridman (19:09.580)

But that said, Robinhood was in a centralized way

Richard Craib (19:15.700)

applied its power to restrict trading

Lex Fridman (19:17.740)

on the stocks that we're referring to.

Lex Fridman (19:19.900)

Do you have a thoughts on actually

Lex Fridman (19:21.860)

like the things that happened?

Richard Craib (19:23.940)

I don't know how much you were paying attention

Lex Fridman (19:26.580)

to sort of the shadiness around the whole thing.

Lex Fridman (19:30.420)

Do you think it was forced to do it?

Lex Fridman (19:32.540)

Or was there something shady going on?

Lex Fridman (19:34.420)

What are your thoughts in general?

Lex Fridman (19:36.060)

Well, I think I wanna see the alternate history.

Richard Craib (19:39.860)

Like I wanna see the counterfactual history

Lex Fridman (19:43.020)

of them not doing that.

Lex Fridman (19:44.740)

How bad would it have gotten for hedge funds?

Lex Fridman (19:47.980)

How much more damage could have been done

Lex Fridman (19:50.220)

if the momentum of these short squeezes could continue?

Lex Fridman (19:53.400)

What happens when there are short squeezes,

Richard Craib (19:57.240)

even if they're in a few stocks,

Lex Fridman (1:00:00.840)

and then they get trampled over.

Richard Craib (1:00:03.920)

That's the terrifying thing, actually.

Lex Fridman (1:00:07.320)

A lot of people have written about this,

Richard Craib (1:00:08.960)

is somehow that little voice that's human morality

Lex Fridman (1:00:13.880)

gets silenced when we get into groups and start chanting.

Lex Fridman (1:00:17.920)

And that's terrifying.

Lex Fridman (1:00:18.760)

But I think maybe I misunderstood.

Richard Craib (1:00:22.320)

I thought that you're saying AI systems can be dangerous,

Lex Fridman (1:00:25.920)

but you just describe how humans can be dangerous.

Lex Fridman (1:00:28.440)

So which is safer?

Lex Fridman (1:00:30.520)

So one thing is, so Wall Street bets these kinds of attacks.

Richard Craib (1:00:38.640)

It's not possible to model, numerize data,

Lex Fridman (1:00:42.080)

and then come up with the idea from the model,

Richard Craib (1:00:45.000)

let's short squeak, just game stop.

Lex Fridman (1:00:46.920)

It's not even framed in that way.

Richard Craib (1:00:49.440)

It's not possible to have that idea.

Lex Fridman (1:00:52.800)

But it is possible for a bunch of humans.

Lex Fridman (1:00:54.960)

So I think this, it's, numera could get very powerful

Lex Fridman (1:00:59.680)

without it being dangerous.

Lex Fridman (1:01:01.840)

But Wall Street bets needs to get a little bit more powerful,

Lex Fridman (1:01:05.480)

and it'll be pretty dangerous.

Richard Craib (1:01:08.240)

Yeah, well, I mean, this is a good place

Lex Fridman (1:01:11.960)

to think about numera data today and numera signals

Lex Fridman (1:01:17.040)

and what that looks like in 10, 20, 30, 50, 100 years.

Lex Fridman (1:01:22.680)

Right now, I guess, maybe you can correct me,

Lex Fridman (1:01:24.520)

but the data that we're working with is like a window.

Lex Fridman (1:01:28.040)

It's an anonymized, obfuscated window

Richard Craib (1:01:32.320)

into a particular aspect, time period of the market.

Lex Fridman (1:01:36.640)

And you can expand that more and more and more and more,

Richard Craib (1:01:40.440)

potentially.

Lex Fridman (1:01:41.160)

You can imagine in different dimensions

Richard Craib (1:01:43.560)

to where it encapsulates all the things that,

Lex Fridman (1:01:47.120)

where you could include kind of human to human communication

Richard Craib (1:01:52.080)

that was available to buy GameStop, for example,

Lex Fridman (1:01:55.960)

on Wall Street bets.

Lex Fridman (1:01:57.640)

So maybe as a step back, can you speak

Lex Fridman (1:02:00.000)

to what is numera signals and what are the different data

Lex Fridman (1:02:05.480)

sets that are involved?

Lex Fridman (1:02:07.600)

So with numera signals, you're still

Richard Craib (1:02:10.800)

providing predictions to us, but you can do it

Lex Fridman (1:02:15.800)

from your own data sets.

Lex Fridman (1:02:18.080)

So numera, it's all you have to model our data

Lex Fridman (1:02:21.040)

to come up with predictions.

Richard Craib (1:02:22.760)

Numa signals is whatever data you can find out there,

Lex Fridman (1:02:26.560)

you can turn it into a signal and give it to us.

Lex Fridman (1:02:29.880)

So it's a way for us to import signals

Lex Fridman (1:02:32.600)

on data we don't yet have.

Lex Fridman (1:02:36.800)

And that's why it's particularly valuable,

Lex Fridman (1:02:38.720)

because it's going to be signals.

Richard Craib (1:02:42.240)

You're only rewarded for signals that are

Lex Fridman (1:02:44.200)

orthogonal to our core signal.

Lex Fridman (1:02:47.600)

So you have to be doing something uncorrelated.

Lex Fridman (1:02:50.040)

And so strange alternative data tends to have that property.

Richard Craib (1:02:55.000)

There isn't too many other signals

Lex Fridman (1:02:57.280)

that are correlated with what's happening on Wall Street

Richard Craib (1:03:02.320)

bets.

Lex Fridman (1:03:02.920)

That's not going to be correlated with the price

Richard Craib (1:03:05.440)

to earnings ratio.

Lex Fridman (1:03:07.760)

And we have some users, as of recently, as of a week ago,

Richard Craib (1:03:11.600)

there was a user that created, I think he's in India,

Lex Fridman (1:03:14.600)

he created a signal that is scraped from Wall Street bets.

Lex Fridman (1:03:21.400)

And now we have that signal as one

Lex Fridman (1:03:24.800)

of our signals in thousands that we use at Numerai.

Lex Fridman (1:03:28.800)

And the structure of the signal is similar,

Lex Fridman (1:03:30.800)

so it's just numbers and time series data.

Richard Craib (1:03:33.200)

It's exactly.

Lex Fridman (1:03:34.120)

And it's just like, you're providing a ranking of stocks.

Lex Fridman (1:03:37.640)

So you just say, a one means you like the stock,

Lex Fridman (1:03:41.440)

zero means you don't like the stock,

Lex Fridman (1:03:43.320)

and you provide that for 5,000 stocks in the world.

Lex Fridman (1:03:46.640)

And they somehow converted the natural language

Richard Craib (1:03:49.600)

that's in the Wall Street bet.

Lex Fridman (1:03:51.880)

Exactly.

Lex Fridman (1:03:52.760)

And they open sourced this Colab notebook.

Lex Fridman (1:03:55.480)

You can go and see it.

Lex Fridman (1:03:57.320)

And so, yeah, it's making a sentiment score.

Lex Fridman (1:04:00.280)

And then turning it into a rank of stocks.

Richard Craib (1:04:02.200)

A sentiment score.

Lex Fridman (1:04:03.160)

Yeah.

Richard Craib (1:04:04.120)

Like, this stock sucks, or this stock is awesome.

Lex Fridman (1:04:07.160)

And then converting.

Richard Craib (1:04:08.000)

That's fascinating.

Lex Fridman (1:04:08.840)

Just even looking at that data would be fascinating.

Lex Fridman (1:04:11.520)

So on the signal side, what's the vision?

Lex Fridman (1:04:15.680)

This long term, what do you see that becoming?

Lex Fridman (1:04:18.240)

So we want to manage all the money in the world.

Lex Fridman (1:04:21.720)

That's Numerai's mission.

Lex Fridman (1:04:23.760)

And to get that, we need to have all the data

Lex Fridman (1:04:27.400)

and have all of the talent.

Richard Craib (1:04:30.200)

Like, there's no way, first principles,

Lex Fridman (1:04:32.400)

if you had really good modeling and really good data

Lex Fridman (1:04:35.480)

that you would lose, right?

Lex Fridman (1:04:37.800)

It's just a question of how much do you need to get really good.

Lex Fridman (1:04:41.440)

So Numerai already has some really nice data

Lex Fridman (1:04:43.960)

that we give out.

Richard Craib (1:04:45.160)

This year, we are 10xing that.

Lex Fridman (1:04:48.360)

And I actually think we'll 10x the amount of data

Richard Craib (1:04:50.600)

we have on Numerai every year for at least the next 10 years.

Lex Fridman (1:04:55.200)

Wow.

Lex Fridman (1:04:55.840)

So it's going to get very big, the data we give out.

Lex Fridman (1:04:59.080)

And signals is more data.

Richard Craib (1:05:03.640)

People with any other random data set

Lex Fridman (1:05:06.400)

can turn that into a signal and give it to us.

Lex Fridman (1:05:09.080)

And in some sense, that kind of data

Lex Fridman (1:05:10.560)

is the edge cases, the weirdness is the,

Lex Fridman (1:05:12.840)

so you're focused on the bulk, the main data.

Lex Fridman (1:05:16.360)

And then there's just weirdness from all over the place

Richard Craib (1:05:18.640)

that just can enter through this back door of Numerai signals.

Lex Fridman (1:05:22.080)

Exactly.

Lex Fridman (1:05:22.640)

And it's also a little bit shorter term.

Lex Fridman (1:05:26.280)

So the signals are about a seven day time horizon.

Lex Fridman (1:05:31.160)

And on Numerai, it's like a 30 day.

Lex Fridman (1:05:33.920)

So it's often for faster situations.

Richard Craib (1:05:38.560)

You've written about a master plan.

Lex Fridman (1:05:40.320)

And you've mentioned, which I love,

Richard Craib (1:05:43.040)

in a similar sort of style of big style thinking,

Lex Fridman (1:05:46.520)

you would like Numerai to manage all of the world's money.

Lex Fridman (1:05:52.920)

So how do we get there from yesterday

Lex Fridman (1:05:56.400)

to several years from now?

Lex Fridman (1:05:59.480)

Like what is the plan?

Lex Fridman (1:06:02.040)

So you've already started to allure to get all the data

Lex Fridman (1:06:06.280)

and get all the talent, humans, models.

Lex Fridman (1:06:11.760)

Exactly.

Richard Craib (1:06:12.680)

I mean, the important thing to note there is,

Lex Fridman (1:06:14.800)

what would that mean?

Lex Fridman (1:06:16.080)

And I think the biggest thing it means

Lex Fridman (1:06:18.000)

is if there was one hedge fund, you

Richard Craib (1:06:22.400)

would have not so much talent wasted

Lex Fridman (1:06:26.200)

on all the other hedge funds.

Richard Craib (1:06:27.760)

Like it's super weird how the industry works.

Lex Fridman (1:06:30.400)

It's like one hedge fund gets a data source and hires a PhD.

Lex Fridman (1:06:34.000)

And another hedge fund has to buy the same data source

Lex Fridman (1:06:36.440)

and hire a PhD.

Lex Fridman (1:06:37.600)

And suddenly, a third of American PhDs

Lex Fridman (1:06:40.000)

are working at hedge funds.

Lex Fridman (1:06:41.280)

And we're not even on Mars.

Lex Fridman (1:06:43.680)

And so in some ways, Numerai, it's

Richard Craib (1:06:46.280)

all about freeing up people who work at hedge funds

Lex Fridman (1:06:49.760)

to go work for Elon.

Richard Craib (1:06:52.480)

Yeah.

Lex Fridman (1:06:53.000)

And also, the people who are working on Numerai problem,

Richard Craib (1:06:58.200)

it feels like a lot of the knowledge

Lex Fridman (1:07:00.240)

there is also transferable to other domains.

Richard Craib (1:07:02.360)

Exactly.

Lex Fridman (1:07:03.920)

One of our top users, he works at NASA Jet Propulsion Lab.

Lex Fridman (1:07:08.080)

And he's amazing.

Lex Fridman (1:07:10.000)

I went to go visit him there.

Lex Fridman (1:07:11.680)

And he's got Numerai posters.

Lex Fridman (1:07:13.800)

And it looks like the movies.

Richard Craib (1:07:16.560)

It looks like Apollo 11 or whatever.

Lex Fridman (1:07:19.320)

Yeah, the point is he didn't quit his job to join full time.

Richard Craib (1:07:26.120)

He's working on getting us to Jupiter's moon.

Lex Fridman (1:07:29.880)

That's his mission, the Europa Klippa mission.

Richard Craib (1:07:31.920)

Actually, literally what you're saying.

Lex Fridman (1:07:33.520)

Literally.

Richard Craib (1:07:34.200)

He's smart enough that we really want his intelligence

Lex Fridman (1:07:37.360)

to reach the stock market.

Richard Craib (1:07:38.480)

Because the stock market's a good thing.

Lex Fridman (1:07:39.600)

Hedge funds are a good thing.

Richard Craib (1:07:40.800)

All kinds of hedge funds, especially.

Lex Fridman (1:07:43.040)

But we don't want him to quit his job.

Lex Fridman (1:07:45.360)

So he can just do Numerai on the weekends.

Lex Fridman (1:07:47.120)

And that's what he does.

Richard Craib (1:07:47.840)

He just made a model and it just automatically submits to us.

Lex Fridman (1:07:50.400)

And he's like one of our best users.

Richard Craib (1:07:53.240)

You mentioned briefly that stock markets are good.

Lex Fridman (1:07:57.400)

From my sort of outsider perspective, is there a sense,

Lex Fridman (1:08:01.920)

do you think trading stocks is closer to gambling?

Lex Fridman (1:08:06.880)

Or is it closer to investing?

Richard Craib (1:08:09.480)

Sometimes it feels like it's gambling as opposed to betting

Lex Fridman (1:08:14.160)

on companies that succeed.

Lex Fridman (1:08:15.560)

And this is maybe connected to our discussion of shorting

Lex Fridman (1:08:17.720)

in general, but from your sense, the way you think about it,

Lex Fridman (1:08:21.000)

is it fundamentally still investing?

Lex Fridman (1:08:23.840)

I do think, I mean, it's a good question.

Richard Craib (1:08:29.720)

I've also seen lately people say, this is like speculation.

Lex Fridman (1:08:33.480)

Is there too much speculation in the market?

Lex Fridman (1:08:35.680)

And it's like, but all the trades are speculative.

Lex Fridman (1:08:38.680)

All the trades have a horizon.

Richard Craib (1:08:40.680)

People want them to work.

Lex Fridman (1:08:44.040)

So I would say that there's certainly

Richard Craib (1:08:48.800)

a lot of aspects of gambling math that applies to investing.

Lex Fridman (1:08:54.920)

Like one thing you don't do in gambling

Richard Craib (1:08:57.360)

is put all your money in one bet.

Lex Fridman (1:09:00.400)

You have bankroll management, and it's a key part of it.

Lex Fridman (1:09:04.280)

And small alterations to your bankroll management

Lex Fridman (1:09:07.600)

might be better than improvements to your skill.

Lex Fridman (1:09:10.680)

And then there are things we care about in our fund.

Lex Fridman (1:09:13.520)

Like we want to make a lot of independent bets.

Richard Craib (1:09:16.560)

We talk about it, like we want to make

Lex Fridman (1:09:18.560)

a lot of independent bets, because that's

Richard Craib (1:09:20.560)

going to be a higher sharp than if you have a lot of bets that

Lex Fridman (1:09:23.400)

depend on each other, like all in one sector.

Lex Fridman (1:09:27.920)

But yeah, I mean, the point is that you

Lex Fridman (1:09:31.320)

want the prices of the stocks to be reflective of their value.

Richard Craib (1:09:39.120)

Of the underlying value of the company.

Lex Fridman (1:09:40.920)

Yeah, you shouldn't have there be like a hedge fund that's

Richard Craib (1:09:45.160)

able to say, well, I've looked at some data,

Lex Fridman (1:09:48.160)

and all of this stuff's super mispriced.

Richard Craib (1:09:51.360)

That's super bad for society if it looks like that to someone.

Lex Fridman (1:09:55.240)

I guess the underlying question then

Richard Craib (1:09:57.600)

is, do you see that the market often drifts away

Lex Fridman (1:10:01.800)

from the underlying value of companies,

Lex Fridman (1:10:04.480)

and it becomes a game in itself?

Lex Fridman (1:10:06.640)

Like with these, whatever they're called,

Richard Craib (1:10:09.920)

like derivatives, like the options, and shorting,

Lex Fridman (1:10:17.720)

and all that kind of stuff.

Richard Craib (1:10:18.880)

It's like layers of game on top of the actual,

Lex Fridman (1:10:22.720)

like what you said, which is like the basic thing

Richard Craib (1:10:25.120)

that the Wall Street Bets was doing,

Lex Fridman (1:10:26.440)

which is like just buying stocks.

Richard Craib (1:10:28.680)

Yeah.

Lex Fridman (1:10:29.720)

There are a lot of games that people

Richard Craib (1:10:31.960)

play that are in the derivatives market.

Lex Fridman (1:10:36.880)

And I think a lot of the stuff people dislike when they look

Richard Craib (1:10:40.240)

at the history of what's happened,

Lex Fridman (1:10:42.480)

they hate like credit default swaps,

Richard Craib (1:10:45.600)

or collateralized debt obligations.

Lex Fridman (1:10:48.040)

Like these are the enemies of 2008.

Lex Fridman (1:10:52.080)

And then the long term capital management thing,

Lex Fridman (1:10:54.760)

it was like they had 30 times leverage, or something.

Richard Craib (1:11:00.000)

You could just go to a gas station

Lex Fridman (1:11:03.280)

and ask anybody at the gas station,

Lex Fridman (1:11:05.720)

is it a good idea to have 30 times leverage?

Lex Fridman (1:11:08.120)

And they just say no.

Richard Craib (1:11:09.520)

It's like common sense just like went out the window.

Lex Fridman (1:11:12.160)

So yeah, I don't respect long term capital management.

Richard Craib (1:11:20.520)

OK.

Lex Fridman (1:11:21.560)

But Numerai doesn't actually use any derivatives

Richard Craib (1:11:24.360)

unless you call shorting derivative.

Lex Fridman (1:11:26.840)

We just we do put money into companies.

Lex Fridman (1:11:29.240)

And that does help the companies we're investing in.

Lex Fridman (1:11:32.240)

It's just in little ways.

Richard Craib (1:11:34.360)

We really did buy Tesla.

Lex Fridman (1:11:36.360)

And it did.

Lex Fridman (1:11:37.520)

And we played some role in its success.

Lex Fridman (1:11:44.080)

Super small, make no mistake.

Lex Fridman (1:11:46.560)

But still, I think that's important.

Lex Fridman (1:11:48.160)

Can I ask you a pothead question,

Lex Fridman (1:11:51.200)

which is what is money, man?

Lex Fridman (1:11:55.720)

So if we just kind of zoom out and look at,

Richard Craib (1:11:59.680)

because I'd love to talk to you about cryptocurrency, which

Lex Fridman (1:12:02.280)

perhaps could be the future of money.

Lex Fridman (1:12:04.400)

In general, how do you think about money?

Lex Fridman (1:12:07.720)

You said Numerai, the vision, the goal

Richard Craib (1:12:10.720)

is to run, to manage the world's money.

Lex Fridman (1:12:15.480)

What is money in your view?

Richard Craib (1:12:19.440)

I don't have a good answer to that.

Lex Fridman (1:12:22.240)

But it's definitely in my personal life,

Richard Craib (1:12:25.200)

it's become more and more warped.

Lex Fridman (1:12:29.200)

And you start to care about the real thing,

Richard Craib (1:12:33.280)

like what's really going on here.

Lex Fridman (1:12:37.040)

Elon talks about things like this,

Lex Fridman (1:12:39.080)

like what is a company, really?

Lex Fridman (1:12:40.880)

It's a bunch of people who show up to work together

Lex Fridman (1:12:43.440)

and they solve a problem.

Lex Fridman (1:12:45.480)

And there might not be a stock out there

Richard Craib (1:12:47.360)

that's trading that represents what they're doing,

Lex Fridman (1:12:49.920)

but it's not the real thing.

Lex Fridman (1:12:52.920)

And being involved in crypto, I put

Lex Fridman (1:12:57.800)

in crowdsale of Ethereum and all these other things

Lex Fridman (1:13:03.440)

and different crypto hedge funds and things

Lex Fridman (1:13:06.240)

that I've invested in.

Lex Fridman (1:13:07.080)

And it's just kind of like, it feels

Lex Fridman (1:13:09.840)

like how I used to think about money stuff

Richard Craib (1:13:13.040)

is just totally warped.

Lex Fridman (1:13:15.960)

Because you stop caring about the price

Lex Fridman (1:13:23.080)

and you care about the product.

Lex Fridman (1:13:26.040)

So by the product, you mean the different mechanisms

Richard Craib (1:13:29.160)

that money is exchanged.

Lex Fridman (1:13:30.440)

I mean, money is ultimately a kind of a little,

Richard Craib (1:13:33.520)

one is a store of wealth, but it's also

Lex Fridman (1:13:36.360)

a mechanism of exchanging wealth.

Lex Fridman (1:13:38.640)

But what wealth means becomes a totally different thing,

Lex Fridman (1:13:42.680)

especially with cryptocurrency, to where it's almost

Richard Craib (1:13:45.960)

like these little contracts, these little agreements,

Lex Fridman (1:13:48.400)

these transactions between human beings

Richard Craib (1:13:50.440)

that represent something that's bigger than just cash being

Lex Fridman (1:13:56.000)

exchanged at 7.11, it feels like.

Richard Craib (1:13:58.080)

Yeah.

Lex Fridman (1:13:58.680)

Maybe I'll answer what is finance.

Richard Craib (1:14:03.280)

It's what are you doing when you have the ability

Lex Fridman (1:14:06.600)

to take out a loan?

Richard Craib (1:14:08.760)

You can bring a whole new future into being with finance.

Lex Fridman (1:14:15.840)

If you couldn't get a student loan to get a college degree,

Richard Craib (1:14:20.240)

you couldn't get a college degree

Lex Fridman (1:14:22.760)

if you didn't have the money.

Lex Fridman (1:14:23.960)

But now, weirdly, you can get it with, and all you have

Lex Fridman (1:14:29.840)

is this loan, which is like, so now you

Richard Craib (1:14:32.360)

can bring a different future into the world.

Lex Fridman (1:14:34.360)

And that's how when I was saying earlier about if you rerun

Richard Craib (1:14:36.960)

American economic history without these things,

Lex Fridman (1:14:40.760)

like you're not allowed to take out loans,

Richard Craib (1:14:42.560)

you're not allowed to have derivatives,

Lex Fridman (1:14:44.880)

you're not allowed to have money,

Richard Craib (1:14:47.680)

it just doesn't really work.

Lex Fridman (1:14:49.000)

And it's a really magic thing how much

Richard Craib (1:14:51.800)

you can do with finance by bringing the future forward.

Lex Fridman (1:14:56.040)

Finance is empowering.

Richard Craib (1:14:58.080)

We sometimes forget this, but it enables innovation.

Lex Fridman (1:15:01.120)

It enables big risk takers and bold builders that ultimately

Richard Craib (1:15:04.720)

make this world better.

Lex Fridman (1:15:06.160)

You said you were early in on cryptocurrency.

Lex Fridman (1:15:09.800)

Can you give your high level overview

Lex Fridman (1:15:12.120)

of just your thoughts about the past, present, and future

Lex Fridman (1:15:15.400)

of cryptocurrency?

Lex Fridman (1:15:17.760)

Yeah, so my friends told me about Bitcoin,

Lex Fridman (1:15:19.880)

and I was interested in equities a lot.

Lex Fridman (1:15:23.800)

And I was like, well, it has no net present value.

Richard Craib (1:15:27.720)

It has no future cash flows.

Lex Fridman (1:15:29.960)

Bitcoin pays no dividends.

Lex Fridman (1:15:33.080)

So I really couldn't get my head around it,

Lex Fridman (1:15:36.120)

that this could be valuable.

Lex Fridman (1:15:39.520)

And then I didn't feel like I was early in cryptocurrency,

Lex Fridman (1:15:44.360)

in fact, because it was like 2014.

Richard Craib (1:15:46.560)

It felt like a long time after Bitcoin.

Lex Fridman (1:15:50.120)

But then I really liked some of the things

Richard Craib (1:15:52.320)

that Ethereum was doing.

Lex Fridman (1:15:54.200)

It seemed like a super visionary thing.

Richard Craib (1:15:57.120)

I was reading something that was just

Lex Fridman (1:15:59.480)

going to change the world when I was reading the white paper.

Lex Fridman (1:16:03.240)

And I liked the different constructs

Lex Fridman (1:16:06.120)

you could have inside of Ethereum

Richard Craib (1:16:08.000)

that you couldn't have on Bitcoin.

Lex Fridman (1:16:10.200)

Like smart contracts and all that kind of stuff?

Richard Craib (1:16:11.800)

Exactly, yeah.

Lex Fridman (1:16:12.640)

And even spoke about different constructions you could have.

Richard Craib (1:16:18.880)

Yeah, that's a cool dance between Bitcoin and Ethereum.

Lex Fridman (1:16:21.920)

It's in the space of ideas.

Richard Craib (1:16:23.640)

It feels so young.

Lex Fridman (1:16:25.240)

Like, I wonder what cryptocurrencies will look like

Richard Craib (1:16:29.360)

in the future.

Lex Fridman (1:16:30.160)

If Bitcoin or Ethereum 2.0 or some version

Richard Craib (1:16:33.680)

will stick around or any of those,

Lex Fridman (1:16:35.920)

who's going to win out?

Lex Fridman (1:16:37.120)

Or if there's even a concept of winning out at all?

Lex Fridman (1:16:39.880)

Is there a cryptocurrency that you especially

Richard Craib (1:16:44.640)

find interesting that technically, financially,

Lex Fridman (1:16:48.960)

philosophically, you think is something

Lex Fridman (1:16:52.200)

you're keeping your eye on?

Lex Fridman (1:16:54.080)

Well, I don't really.

Richard Craib (1:16:55.360)

I'm not looking to invest in cryptocurrencies anymore.

Lex Fridman (1:16:59.200)

But I mean, and many are almost identical.

Richard Craib (1:17:05.040)

I mean, there wasn't too much difference

Lex Fridman (1:17:09.600)

between even Ethereum and Bitcoin in some ways, right?

Lex Fridman (1:17:13.520)

But there are some that I like the privacy ones.

Lex Fridman (1:17:15.880)

I mean, I like Zcash for its coolness.

Richard Craib (1:17:19.120)

It's actually a different kind of invention

Lex Fridman (1:17:23.640)

compared to some of the other things.

Lex Fridman (1:17:25.560)

OK, can you speak just briefly to privacy?

Lex Fridman (1:17:29.520)

Is there some mechanism of preserving

Lex Fridman (1:17:31.200)

some privacy of the universe?

Lex Fridman (1:17:32.680)

So I guess everything is public.

Richard Craib (1:17:34.720)

Yeah.

Lex Fridman (1:17:35.200)

Is that the problem?

Richard Craib (1:17:36.360)

Yeah, none of the transactions are private.

Lex Fridman (1:17:40.000)

And so I have some numeraire.

Lex Fridman (1:17:46.640)

And you can just see it.

Lex Fridman (1:17:48.120)

In fact, you can go to a website and it says like,

Richard Craib (1:17:50.040)

you can go to like, etherscan.

Lex Fridman (1:17:51.520)

And it'll say like, numeraire founder.

Lex Fridman (1:17:54.640)

And I'm like, how the hell do you guys know this?

Lex Fridman (1:17:57.720)

So they can reverse the near, whatever that's called.

Richard Craib (1:18:00.200)

Yeah, and so they can see me move it, too.

Lex Fridman (1:18:02.160)

They can see me.

Lex Fridman (1:18:02.920)

Oh, why is he moving it?

Lex Fridman (1:18:04.040)

Yeah.

Lex Fridman (1:18:06.040)

So but yeah, Zcash.

Lex Fridman (1:18:10.520)

Also, when you can make private transactions,

Richard Craib (1:18:12.640)

you can also play different games.

Lex Fridman (1:18:14.400)

Yes.

Lex Fridman (1:18:15.560)

And it's unclear.

Lex Fridman (1:18:17.840)

It's like what's quite cool about Zcash

Richard Craib (1:18:19.360)

is I wonder what games are being played there.

Lex Fridman (1:18:21.640)

No one will know.

Lex Fridman (1:18:23.040)

So from a deeply analytical perspective,

Lex Fridman (1:18:27.200)

can you describe why Dogecoin is going to win?

Richard Craib (1:18:31.200)

Which it surely will.

Lex Fridman (1:18:32.440)

Like it very likely will take over the world.

Lex Fridman (1:18:34.960)

And once we expand out into the universe,

Lex Fridman (1:18:37.360)

we'll take over the universe.

Richard Craib (1:18:40.080)

Or on a more serious note, like what

Lex Fridman (1:18:42.480)

are your thoughts on the recent success of Dogecoin

Richard Craib (1:18:45.160)

where you've spoken to sort of the meme stocks,

Lex Fridman (1:18:49.200)

the memetics of the whole thing, that it feels like the joke can

Richard Craib (1:18:55.240)

become the reality.

Lex Fridman (1:18:58.160)

Like the meme, the joke has power in this world.

Richard Craib (1:19:02.400)

Yeah.

Lex Fridman (1:19:02.880)

It's fascinating.

Richard Craib (1:19:04.080)

Exactly.

Lex Fridman (1:19:05.520)

It's like why is it correlated with Elon tweeting about it?

Lex Fridman (1:19:12.240)

It's not just Elon alone tweeting, right?

Lex Fridman (1:19:15.040)

It's like Elon tweeting and that becomes

Richard Craib (1:19:17.960)

a catalyst for everybody on the internet kind of like spreading

Lex Fridman (1:19:22.080)

the joke, right?

Richard Craib (1:19:22.960)

Exactly.

Lex Fridman (1:19:23.480)

The joke of it.

Lex Fridman (1:19:24.200)

So it's the initial spark of the fire for Wall Street

Lex Fridman (1:19:28.560)

bets type of situation.

Richard Craib (1:19:30.520)

Yeah.

Lex Fridman (1:19:31.040)

And that's fascinating because jokes

Richard Craib (1:19:33.680)

seem to spread faster than other mechanisms.

Lex Fridman (1:19:37.880)

Like funny shit is very effective at captivating

Richard Craib (1:19:43.440)

the discourse on the internet.

Lex Fridman (1:19:47.160)

Yeah, and I think you can have, like I like the one meme,

Richard Craib (1:19:51.520)

like Doge, I haven't heard that name in a long time.

Lex Fridman (1:19:57.760)

Like I think back to that meme often.

Richard Craib (1:20:00.760)

That's like funny.

Lex Fridman (1:20:01.640)

And every time I think back to it,

Lex Fridman (1:20:04.120)

there's a little probability that I might buy it, right?

Lex Fridman (1:20:08.680)

And so imagine you have millions of people who have had

Richard Craib (1:20:12.080)

all these great jokes, told them,

Lex Fridman (1:20:14.400)

and every now and then they reminisce,

Richard Craib (1:20:16.320)

oh, that was really funny.

Lex Fridman (1:20:17.840)

And then they're like, let me buy some.

Richard Craib (1:20:21.960)

Wouldn't that be interesting if we travel in time

Lex Fridman (1:20:25.440)

like multiple centuries where the entirety

Lex Fridman (1:20:28.720)

of the communication of the human species is like humor?

Lex Fridman (1:20:33.400)

Like it's all just jokes.

Richard Craib (1:20:35.480)

Like we're high on probably some really advanced drugs

Lex Fridman (1:20:39.840)

and we're all just laughing nonstop.

Richard Craib (1:20:42.480)

It's a weird like dystopian future of just humor.

Lex Fridman (1:20:47.320)

Elon has made me realize how like good it feels

Richard Craib (1:20:53.200)

to just not take shit seriously every once in a while

Lex Fridman (1:20:55.400)

and just relieve like the pressure of the world.

Richard Craib (1:20:58.600)

At the same time, the reason I don't always like

Lex Fridman (1:21:03.240)

when people finish their sentences with lol

Richard Craib (1:21:06.360)

is like when you don't take anything seriously.

Lex Fridman (1:21:11.360)

When everything becomes a joke,

Richard Craib (1:21:13.840)

then it feels like that way of thinking

Lex Fridman (1:21:20.200)

feels like it will destroy the world.

Richard Craib (1:21:22.280)

It's like, I often think of like,

Lex Fridman (1:21:24.000)

will memes save the world or destroy it?

Richard Craib (1:21:25.880)

Because I think both are possible directions.

Lex Fridman (1:21:28.560)

Yeah, I think this is a big problem.

Richard Craib (1:21:30.320)

I mean, I always felt that about America,

Lex Fridman (1:21:33.200)

a lot of people are telling jokes kind of all the time

Lex Fridman (1:21:36.320)

and they're kind of good at it.

Lex Fridman (1:21:37.800)

And you take someone aside, an American,

Richard Craib (1:21:42.080)

you're like, I really wanna have a sincere conversation.

Lex Fridman (1:21:44.240)

It's like hard to even keep a straight face

Richard Craib (1:21:46.760)

because everything is so, there's so much levity.

Lex Fridman (1:21:50.400)

So it's complicated.

Richard Craib (1:21:51.320)

I like how sincere actually like your Twitter can be.

Lex Fridman (1:21:54.920)

You're like, I am in love with the world today.

Richard Craib (1:21:57.920)

I get so much shit for it, it's hilarious.

Lex Fridman (1:22:00.640)

I'm never gonna stop because I realize like,

Richard Craib (1:22:03.120)

you have to be able to sometimes just be real

Lex Fridman (1:22:05.240)

and be positive and just be, say the cliche things,

Richard Craib (1:22:09.200)

which ultimately those things actually capture

Lex Fridman (1:22:11.960)

some fundamental truths about life.

Lex Fridman (1:22:15.040)

But it's a dance.

Lex Fridman (1:22:16.560)

And I think Elon does a good job of that

Richard Craib (1:22:20.560)

from an engineering perspective of being able to joke,

Lex Fridman (1:22:22.960)

but everyone's mostly to pull back and be like,

Richard Craib (1:22:26.800)

here's real problems, let's solve them and so on.

Lex Fridman (1:22:29.720)

And then be able to jump back to a joke.

Lex Fridman (1:22:31.800)

So it's ultimately, I think, I guess a skill that we

Lex Fridman (1:22:36.280)

have to learn.

Lex Fridman (1:22:39.040)

But I guess your advice is to invest everything

Lex Fridman (1:22:41.760)

anyone listening owns into Dogecoin.

Richard Craib (1:22:44.640)

That's what I heard from this interaction.

Lex Fridman (1:22:46.120)

Yeah, no, exactly.

Richard Craib (1:22:46.960)

Yeah, our hedge fund is unavailable.

Lex Fridman (1:22:50.360)

Just go straight to Dogecoin.

Richard Craib (1:22:52.320)

You're running a successful company.

Lex Fridman (1:22:55.120)

It's just interesting because my mind has been in that space

Richard Craib (1:22:58.600)

of potentially just being one of the millions of other

Lex Fridman (1:23:01.920)

entrepreneurs.

Richard Craib (1:23:03.360)

What's your advice on how to build a successful startup,

Lex Fridman (1:23:08.400)

how to build a successful company?

Richard Craib (1:23:10.480)

I think that one thing I do like,

Lex Fridman (1:23:13.760)

and it might be a particular thing about America,

Lex Fridman (1:23:16.880)

but there is something about playing,

Lex Fridman (1:23:20.600)

tell people what you really want to happen in the world.

Richard Craib (1:23:23.960)

Don't stop.

Lex Fridman (1:23:25.400)

It's not gonna make it,

Richard Craib (1:23:28.720)

like if you're asking someone to invest in your company,

Lex Fridman (1:23:31.080)

don't say, I think maybe one day we might make

Richard Craib (1:23:33.720)

a million dollars.

Lex Fridman (1:23:35.920)

When you actually believe something else,

Richard Craib (1:23:38.520)

you actually believe you're actually more optimistic,

Lex Fridman (1:23:41.480)

but you're toning down your optimism because you want

Richard Craib (1:23:45.240)

to appear like low risk.

Lex Fridman (1:23:50.080)

But actually it's super high risk if your company

Richard Craib (1:23:53.320)

becomes mediocre because no one wants to work

Lex Fridman (1:23:56.720)

in a mediocre company.

Richard Craib (1:23:57.560)

No one wants to invest in a mediocre company.

Lex Fridman (1:24:00.000)

So you should play the real game.

Lex Fridman (1:24:02.640)

And obviously this doesn't apply to all businesses,

Lex Fridman (1:24:04.480)

but if you play a venture backed startup kind of game,

Richard Craib (1:24:07.720)

like play for keeps, play to win, go big.

Lex Fridman (1:24:11.600)

And it's very hard to do that.

Richard Craib (1:24:13.240)

I've always feel like, yeah,

Lex Fridman (1:24:18.200)

you can start narrowing your focus because 10 people

Richard Craib (1:24:22.040)

are telling you, you gotta care about this boring thing

Lex Fridman (1:24:26.200)

that won't matter five years from now.

Lex Fridman (1:24:28.000)

And you should push back and do the real,

Lex Fridman (1:24:31.600)

play the real game.

Lex Fridman (1:24:32.440)

So be bold.

Lex Fridman (1:24:33.840)

So both, I mean, there's an interesting duality there.

Lex Fridman (1:24:37.840)

So there's the way you speak to other people

Lex Fridman (1:24:41.080)

about like your plans and what you are like privately

Richard Craib (1:24:45.360)

just in your own mind.

Lex Fridman (1:24:48.120)

And maybe it's connected with what you were saying about,

Richard Craib (1:24:50.160)

yeah, sincerity as well.

Lex Fridman (1:24:51.920)

Like if you appear to be sincerely optimistic

Richard Craib (1:24:55.800)

about something that's big or crazy,

Lex Fridman (1:24:59.680)

it's putting yourself up to be kind of like

Richard Craib (1:25:01.640)

ridiculed or something.

Lex Fridman (1:25:03.160)

And so if you say, my mission is to, yeah, go to Mars,

Richard Craib (1:25:08.000)

it's just so bonkers that it's hard to say.

Lex Fridman (1:25:12.640)

It is, but one powerful thing, just like you said,

Richard Craib (1:25:17.680)

is if you say it and you believe it,

Lex Fridman (1:25:20.240)

then actually amazing people come and work with you.

Richard Craib (1:25:25.680)

Exactly.

Lex Fridman (1:25:26.520)

It's not just skill, but the dreams.

Richard Craib (1:25:28.800)

There's something about optimism that,

Lex Fridman (1:25:31.640)

like that fire that you have when you're optimistic

Richard Craib (1:25:33.960)

of actually having the hope of building

Lex Fridman (1:25:36.200)

something totally cool, something totally new,

Richard Craib (1:25:38.960)

that when those people get in a room together,

Lex Fridman (1:25:41.120)

like they can actually do it.

Richard Craib (1:25:42.840)

Yeah.

Lex Fridman (1:25:43.840)

Yeah, there's, yeah, that's,

Lex Fridman (1:25:47.480)

and also makes life really fun when you're in that room.

Lex Fridman (1:25:50.800)

So all of that together, ultimately,

Richard Craib (1:25:55.080)

I don't know, that's what makes this crazy ride

Lex Fridman (1:25:57.160)

of a startup really look fun.

Lex Fridman (1:25:59.640)

And Elon is an example of a person who succeeded at that.

Lex Fridman (1:26:02.520)

There's not many other inspiring figures, which is sad.

Richard Craib (1:26:06.040)

I used to be at Google and there's something that happens

Lex Fridman (1:26:11.120)

that sometimes when the company grows

Richard Craib (1:26:13.200)

bigger and bigger and bigger,

Lex Fridman (1:26:14.440)

where that kind of ambition kind of quiets down a little bit.

Richard Craib (1:26:18.760)

Yeah.

Lex Fridman (1:26:19.600)

Google had this ambition, still does,

Richard Craib (1:26:21.840)

of making the world's information accessible to everyone.

Lex Fridman (1:26:24.880)

And I remember, I don't know, that's beautiful.

Richard Craib (1:26:28.320)

I still love that dream of that they used to scan books,

Lex Fridman (1:26:33.160)

but just in every way possible

Richard Craib (1:26:34.760)

make the world's information accessible.

Lex Fridman (1:26:37.480)

Same with Wikipedia.

Richard Craib (1:26:38.640)

Every time I open up Wikipedia,

Lex Fridman (1:26:40.480)

I'm just awe inspired by how awesome humans are, man.

Lex Fridman (1:26:47.200)

And creating this together,

Lex Fridman (1:26:48.600)

I don't know what the meanings are over there,

Lex Fridman (1:26:50.280)

but it's just beautiful.

Lex Fridman (1:26:52.720)

Like what they've created is incredible.

Lex Fridman (1:26:55.200)

And I'd love to be able to be part of something like that.

Lex Fridman (1:26:58.880)

And you're right, for that, you have to be bold.

Lex Fridman (1:27:01.800)

And there's, and strange to me also,

Lex Fridman (1:27:03.480)

I think you're right that there's

Lex Fridman (1:27:04.600)

how many boring companies there are.

Lex Fridman (1:27:06.760)

Something I always talk about, especially in FinTech,

Richard Craib (1:27:08.880)

it's like, why am I excited about, this is so lame.

Lex Fridman (1:27:13.400)

Like what is, this isn't even important.

Richard Craib (1:27:16.360)

Even if you succeed, this is gonna be like terrible.

Lex Fridman (1:27:19.320)

This is not good.

Lex Fridman (1:27:21.040)

And it's just strange how people can kind of

Lex Fridman (1:27:23.320)

get fake enthusiastic about like boring ideas

Richard Craib (1:27:27.720)

when there's so many bigger ideas that,

Lex Fridman (1:27:32.600)

yeah, I mean, you read these things,

Richard Craib (1:27:33.680)

like this company raises money,

Lex Fridman (1:27:35.000)

and it's just like, that's a lot of money

Richard Craib (1:27:36.400)

for the worst idea I've ever heard.

Lex Fridman (1:27:38.520)

Some ideas are really big.

Lex Fridman (1:27:41.440)

So like I worked on autonomous vehicles quite a bit.

Lex Fridman (1:27:44.520)

And there's so many ways in which you can present

Richard Craib (1:27:48.080)

that idea to yourself, to the team you work with,

Lex Fridman (1:27:50.680)

to just, yeah, like to yourself when you're quietly

Richard Craib (1:27:53.120)

looking in the mirror in the morning,

Lex Fridman (1:27:55.200)

that's really boring or really exciting.

Richard Craib (1:27:58.320)

Like if you're really ambitious with autonomous vehicles,

Lex Fridman (1:28:01.880)

it changes the nature of like human robot interaction,

Richard Craib (1:28:06.320)

it changes the nature of how we move.

Lex Fridman (1:28:08.160)

Forget money, forget all that stuff.

Richard Craib (1:28:09.920)

It changes like everything about robotics and AI,

Lex Fridman (1:28:13.240)

machine learning, it changes everything about manufacture.

Richard Craib (1:28:16.440)

I mean, cars, transportation is so fundamentally connected

Lex Fridman (1:28:20.080)

to cars, and if that changes,

Richard Craib (1:28:22.320)

it's changing the fabric of society,

Lex Fridman (1:28:24.560)

of movies, of everything.

Lex Fridman (1:28:27.360)

And if you go bold and take risks

Lex Fridman (1:28:29.560)

and be willing to go bankrupt with your company,

Richard Craib (1:28:33.800)

as opposed to cautiously, you can really change the world.

Lex Fridman (1:28:37.520)

And it's so sad for me to see all these autonomous companies,

Richard Craib (1:28:40.560)

autonomous vehicle companies,

Lex Fridman (1:28:41.520)

they're like really more focused about fundraising

Lex Fridman (1:28:45.160)

and kind of like smoke and mirrors,

Lex Fridman (1:28:46.400)

they're really afraid,

Richard Craib (1:28:48.240)

the entirety of their marketing is grounded in fear

Lex Fridman (1:28:51.760)

and presenting enough smoke to where they keep raising funds

Lex Fridman (1:28:54.960)

so they can cautiously use technology of a previous decade

Lex Fridman (1:28:59.080)

or previous two decades to kind of test vehicles

Richard Craib (1:29:02.400)

here and there, as opposed to do crazy things

Lex Fridman (1:29:04.960)

and bold and go huge at scale, do huge data collection.

Lex Fridman (1:29:10.080)

And yeah, so that's just an example.

Lex Fridman (1:29:12.760)

Like the idea can be big,

Lex Fridman (1:29:14.400)

but if you don't allow yourself to take that idea

Lex Fridman (1:29:17.080)

and think really big with it,

Richard Craib (1:29:20.160)

then you're not gonna make anything happen.

Lex Fridman (1:29:22.320)

Yeah, you're absolutely right in that.

Lex Fridman (1:29:24.200)

So you've been connected in your work

Lex Fridman (1:29:28.880)

with a bunch of amazing people.

Lex Fridman (1:29:31.280)

How much interaction do you have with investors,

Lex Fridman (1:29:34.440)

that whole process is an entire mystery to me.

Richard Craib (1:29:36.840)

Is there some people that just have influence

Lex Fridman (1:29:38.600)

on the trajectory of your thinking completely,

Lex Fridman (1:29:42.960)

or is it just this collective energy behind the company?

Lex Fridman (1:29:46.920)

Yeah, I mean, I came here and I was amazed how,

Richard Craib (1:29:52.160)

yeah, people would, I was only here for a few months

Lex Fridman (1:29:54.840)

and I met some incredible investors

Lex Fridman (1:29:57.160)

and I'd almost run out of money.

Lex Fridman (1:29:59.720)

And once they invested, I was like,

Richard Craib (1:30:04.880)

I am not gonna let you down.

Lex Fridman (1:30:06.800)

And I was like, okay, I'm gonna send them

Richard Craib (1:30:08.280)

like an email update every like three minutes.

Lex Fridman (1:30:11.600)

And then they don't care at all.

Lex Fridman (1:30:14.160)

So they kind of wanna, I don't know, like,

Lex Fridman (1:30:15.720)

so for some, I like it when it's just like,

Richard Craib (1:30:18.200)

they're always available to talk,

Lex Fridman (1:30:20.040)

but a lot of building a business,

Richard Craib (1:30:23.360)

especially a high tech business,

Lex Fridman (1:30:26.560)

there's a little for them to add, right?

Richard Craib (1:30:29.240)

There's little for them to add on product.

Lex Fridman (1:30:31.240)

There's a lot for them to add on like business development.

Lex Fridman (1:30:34.320)

And if we are doing product research,

Lex Fridman (1:30:36.400)

which is for us research into the market,

Richard Craib (1:30:39.000)

research into how to make a great hedge fund,

Lex Fridman (1:30:41.800)

and we do that for years,

Richard Craib (1:30:44.920)

there's not much to tell the investors.

Lex Fridman (1:30:47.120)

So that basically is like, I believe in you.

Richard Craib (1:30:49.040)

There's something, I like the cut of your jib.

Lex Fridman (1:30:52.440)

There's something in your idea, in your ambition,

Richard Craib (1:30:55.760)

in your plans that I like.

Lex Fridman (1:30:57.880)

And it's almost like a pat on the back.

Richard Craib (1:30:59.800)

It's like, go get them kid.

Lex Fridman (1:31:01.560)

Yeah, it is a bit like that.

Lex Fridman (1:31:02.720)

And that's cool.

Lex Fridman (1:31:04.360)

That's a good way to do it.

Richard Craib (1:31:05.280)

I'm glad they do it that way.

Lex Fridman (1:31:07.120)

Like the one meeting I had,

Richard Craib (1:31:08.800)

which was like really good with this

Lex Fridman (1:31:10.240)

was meeting Howard Morgan,

Richard Craib (1:31:13.280)

who's actually a co founder of Renaissance Technologies

Lex Fridman (1:31:16.760)

in the like 1980s and worked with Jim Simons.

Lex Fridman (1:31:21.000)

And he was in the room

Lex Fridman (1:31:25.840)

and I was meeting some other guy and he was in the room

Lex Fridman (1:31:28.200)

and I was explaining how quantitative finance works.

Lex Fridman (1:31:33.000)

I was like, so they use mathematical models.

Lex Fridman (1:31:36.600)

And then he was like, yeah, I started Renaissance.

Lex Fridman (1:31:40.480)

I know a bit about this.

Lex Fridman (1:31:43.160)

And then I was like, oh my God.

Lex Fridman (1:31:46.680)

So yeah, but then, and then I think he kind of said, well,

Richard Craib (1:31:50.320)

yeah, he said, well, cause I was talking,

Lex Fridman (1:31:52.160)

he was working at first round capital as a partner

Lex Fridman (1:31:55.920)

and they kind of said, they didn't want to invest.

Lex Fridman (1:31:59.200)

And then I wrote a blog post describing the idea

Lex Fridman (1:32:01.920)

and I was like, I really think you guys should invest.

Lex Fridman (1:32:03.640)

And then they end up.

Richard Craib (1:32:04.560)

Oh, interesting.

Lex Fridman (1:32:05.760)

You convinced them on that.

Richard Craib (1:32:06.600)

That must be good.

Lex Fridman (1:32:07.440)

Yeah, cause they're like,

Richard Craib (1:32:08.280)

we don't really invest in hedge funds.

Lex Fridman (1:32:09.120)

And I was like, you don't see like what I'm doing.

Lex Fridman (1:32:11.240)

This is so a tech company, not a hedge fund, right?

Lex Fridman (1:32:14.400)

Yeah, and Numerai is brilliant.

Richard Craib (1:32:15.800)

It's, when it caught my eye,

Lex Fridman (1:32:18.240)

there's something special there.

Lex Fridman (1:32:19.240)

So I really do hope you succeed in the,

Lex Fridman (1:32:22.720)

obviously it's a risky thing you're taking on,

Richard Craib (1:32:24.880)

the ambition of it, the size of it,

Lex Fridman (1:32:26.560)

but I do hope you succeed.

Richard Craib (1:32:28.320)

You mentioned Jim Simons.

Lex Fridman (1:32:30.400)

He comes up in another world of mine really often on the,

Richard Craib (1:32:34.520)

he's just a brilliant guy on the mathematics side

Lex Fridman (1:32:37.720)

as a mathematician,

Lex Fridman (1:32:38.760)

but he's also brilliant finance hedge fund manager guy.

Lex Fridman (1:32:44.120)

Have you gotten a chance to interact with him at all?

Richard Craib (1:32:47.040)

Have you learned anything from him on the math,

Lex Fridman (1:32:51.080)

on the finance, on the philosophy, life side, things?

Richard Craib (1:32:54.960)

I've played poker with him.

Lex Fridman (1:32:56.280)

It was pretty cool.

Richard Craib (1:32:57.120)

It was like, actually in the show, Billions,

Lex Fridman (1:32:59.680)

they kind of do a little thing about this poker tournament

Richard Craib (1:33:02.680)

thing with all the hedge fund managers.

Lex Fridman (1:33:04.240)

And that's real life thing.

Lex Fridman (1:33:06.840)

And they have a lot of like world series of bracelet,

Lex Fridman (1:33:09.760)

world series poker bracelets holders,

Lex Fridman (1:33:11.800)

but it's kind of Jim's thing.

Lex Fridman (1:33:13.520)

And I met him there and yeah, it was kind of brief,

Lex Fridman (1:33:19.200)

but I was just like, he's like, oh, how do you,

Lex Fridman (1:33:21.480)

why are you here?

Lex Fridman (1:33:22.320)

And I was like, oh, Howard sent me, you know,

Lex Fridman (1:33:23.880)

he's like, go play this tournament,

Richard Craib (1:33:25.960)

meet some of the other players.

Lex Fridman (1:33:27.800)

And then...

Lex Fridman (1:33:29.000)

Was it Texas Holdem?

Lex Fridman (1:33:30.120)

Yeah, Texas Holdem tournament, yeah.

Lex Fridman (1:33:32.240)

Do you play poker yourself or was it?

Lex Fridman (1:33:33.800)

Yeah, I do.

Richard Craib (1:33:34.880)

I mean, it was crazy.

Lex Fridman (1:33:36.440)

On my right was the CEO,

Richard Craib (1:33:39.080)

who's the current CEO of Renaissance, Peter Brown.

Lex Fridman (1:33:42.200)

And Peter Muller, who's a hedge fund manager at PDT.

Lex Fridman (1:33:49.160)

And yeah, I mean, it was just like,

Lex Fridman (1:33:50.440)

and then, you know, just everyone.

Lex Fridman (1:33:51.760)

And then all these bracelet world series,

Lex Fridman (1:33:53.200)

like people that I know from like TV.

Lex Fridman (1:33:56.960)

And Robert Mercer, who's fucking crazy.

Lex Fridman (1:34:01.440)

Who's that?

Richard Craib (1:34:02.280)

He's the guy who donated the most money to Trump.

Lex Fridman (1:34:08.280)

And he's just like...

Richard Craib (1:34:09.120)

It's a lot of personality.

Lex Fridman (1:34:10.240)

Character, yeah, geez, it's crazy.

Lex Fridman (1:34:13.920)

So it's quite cool how, yeah, like the, it was really fun.

Lex Fridman (1:34:17.600)

And then I managed to knock out Peter Muller.

Richard Craib (1:34:19.600)

I have a, I got a little trophy for knocking him out

Lex Fridman (1:34:22.680)

because he was a previous champion.

Richard Craib (1:34:24.440)

In fact, I think he's won the most.

Lex Fridman (1:34:25.880)

I think he's won three times.

Richard Craib (1:34:27.560)

Super smart guy.

Lex Fridman (1:34:30.520)

But I will say Jim outlasted me in the tournament.

Lex Fridman (1:34:35.520)

And they're all extremely good at poker,

Lex Fridman (1:34:41.680)

but they're also, so it was a $10,000 buy in.

Lex Fridman (1:34:45.880)

And I was like, this is kind of expensive,

Lex Fridman (1:34:50.400)

but it all goes to charity, Jim's math charity.

Lex Fridman (1:34:54.440)

But then the way they play, they have like rebis

Lex Fridman (1:34:58.360)

and like they all do a shit ton of rebis

Richard Craib (1:35:01.080)

because it's for charity.

Lex Fridman (1:35:02.720)

So immediately they're like going all in

Lex Fridman (1:35:06.240)

and I'm like, man, like, so I ended up adding more as well.

Lex Fridman (1:35:12.000)

So it's like you couldn't play at all without doing that.

Richard Craib (1:35:15.360)

Yeah, the stakes are high.

Lex Fridman (1:35:16.640)

But you're connected to a lot of these folks

Richard Craib (1:35:18.760)

that are kind of titans of just of economics

Lex Fridman (1:35:25.760)

and tech in general.

Lex Fridman (1:35:27.200)

Do you feel a burden from this?

Lex Fridman (1:35:28.720)

You're a young guy.

Richard Craib (1:35:30.200)

I did feel a bit out of place there.

Lex Fridman (1:35:33.760)

The company was quite new

Lex Fridman (1:35:35.640)

and they also don't speak about things, right?

Lex Fridman (1:35:39.920)

So it's not like going to meet a famous rocket engineer

Richard Craib (1:35:44.320)

who will tell you how to make a rocket.

Lex Fridman (1:35:46.200)

They do not want to tell you anything

Richard Craib (1:35:48.040)

about how to make a hedge fund.

Lex Fridman (1:35:49.600)

It's like all secretive and that part I didn't like.

Lex Fridman (1:35:54.200)

And they were also kind of making fun of me a little bit.

Lex Fridman (1:35:57.720)

Like they would say, like they'd call me like,

Richard Craib (1:36:00.600)

I don't know, the Bitcoin kid or.

Lex Fridman (1:36:02.000)

Yeah, yeah, yeah.

Lex Fridman (1:36:02.840)

And then they would say, even things like,

Lex Fridman (1:36:05.280)

member Peter, yeah, said to me something like,

Richard Craib (1:36:08.880)

I don't think AI is gonna have a big role in finance.

Lex Fridman (1:36:12.960)

And I was like, hearing this from the CEO of Renaissance

Richard Craib (1:36:16.040)

was like weird to hear because I was like,

Lex Fridman (1:36:17.960)

of course it will.

Lex Fridman (1:36:19.080)

And he's like, but he can see,

Lex Fridman (1:36:21.200)

I can see it having a really big impact

Richard Craib (1:36:23.400)

on things like self driving cars.

Lex Fridman (1:36:25.600)

But finance, it's too noisy and whatever.

Lex Fridman (1:36:28.280)

And so I don't think it's like the perfect application.

Lex Fridman (1:36:30.640)

And I was like, that was interesting to hear

Richard Craib (1:36:32.520)

because it's like, and I think it was that same day

Lex Fridman (1:36:35.560)

that Libra, I think it is, the poker playing AI

Richard Craib (1:36:40.840)

started to beat like the human.

Lex Fridman (1:36:42.840)

So it was kind of funny hearing them like say,

Richard Craib (1:36:44.480)

oh, I'm not sure AI could ever attack that problem.

Lex Fridman (1:36:47.280)

And then that very day it's attacking the problem

Richard Craib (1:36:49.360)

of the game we're playing.

Lex Fridman (1:36:51.080)

Well, there's a kind of a magic to somebody

Richard Craib (1:36:55.720)

who's exceptionally successful looking at you,

Lex Fridman (1:36:59.920)

giving you respect, but also saying that what you're doing

Richard Craib (1:37:03.560)

is not going to succeed in a sense.

Lex Fridman (1:37:06.280)

Like they're not really saying it,

Lex Fridman (1:37:08.200)

but I tend to believe from my interactions with people

Lex Fridman (1:37:11.240)

that it's a kind of prod to say like, prove me wrong.

Richard Craib (1:37:14.880)

Yeah.

Lex Fridman (1:37:15.720)

That's ultimately, that's how those guys talk.

Richard Craib (1:37:18.040)

They see good talent and they're like.

Lex Fridman (1:37:20.440)

Yeah.

Lex Fridman (1:37:21.280)

And I think they're also saying

Lex Fridman (1:37:22.400)

it's not gonna succeed quickly in some way.

Richard Craib (1:37:25.880)

They're like, this is gonna take a long time

Lex Fridman (1:37:29.080)

and maybe that's good to know.

Lex Fridman (1:37:32.760)

And certainly AI in trading,

Lex Fridman (1:37:36.320)

that's one of the most philosophically interesting questions

Richard Craib (1:37:42.280)

about artificial intelligence and the nature of money.

Lex Fridman (1:37:45.920)

Because it's like, how much can you extract

Richard Craib (1:37:48.920)

in terms of patterns from all of these millions

Lex Fridman (1:37:52.360)

of humans interacting using this methodology of money?

Richard Craib (1:37:57.240)

It's like one of the open questions

Lex Fridman (1:37:58.680)

in the artificial intelligence.

Richard Craib (1:37:59.680)

In that sense, you converting into a data set

Lex Fridman (1:38:02.360)

is one of like the biggest gifts to the research community,

Richard Craib (1:38:07.120)

to the whole, anyone who loves data science and AI,

Lex Fridman (1:38:11.080)

this is kind of fascinating.

Richard Craib (1:38:14.080)

I'd love to see where this goes actually.

Lex Fridman (1:38:15.680)

I think I say sometimes long before AGI destroys the world,

Richard Craib (1:38:19.840)

a narrow intelligence will win all the money

Lex Fridman (1:38:21.800)

in the stock market.

Richard Craib (1:38:23.240)

Way, like just a narrow AI.

Lex Fridman (1:38:25.440)

Yeah.

Lex Fridman (1:38:26.360)

And I don't know if I'm gonna be the one who invents that.

Lex Fridman (1:38:29.920)

So I'm building Numerai to make sure

Richard Craib (1:38:31.960)

that that narrow AI uses our data.

Lex Fridman (1:38:35.840)

So you're giving a platform

Richard Craib (1:38:37.000)

to where millions of people can participate

Lex Fridman (1:38:38.800)

and do build that narrow AI themselves.

Richard Craib (1:38:43.280)

People love it when I ask this kind of question

Lex Fridman (1:38:45.600)

about books, about ideas and philosophers and so on.

Richard Craib (1:38:50.560)

I was wondering if you had books or ideas,

Lex Fridman (1:38:56.760)

philosophers, thinkers that had an influence

Richard Craib (1:38:59.120)

on your life when you were growing up

Lex Fridman (1:39:01.440)

or just today that you would recommend

Richard Craib (1:39:04.520)

that people check out blog posts, podcasts, videos,

Lex Fridman (1:39:08.880)

all that kind of stuff.

Richard Craib (1:39:09.720)

Is there something that just kind of had an impact on you

Lex Fridman (1:39:12.080)

that you couldn't recommend?

Richard Craib (1:39:13.760)

A super kind of obvious one that I really,

Lex Fridman (1:39:19.000)

I was reading Zero to One while coming up with Numerai.

Richard Craib (1:39:22.240)

It was like halfway through the book.

Lex Fridman (1:39:24.880)

And I really do like a lot of the ideas there.

Lex Fridman (1:39:27.160)

And it's also about kind of thinking big

Lex Fridman (1:39:29.360)

and also it's like peculiar little book.

Richard Craib (1:39:34.760)

It's like why, like there's a little picture

Lex Fridman (1:39:36.400)

of the hipster versus Unabomber.

Lex Fridman (1:39:38.800)

And it's a weird little book.

Lex Fridman (1:39:40.480)

So I like, there's kind of like some depth there.

Richard Craib (1:39:42.760)

In terms of a book on a, if you're thinking

Lex Fridman (1:39:44.920)

of doing a startup, that's a good book.

Richard Craib (1:39:47.320)

A book I like a lot is maybe my favorite book

Lex Fridman (1:39:52.320)

is David Deutsch's Beginning of Infinity.

Richard Craib (1:39:56.200)

I just found that so optimistic.

Lex Fridman (1:40:00.360)

It puts you, everything you read in science,

Richard Craib (1:40:03.800)

it like makes the world feel like kind of colder

Lex Fridman (1:40:06.520)

because like it's like we're just coming from evolution

Lex Fridman (1:40:10.640)

and coming from nothing should be this way or whatever.

Lex Fridman (1:40:15.200)

And humans are not very powerful.

Richard Craib (1:40:16.800)

We're just like scum on the earth.

Lex Fridman (1:40:19.240)

And the way David Deutsch sees things

Lex Fridman (1:40:20.960)

and argues, he argues them with the same rigor

Lex Fridman (1:40:23.840)

that the cynics often use

Lex Fridman (1:40:26.200)

and then has a much better conclusion.

Lex Fridman (1:40:29.840)

That's some of the statements of things like,

Richard Craib (1:40:33.480)

anything that doesn't violate the laws

Lex Fridman (1:40:35.000)

of physics can be solved.

Lex Fridman (1:40:39.280)

So ultimately arriving at a hopeful,

Lex Fridman (1:40:41.800)

like a hopeful path forward.

Richard Craib (1:40:42.640)

Yeah, without being like a hippie.

Lex Fridman (1:40:45.680)

You've mentioned kind of advice for startups.

Richard Craib (1:40:47.640)

Is there, in general, whether you do a startup or not,

Lex Fridman (1:40:50.600)

do you have advice for young people today?

Richard Craib (1:40:52.680)

You're like an example of somebody

Lex Fridman (1:40:54.160)

who's paved their own path

Lex Fridman (1:40:56.240)

and were, I would say exceptionally successful.

Lex Fridman (1:40:59.200)

Is there advice, somebody who's like 20 today, 18,

Richard Craib (1:41:02.800)

undergrad or thinking about going to college,

Lex Fridman (1:41:05.640)

or in college and so on that you would give them?

Richard Craib (1:41:09.520)

I think I often tell young people don't start companies.

Lex Fridman (1:41:13.560)

Is it not, don't start a company

Richard Craib (1:41:16.040)

unless you're prepared to make it your life's work.

Lex Fridman (1:41:19.120)

Like that's a really good way of putting it.

Lex Fridman (1:41:22.480)

And a lot of people think, well, this semester

Lex Fridman (1:41:25.760)

I'm gonna take a semester off.

Lex Fridman (1:41:27.280)

And in that one semester,

Lex Fridman (1:41:28.560)

I'm gonna start a company and sell it or whatever.

Lex Fridman (1:41:31.240)

And it's just like, what are you talking about?

Lex Fridman (1:41:33.360)

It doesn't really work that way.

Richard Craib (1:41:34.960)

You should be like super into the idea,

Lex Fridman (1:41:37.200)

so into it that you wanna spend a really long time on it.

Lex Fridman (1:41:41.560)

Is that more about psychology or actual time allocation?

Lex Fridman (1:41:44.040)

Like, is it literally the fact that you need to give 100%

Lex Fridman (1:41:47.560)

for potentially years for it to succeed?

Lex Fridman (1:41:49.840)

Or is it more about just the mindset that's required?

Richard Craib (1:41:53.600)

Yeah, I mean, I think, well, any, I think, yeah,

Lex Fridman (1:41:55.800)

you don't wanna have,

Richard Craib (1:41:56.640)

certainly don't wanna have a plan to sell the company

Lex Fridman (1:42:00.800)

like quickly or something,

Richard Craib (1:42:02.160)

or it's like a company that has a very,

Lex Fridman (1:42:05.560)

it's like a big fashion component.

Richard Craib (1:42:07.360)

Like it'll only work now.

Lex Fridman (1:42:08.720)

It's like an app or something.

Lex Fridman (1:42:12.640)

So yeah, that's a big one.

Lex Fridman (1:42:14.960)

And then I also think something I've thought about recently

Richard Craib (1:42:18.880)

is I had a job as a quant at a fund

Lex Fridman (1:42:23.480)

for about two and a half years.

Lex Fridman (1:42:25.920)

And part of me thinks if I had spent

Lex Fridman (1:42:29.080)

another two years there,

Richard Craib (1:42:31.080)

I would have learned a lot more

Lex Fridman (1:42:34.360)

and had even more knowledge to be where,

Richard Craib (1:42:38.480)

to basically accelerate how long Numerai took.

Lex Fridman (1:42:41.160)

So the idea that you can sit in an air conditioned room

Lex Fridman (1:42:44.720)

and get free food,

Lex Fridman (1:42:46.280)

or even sit at home now in your underwear

Lex Fridman (1:42:48.880)

and make a huge amount of money and learn whatever you want

Lex Fridman (1:42:53.400)

and get, it's just crazy.

Richard Craib (1:42:55.080)

It's such a good deal.

Lex Fridman (1:42:56.480)

Yeah, oh, that's interesting.

Richard Craib (1:42:57.920)

That's the case for, I was terrified of that.

Lex Fridman (1:43:00.360)

Like at Google, I thought I would become really comfortable

Richard Craib (1:43:04.040)

in that air conditioned room.

Lex Fridman (1:43:06.160)

And that, I was afraid the quant situation is,

Richard Craib (1:43:10.640)

I mean, what you present is really brilliant

Lex Fridman (1:43:13.280)

that it's exceptionally valuable, the lessons you learn,

Richard Craib (1:43:17.640)

because you get to get paid while you learn from others.

Lex Fridman (1:43:21.120)

If you see that, if you see jobs

Richard Craib (1:43:24.160)

in the space of your passion that way,

Lex Fridman (1:43:27.800)

that it's just an education.

Richard Craib (1:43:29.360)

It's like the best kind of education.

Lex Fridman (1:43:31.400)

But of course you have, from my perspective,

Richard Craib (1:43:34.200)

you have to be really careful on that to get comfortable.

Lex Fridman (1:43:37.320)

Again, in a relationship, then you buy a house

Richard Craib (1:43:39.760)

or whatever the hell it is, and then you get,

Lex Fridman (1:43:42.720)

and then you convince yourself like,

Richard Craib (1:43:44.200)

well, I have to pay these fees for the car,

Lex Fridman (1:43:46.640)

for the house, blah, blah, blah.

Lex Fridman (1:43:48.240)

And then there's momentum and all of a sudden

Lex Fridman (1:43:50.360)

you're on your death bed and there's grandchildren

Lex Fridman (1:43:53.080)

and you're drinking whiskey

Lex Fridman (1:43:55.160)

and complaining about kids these days.

Lex Fridman (1:43:56.960)

So I'm afraid of that momentum, but you're right.

Lex Fridman (1:44:00.800)

Like there's something special about the education

Richard Craib (1:44:04.120)

you get working at these companies.

Lex Fridman (1:44:06.320)

Yeah, and I remember on my desk,

Richard Craib (1:44:08.200)

I had like a bunch of papers on quant finance,

Lex Fridman (1:44:11.320)

a bunch of papers on optimization,

Lex Fridman (1:44:13.280)

and then the paper on Ethereum, just on my desk as well,

Lex Fridman (1:44:16.680)

and the white paper, and it's like,

Richard Craib (1:44:19.280)

it's been amazing how kind of, and you can learn about,

Lex Fridman (1:44:23.680)

so that, I also thought, I think this idea

Richard Craib (1:44:25.680)

of like learning about intersections of things,

Lex Fridman (1:44:28.440)

I don't think there are too many people that know

Richard Craib (1:44:30.240)

like as much about crypto and quant finance

Lex Fridman (1:44:33.640)

and machine learning as I do.

Lex Fridman (1:44:36.280)

And that's a really nice set of three things

Lex Fridman (1:44:39.280)

to know stuff about.

Lex Fridman (1:44:40.640)

And that was because I had like free time in my job.

Lex Fridman (1:44:45.160)

Okay, let me ask the perfectly impractical,

Lex Fridman (1:44:48.640)

but the most important question.

Lex Fridman (1:44:49.840)

What's the meaning of all the things you're trying to do

Lex Fridman (1:44:53.760)

so many amazing things, why?

Lex Fridman (1:44:56.200)

What's the meaning of this life of yours or ours?

Richard Craib (1:45:00.800)

I don't know.

Lex Fridman (1:45:01.840)

Humans.

Richard Craib (1:45:03.400)

Yeah, so I have yet had some people say,

Lex Fridman (1:45:06.300)

asking what meaning of life is,

Richard Craib (1:45:07.760)

is like asking the wrong question or something.

Lex Fridman (1:45:10.160)

The question is wrong.

Richard Craib (1:45:11.160)

Yeah.

Lex Fridman (1:45:12.000)

No, usually people get too nervous to be able to say that

Richard Craib (1:45:14.640)

because it's like your question sucks.

Lex Fridman (1:45:17.480)

I don't think there's an answer.

Richard Craib (1:45:18.800)

It's like the searching for it.

Lex Fridman (1:45:21.120)

It's like sometimes asking it.

Richard Craib (1:45:22.880)

It's like sometimes sitting back and looking up at the stars

Lex Fridman (1:45:25.560)

and being like, huh, I wonder if there's aliens up there.

Richard Craib (1:45:29.800)

There's a useful like a palette cleanser aspect to it

Lex Fridman (1:45:36.520)

because it kind of wakes you up to like all the little busy,

Richard Craib (1:45:39.600)

hurried day to day activities, all the meetings,

Lex Fridman (1:45:42.680)

all the things you'd like a part of.

Richard Craib (1:45:45.220)

We're just like ants, a part of a system,

Lex Fridman (1:45:47.640)

a part of another system.

Lex Fridman (1:45:48.920)

And then asking this bigger question

Lex Fridman (1:45:52.120)

allows you to kind of zoom out and think about it.

Lex Fridman (1:45:53.920)

But there's ultimately,

Lex Fridman (1:45:56.240)

I think it's an impossible thing for a limited capacity,

Richard Craib (1:45:58.880)

like cognitive capacity to capture.

Lex Fridman (1:46:01.520)

But it's fun to listen to somebody

Richard Craib (1:46:03.040)

who's exceptionally successful, exceptionally busy now,

Lex Fridman (1:46:07.240)

who's also young like you,

Richard Craib (1:46:09.800)

to ask these kinds of questions about like death.

Lex Fridman (1:46:14.640)

You know, do you consider your own mortality kind of thing

Lex Fridman (1:46:18.280)

and life, whether that enters your mind?

Lex Fridman (1:46:21.760)

Because it often doesn't,

Richard Craib (1:46:23.240)

it kind of almost gets in the way.

Lex Fridman (1:46:24.920)

Yeah.

Richard Craib (1:46:26.260)

It's amazing how many things you can like that are trivial

Lex Fridman (1:46:28.840)

that could occupy a lot of your mind

Richard Craib (1:46:31.280)

until something bad happens or something flips you.

Lex Fridman (1:46:36.400)

And then you start thinking about the people you love

Richard Craib (1:46:38.480)

that are in your life.

Lex Fridman (1:46:39.760)

Then you started thinking about like,

Richard Craib (1:46:41.080)

holy shit, this ride ends.

Lex Fridman (1:46:42.840)

Exactly.

Richard Craib (1:46:43.960)

Yeah, I just had COVID and I had it quite bad.

Lex Fridman (1:46:48.960)

What wasn't really bad is just like,

Richard Craib (1:46:51.720)

I also got a simultaneous like lung infection.

Lex Fridman (1:46:55.720)

So I had like almost like bronchitis or whatever.

Richard Craib (1:46:59.280)

I don't even, I don't understand that stuff,

Lex Fridman (1:47:01.960)

but I started, and then you're forced to be isolated.

Richard Craib (1:47:06.280)

Right.

Lex Fridman (1:47:07.120)

And so it's actually kind of nice

Richard Craib (1:47:08.880)

because it's very depressing.

Lex Fridman (1:47:12.440)

And then I've heard stories of, I think it's Sean Parker.

Richard Craib (1:47:15.800)

He had like all these diseases as a child

Lex Fridman (1:47:18.240)

and he had to like just stay in bed for years.

Lex Fridman (1:47:20.640)

And then he like made Napster.

Lex Fridman (1:47:23.320)

It's like pretty cool.

Lex Fridman (1:47:24.840)

So yeah, I had about 15 days of this recently,

Lex Fridman (1:47:27.680)

just last month.

Lex Fridman (1:47:28.520)

And it feels like it did shock me

Lex Fridman (1:47:30.400)

into a new kind of energy and ambition.

Richard Craib (1:47:34.640)

Were there moments when you were just like terrified

Lex Fridman (1:47:37.960)

at the combination of loneliness?

Lex Fridman (1:47:39.680)

And like, you know, the thing about COVID is like,

Lex Fridman (1:47:43.120)

there's some degree of uncertainty.

Richard Craib (1:47:45.480)

Like it feels like it's a new thing, a new monster

Lex Fridman (1:47:48.400)

that's arrived on this earth.

Lex Fridman (1:47:50.920)

And so, you know, dealing with it alone,

Lex Fridman (1:47:54.280)

a lot of people are dying.

Richard Craib (1:47:55.680)

It's like wondering like.

Lex Fridman (1:47:57.240)

Yeah, you do wonder, I mean, for sure.

Lex Fridman (1:47:59.200)

And then there are the even new strains in South Africa,

Lex Fridman (1:48:03.160)

which is where I was.

Lex Fridman (1:48:04.000)

And maybe the new strain had some interaction

Lex Fridman (1:48:06.920)

with my genes and I'm just gonna die.

Lex Fridman (1:48:09.520)

But ultimately it was liberating somehow.

Lex Fridman (1:48:11.440)

I loved it.

Richard Craib (1:48:12.600)

Oh, I loved that I got out of it.

Lex Fridman (1:48:15.440)

Okay.

Richard Craib (1:48:16.280)

Because it also affects your mind.

Lex Fridman (1:48:17.120)

You get confused, you get confusion

Lex Fridman (1:48:18.680)

and kind of a lot of fatigue

Lex Fridman (1:48:21.360)

and you can't do your usual tricks

Richard Craib (1:48:23.360)

of psyching yourself out of it.

Lex Fridman (1:48:24.800)

So, you know, sometimes it's like, oh man, I feel tired.

Richard Craib (1:48:27.440)

Okay, I'm just gonna go have coffee and then I'll be fine.

Lex Fridman (1:48:29.720)

It's like, now it's like, I feel tired.

Richard Craib (1:48:31.440)

I don't even wanna get out of bed to get coffee

Lex Fridman (1:48:33.440)

because I feel so tired.

Lex Fridman (1:48:34.680)

And then you have to confront,

Lex Fridman (1:48:37.120)

there's no like quick fix cure and you're trapped at home.

Lex Fridman (1:48:40.840)

But that, so now you have this little thing

Lex Fridman (1:48:43.320)

that happened to you that was a reminder

Richard Craib (1:48:44.920)

that you're mortal and you get to carry that flag

Lex Fridman (1:48:48.160)

in trying to create something special in this world.

Lex Fridman (1:48:53.280)

Right?

Lex Fridman (1:48:54.120)

With Numerai.

Richard Craib (1:48:54.960)

Listen, this was like one of my favorite conversation

Lex Fridman (1:48:58.320)

because the way you think about this world of money

Lex Fridman (1:49:03.280)

and just this world in general is so clear

Lex Fridman (1:49:05.200)

and you're able to explain it so eloquently.

Richard Craib (1:49:08.680)

Richard, that was really fun.

Lex Fridman (1:49:09.960)

Really appreciate you talking to me.

Richard Craib (1:49:11.000)

Thank you.

Lex Fridman (1:49:11.840)

Thank you.

Richard Craib (1:49:13.160)

Thanks for listening to this conversation with Richard Crave

Lex Fridman (1:49:15.680)

and thank you to our sponsors,

Richard Craib (1:49:17.800)

Audible Audio Books, Trial Labs,

Lex Fridman (1:49:20.240)

Machine Learning Company, Blinkist app

Richard Craib (1:49:22.840)

that summarizes books and Athletic Greens

Lex Fridman (1:49:25.640)

all in one nutrition drink.

Richard Craib (1:49:27.560)

Click the sponsor links to get a discount

Lex Fridman (1:49:29.840)

and to support this podcast.

Lex Fridman (1:49:32.080)

And now let me leave you with some words from Warren Buffett.

Lex Fridman (1:49:36.640)

Games are won by players who focus on the playing field,

Richard Craib (1:49:40.600)

not by those whose eyes are glued to the scoreboard.

Lex Fridman (1:49:44.600)

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

Richard Craib (20:00.520)

they affect kind of all the other shorts too.

Lex Fridman (20:02.920)

And suddenly brokers are saying things like,

Richard Craib (20:05.680)

you need to put up more collateral.

Lex Fridman (20:07.400)

So we had a short, it wasn't GameStop luckily,

Richard Craib (20:12.080)

it was Blackberry and it went up like 100% in a day.

Lex Fridman (20:15.440)

It was one of these meme stocks, super bad company.

Lex Fridman (20:17.640)

The AIs don't like it, okay?

Lex Fridman (20:19.720)

The AIs think it's going down.

Lex Fridman (20:21.240)

What's a meme stock?

Lex Fridman (20:22.400)

A meme stock is kind of a new term for these stocks

Richard Craib (20:26.360)

that catch memetic momentum on Reddit.

Lex Fridman (20:32.080)

And so the meme stocks were GameStop, the biggest one,

Richard Craib (20:36.080)

GameStonk, as Elon calls it, AMC.

Lex Fridman (20:39.440)

And Blackberry was one, Nokia was one.

Lex Fridman (20:44.880)

So these are high short interest stocks as well.

Lex Fridman (20:47.960)

So these are targeted stocks.

Richard Craib (20:50.040)

Some people say, oh, isn't it adorable

Lex Fridman (20:53.360)

that these people are investing money

Lex Fridman (20:56.280)

in these companies that are nostalgic?

Lex Fridman (20:59.440)

It's like, you go into the AMC movie theater,

Richard Craib (21:01.640)

it's like nostalgic.

Lex Fridman (21:02.480)

It's like, no, it's not why they're doing it.

Richard Craib (21:05.040)

It's that they had a lot of short interest.

Lex Fridman (21:07.000)

That was the main thing.

Lex Fridman (21:08.320)

And so there were high chance of short squeeze.

Lex Fridman (21:11.520)

In saying, I would love to see an alternate history,

Lex Fridman (21:14.040)

do you have a sense that that,

Lex Fridman (21:17.080)

what is your prediction of what that history would look like?

Richard Craib (21:19.840)

Well, you wouldn't have needed very many more days

Lex Fridman (21:23.040)

of that kind of chaos to hurt hedge funds.

Richard Craib (21:28.440)

I think it's underrated how damaging it could have been.

Lex Fridman (21:32.840)

Because when your shorts go up,

Richard Craib (21:38.440)

your collateral requirements for them go up.

Lex Fridman (21:40.520)

It's similar to Robinhood.

Richard Craib (21:41.880)

Like we have a prime broker that says, said to us,

Lex Fridman (21:46.080)

you need to put up like $40 per $100 of short exposure.

Lex Fridman (21:51.480)

And then the next day they said,

Lex Fridman (21:52.560)

actually you have to put up all of it, 100%.

Lex Fridman (21:56.480)

And we were like, what?

Lex Fridman (21:58.960)

But if that happens to all the short,

Richard Craib (22:02.440)

all the commonly held hedge fund shorts,

Lex Fridman (22:05.560)

because they're all kind of holding the same things.

Richard Craib (22:08.320)

If that happens, not only do you have to cover the short,

Lex Fridman (22:13.320)

which means you're buying the bad companies,

Richard Craib (22:16.160)

you need to sell your good companies

Lex Fridman (22:18.720)

in order to cover the short.

Lex Fridman (22:20.880)

So suddenly like all the good companies,

Lex Fridman (22:23.640)

all the ones that the hedge funds like are coming down

Lex Fridman (22:26.200)

and all the ones that the hedge funds hate are going up

Lex Fridman (22:29.920)

in a cascading way.

Lex Fridman (22:33.920)

So I believe that if you could have had a few more days

Lex Fridman (22:37.120)

of GameStop doubling, AMC doubling,

Richard Craib (22:40.840)

you would have had more and more hedge fund deleveraging.

Lex Fridman (22:45.480)

But so hedge funds, I mean, they get a lot of shit,

Lex Fridman (22:48.360)

but they, do you have a sense

Lex Fridman (22:51.320)

that they do some good for the world?

Richard Craib (22:53.480)

I mean, ultimately, so, okay.

Lex Fridman (22:55.320)

First of all, Wall Street bets itself

Richard Craib (22:57.040)

is a kind of distributed hedge fund.

Lex Fridman (22:59.320)

Numeri is a kind of hedge fund.

Lex Fridman (23:01.520)

So I got, hedge fund is a very broad category.

Lex Fridman (23:04.000)

I mean, like if some of those were destroyed,

Lex Fridman (23:07.200)

would that be good for the world?

Lex Fridman (23:08.880)

Or would there be coupled

Richard Craib (23:11.960)

with the destroying the evil shorting,

Lex Fridman (23:15.160)

would there be just a lot of pain

Lex Fridman (23:16.640)

in terms of investment in good companies?

Lex Fridman (23:18.960)

Yeah, a thing I like to tell people

Richard Craib (23:21.360)

if they hate hedge funds is,

Lex Fridman (23:23.680)

I don't think you want to rerun American economic history

Richard Craib (23:27.960)

without hedge funds.

Lex Fridman (23:29.680)

So on mass they're, yeah, they're good.

Richard Craib (23:34.000)

Yeah, you really wouldn't want to.

Lex Fridman (23:36.040)

Because hedge funds are kind of like picking up,

Richard Craib (23:38.640)

they're making liquidity, right, in stocks.

Lex Fridman (23:41.200)

And so if you love venture capitalists,

Richard Craib (23:45.520)

they're investing in new technology, it's so good.

Lex Fridman (23:48.320)

You have to also kind of like hedge funds

Richard Craib (23:50.560)

because they're the reason venture capitalists exist

Lex Fridman (23:54.040)

because their companies can have a liquidity event

Richard Craib (23:56.520)

when they go to the public markets.

Lex Fridman (23:58.960)

So it's kind of essential that we have them.

Richard Craib (24:01.960)

There are many different kinds of them.

Lex Fridman (24:04.440)

I believe we could maybe get away

Richard Craib (24:06.240)

with only having an AI hedge fund.

Lex Fridman (24:11.560)

But we don't necessarily need

Richard Craib (24:13.120)

these evil billions type hedge funds

Lex Fridman (24:15.280)

that make the media and try to kill companies.

Lex Fridman (24:18.600)

But we definitely need hedge funds.

Lex Fridman (24:20.160)

Maybe from your perspective,

Richard Craib (24:21.760)

because you run such an organization

Lex Fridman (24:27.360)

and Vlad, the CEO of Robinhood,

Richard Craib (24:30.760)

sort of had to make decisions really quickly,

Lex Fridman (24:32.720)

probably had to wake up

Richard Craib (24:33.680)

in the middle of the night kind of thing.

Lex Fridman (24:36.920)

And he also had a conversation with Elon Musk on Clubhouse,

Richard Craib (24:40.680)

which I just signed up for.

Lex Fridman (24:42.360)

It was a fascinating,

Richard Craib (24:43.560)

one of the great journalistic performances of our time

Lex Fridman (24:47.480)

with Elon Musk.

Richard Craib (24:49.080)

Pull a surprise for Elon.

Lex Fridman (24:51.080)

How hilarious would it be if he gets a pull a surprise?

Lex Fridman (24:55.880)

And then his Wikipedia would be like journalist

Lex Fridman (24:58.240)

and part time entrepreneur.

Richard Craib (25:00.360)

Business magnate.

Lex Fridman (25:01.200)

Business magnate.

Richard Craib (25:02.840)

As you know, I don't know if you can comment

Lex Fridman (25:05.520)

on any aspects of that,

Lex Fridman (25:06.720)

but like, if you were Vlad,

Lex Fridman (25:08.600)

how would you do things differently?

Lex Fridman (25:10.080)

What are your thoughts about his interaction with Elon?

Lex Fridman (25:13.320)

How he should have played it differently?

Richard Craib (25:15.640)

Like, I guess there's a lot of aspects to this interaction.

Lex Fridman (25:19.400)

One is about transparency.

Richard Craib (25:20.920)

Like how much do you want to tell people

Lex Fridman (25:23.240)

about really what went down?

Richard Craib (25:24.880)

There's NDAs potentially involved.

Lex Fridman (25:27.320)

How much in private do you want to push back

Lex Fridman (25:32.720)

and say, no, fuck you, to centralize power?

Lex Fridman (25:36.240)

Whatever the phone calls you're getting,

Richard Craib (25:37.560)

which I'm sure he was getting some kind of phone calls

Lex Fridman (25:40.200)

that might not be contractual.

Richard Craib (25:42.080)

Like it's not contracts that are forcing him,

Lex Fridman (25:44.160)

but he was being, what do you call it?

Richard Craib (25:47.400)

Like pressured to behave in certain kinds of ways

Lex Fridman (25:49.960)

from all kinds of directions.

Lex Fridman (25:51.560)

Like what do you take from this whole situation?

Lex Fridman (25:55.760)

I was very excited to see Vlad's response.

Richard Craib (25:58.320)

I mean, it's pretty cool to have him talk to Elon.

Lex Fridman (26:00.920)

And one of the things that like struck me

Richard Craib (26:02.840)

in the first like few seconds of Vlad speaking was like,

Lex Fridman (26:06.320)

I was like, is Vlad like a boomer?

Richard Craib (26:10.920)

Like, but hear me out.

Lex Fridman (26:13.120)

Like he seemed like a 55 year old man

Richard Craib (26:16.200)

talking to a 20 year old.

Lex Fridman (26:18.080)

Elon was like the 20 year old.

Lex Fridman (26:19.920)

And he's like the 55 year old man.

Lex Fridman (26:21.920)

You can see why Citadel are NMR buddies, right?

Richard Craib (26:25.280)

Like you can, you can see why.

Lex Fridman (26:27.800)

It's like, this is a nice, it's not a bad thing.

Richard Craib (26:30.200)

It's like, he's got a respectable professional attitude.

Lex Fridman (26:35.400)

Well, he also tried to do like a jokey thing.

Richard Craib (26:38.120)

Like, no, we're not being ageist here.

Lex Fridman (26:41.360)

Boomer, but like a 60 year old CEO of Bank of America

Richard Craib (26:46.880)

would try to make a joke for the kids.

Lex Fridman (26:48.400)

That's what Vlad's like.

Lex Fridman (26:49.840)

Yeah, I was like, what is this?

Lex Fridman (26:51.760)

This guy's like, what is he, 30?

Richard Craib (26:53.680)

Yeah.

Lex Fridman (26:55.040)

And I'm like, this is weird.

Richard Craib (26:56.840)

Yeah.

Lex Fridman (26:58.040)

But I think, and maybe that's also what I like

Richard Craib (27:00.480)

about Elon's kind of influence on American business

Lex Fridman (27:03.520)

is like, he's super like anti the professional.

Lex Fridman (27:07.280)

Like why say, you know, a hundred words about nothing?

Lex Fridman (27:13.120)

And so I liked how he was cutting in and saying,

Lex Fridman (27:15.320)

Vlad, what do you mean?

Lex Fridman (27:16.160)

Spill the beans, bro.

Richard Craib (27:17.440)

Yeah, so you don't have to be courteous.

Lex Fridman (27:19.520)

It's like the first principles thinking.

Lex Fridman (27:21.040)

It's like, what the hell happened?

Lex Fridman (27:23.200)

Yes.

Richard Craib (27:24.040)

Let's just talk like normal people.

Lex Fridman (27:26.600)

The problem of course is, you know, for Elon,

Lex Fridman (27:31.600)

it's cost them, what is it?

Lex Fridman (27:33.320)

Tens of millions of dollars is tweeting like that.

Lex Fridman (27:36.520)

But perhaps it's a worthy price to pay

Lex Fridman (27:39.680)

because ultimately there's something magical

Richard Craib (27:42.160)

about just being real and honest

Lex Fridman (27:45.840)

and just going off the cuff and making the mistakes

Lex Fridman (27:48.320)

and paying for them, but just being real.

Lex Fridman (27:50.800)

And then moments like this,

Richard Craib (27:52.880)

that was an opportunity for Vlad to be that.

Lex Fridman (27:55.080)

And it felt like he wasn't.

Lex Fridman (27:57.400)

Do you think we'll ever find out what really went down

Lex Fridman (28:02.600)

if there was something shady underneath it all?

Richard Craib (28:05.200)

Yeah, I mean, it would be sad if nothing shady happened,

Lex Fridman (28:09.800)

but his presence made it shady.

Richard Craib (28:12.120)

Sometimes I feel like that would mark Zuckerberg,

Lex Fridman (28:15.240)

the CEO of Facebook.

Richard Craib (28:18.000)

Sometimes I feel like, yeah,

Lex Fridman (28:19.040)

there's a lot of shitty things that Facebook is doing,

Lex Fridman (28:22.280)

but sometimes I think he makes it look worse

Lex Fridman (28:25.120)

by the way he presents himself about those things.

Richard Craib (28:28.440)

Like I honestly think that a large amount of people

Lex Fridman (28:31.400)

at Facebook just have a huge unstable chaotic system

Lex Fridman (28:36.800)

and they're all, not all, but mass are trying to do good

Lex Fridman (28:40.760)

with this chaotic system.

Lex Fridman (28:42.320)

But the presentation is like,

Lex Fridman (28:43.760)

it sounds like there's a lot of back room conversations

Richard Craib (28:47.680)

that are trying to manipulate people.

Lex Fridman (28:49.640)

And there's something about the realness that Elon has

Richard Craib (28:53.400)

that it feels like CEO should have

Lex Fridman (28:55.720)

and Vlad had that opportunity.

Richard Craib (28:57.200)

I think Mark Zuckerberg had that too when he was younger.

Lex Fridman (28:59.840)

Younger.

Lex Fridman (29:00.680)

And somebody said, you gotta be more professional, man.

Lex Fridman (29:02.800)

You can't say, you know, lol to an interview.

Lex Fridman (29:06.880)

And then suddenly he became like this distant person

Lex Fridman (29:10.720)

that was hot.

Richard Craib (29:11.560)

Like you'd rather have him make mistakes,

Lex Fridman (29:13.400)

but be honest than be like professional

Lex Fridman (29:16.080)

and never make mistakes.

Lex Fridman (29:18.000)

Yeah, one of the difficult hires I think

Richard Craib (29:22.000)

is like marketing people or like PR people

Lex Fridman (29:25.680)

is you have to hire people that get the fact

Richard Craib (29:28.320)

that you can say lol on an interview.

Lex Fridman (29:31.000)

Or, you know, take risks as opposed to what the PR,

Richard Craib (29:37.200)

I've talked to quite a few big CEOs

Lex Fridman (29:39.720)

and the people around them are trying to constantly

Lex Fridman (29:44.720)

minimize risk of like, what if he says the wrong thing?

Lex Fridman (29:48.000)

What if she says the wrong thing?

Richard Craib (29:49.480)

It's like, what, be careful.

Lex Fridman (29:51.360)

It's constantly like, ooh, like, I don't know.

Lex Fridman (29:53.720)

And there's this nervous energy that builds up over time

Lex Fridman (29:56.560)

with larger and larger teams where the whole thing,

Richard Craib (29:59.720)

like I visited YouTube, for example.

Lex Fridman (30:01.520)

Everybody I talked at YouTube, incredible engineering

Lex Fridman (30:05.080)

and incredible system, but everybody's scared.

Lex Fridman (30:08.520)

Like, let's be honest about this like madness

Richard Craib (30:13.080)

that we have going on of huge amounts of video

Lex Fridman (30:15.840)

that we can't possibly ever handle.

Richard Craib (30:17.640)

There's a bunch of hate on YouTube.

Lex Fridman (30:19.680)

There's this chaos of comments,

Richard Craib (30:21.480)

bunch of conspiracy theories, some of which might be true.

Lex Fridman (30:24.280)

And then just like this mess that we're dealing with

Lex Fridman (30:27.360)

and it's exciting, it's beautiful.

Lex Fridman (30:30.160)

It's a place where like democratizes education,

Richard Craib (30:32.840)

all that kind of stuff.

Lex Fridman (30:34.200)

And instead they're all like sitting in like,

Richard Craib (30:36.920)

trying to be very polite and saying like,

Lex Fridman (30:38.720)

well, we just want to improve the health of our platforms.

Richard Craib (30:41.640)

Like, it's like this discussion like,

Lex Fridman (30:44.720)

all right, man, let's just be real.

Richard Craib (30:46.040)

Let's both advertise how amazing this fricking thing is,

Lex Fridman (30:50.480)

but also to say like, we don't know what we're doing.

Richard Craib (30:53.200)

We have all these Nazis posting videos on YouTube.

Lex Fridman (30:56.040)

We don't know how to like handle it.

Lex Fridman (30:58.040)

And just being real like that,

Lex Fridman (31:00.000)

because I suppose that's just a skill.

Richard Craib (31:02.240)

Maybe it can't be taught, but over time,

Lex Fridman (31:05.080)

the whatever the dynamics of the company is,

Richard Craib (31:06.960)

it does seem like Zuckerberg and others get worn down.

Lex Fridman (31:09.840)

They just get tired.

Richard Craib (31:10.960)

Yeah.

Lex Fridman (31:11.920)

They get tired of.

Richard Craib (31:12.760)

Not being real.

Lex Fridman (31:13.720)

Of not being real, which is sad.

Lex Fridman (31:16.360)

So let's talk about Numerai,

Lex Fridman (31:17.720)

which is an incredible company system idea, I think,

Lex Fridman (31:24.680)

but good place to start.

Lex Fridman (31:26.520)

What is Numerai and how does it work?

Lex Fridman (31:30.600)

So Numerai is the first hedge fund

Lex Fridman (31:32.960)

that gives away all of its data.

Lex Fridman (31:35.160)

So this is like probably the last thing

Lex Fridman (31:37.640)

a hedge fund would do, right?

Lex Fridman (31:39.160)

Why would we give away a data?

Lex Fridman (31:40.280)

It's like giving away your edge.

Lex Fridman (31:42.960)

But the reason we do it is because we're looking for people

Lex Fridman (31:46.840)

to model our data.

Lex Fridman (31:49.080)

And the way we do it is by obfuscating the data.

Lex Fridman (31:52.840)

So when you look at Numerai's data

Richard Craib (31:55.360)

that you can download for free,

Lex Fridman (31:57.520)

it just looks like a million rows

Richard Craib (32:00.440)

of numbers between zero and one.

Lex Fridman (32:02.920)

And you have no idea what the columns mean,

Lex Fridman (32:05.280)

but you do know that if you're good at machine learning

Lex Fridman (32:09.120)

or have done regressions before,

Richard Craib (32:11.520)

you know that I can still find patterns in this data,

Lex Fridman (32:14.640)

even though I don't know what the features mean.

Lex Fridman (32:17.560)

And the data itself is a time series data.

Lex Fridman (32:20.480)

And even though it's obfuscated, anonymized,

Lex Fridman (32:24.480)

what is the source data like approximately?

Lex Fridman (32:26.920)

What are we talking about?

Lex Fridman (32:27.960)

So we are buying data from lots of different data vendors

Lex Fridman (32:32.360)

and they would also never want us to share that data.

Lex Fridman (32:36.600)

So we have strict contracts with them.

Lex Fridman (32:38.560)

So we only can,

Lex Fridman (32:41.200)

but that's the kind of data you could never buy yourself

Lex Fridman (32:43.480)

unless you had maybe a million dollars a year

Richard Craib (32:46.760)

of budget to buy data.

Lex Fridman (32:48.760)

So what's happened with the hedge fund industry

Richard Craib (32:50.600)

is you have a lot of talented people

Lex Fridman (32:54.600)

who used to be able to trade and still can trade,

Lex Fridman (32:59.640)

but now they have such a data disadvantage,

Lex Fridman (33:02.200)

it would never make sense for them to trade themselves.

Lex Fridman (33:07.120)

But Numerai, by giving away this obfuscated data,

Lex Fridman (33:09.640)

we can give them a really, really high quality data set

Richard Craib (33:12.240)

that would otherwise be very expensive.

Lex Fridman (33:14.800)

And they can use whatever new machine learning technique

Richard Craib (33:18.680)

they want to find patterns in that data

Lex Fridman (33:22.040)

that we can use in our hedge fund.

Lex Fridman (33:24.000)

And so how much variety is there in underlying data?

Lex Fridman (33:27.080)

We're talking about,

Richard Craib (33:29.560)

I apologize if I'm using the wrong terms,

Lex Fridman (33:31.080)

but one is just like the stock price.

Richard Craib (33:34.360)

The other, there's like options and all that kind of stuff,

Lex Fridman (33:36.840)

like the, what are they called, order books or whatever.

Richard Craib (33:41.520)

Is there maybe other totally unrelated

Lex Fridman (33:45.320)

directly to the stock market data,

Lex Fridman (33:47.120)

like natural language as well, all that kind of stuff?

Lex Fridman (33:51.320)

Yeah, we were really focused on stock data

Richard Craib (33:55.120)

that's specific to stocks.

Lex Fridman (33:56.560)

So things like you can have like a,

Richard Craib (33:59.840)

every stock has like a PE ratio.

Lex Fridman (34:01.880)

For some stocks, it's not as meaningful,

Lex Fridman (34:03.800)

but every stock has that.

Lex Fridman (34:05.560)

Every stock has one year momentum,

Lex Fridman (34:07.560)

how much they went up in the last year,

Lex Fridman (34:10.680)

but those are very common factors.

Lex Fridman (34:12.520)

But we try to get lots and lots of those factors

Lex Fridman (34:15.400)

that we have for many, many years,

Richard Craib (34:17.040)

like 15, 20 years history.

Lex Fridman (34:21.520)

And then the setup of the problem is commonly in quant

Richard Craib (34:25.680)

called like cross sectional global equity.

Lex Fridman (34:28.120)

You're not really trying to say,

Richard Craib (34:29.720)

I believe this stock will go up.

Lex Fridman (34:31.920)

You're trying to say the relative position of this stock

Richard Craib (34:36.240)

in feature space makes it not a bad buy in a portfolio.

Lex Fridman (34:41.720)

So it captures some period of time

Lex Fridman (34:44.480)

and you're trying to find the patterns,

Lex Fridman (34:46.040)

the dynamics captured by the data of that period of time

Richard Craib (34:49.920)

in order to make short term predictions

Lex Fridman (34:51.760)

about what's going to happen.

Richard Craib (34:53.160)

Yeah, so our predictions are also not that short.

Lex Fridman (34:55.720)

We're not really caring about things like order books

Lex Fridman (34:58.760)

and tech data, not high frequency at all.

Lex Fridman (35:02.360)

We're actually holding things for quite a bit longer.

Lex Fridman (35:05.080)

So our prediction time horizon is about one month.

Lex Fridman (35:07.800)

We end up holding stocks

Richard Craib (35:08.840)

for maybe like three or four months.

Lex Fridman (35:10.760)

So I kind of believe that's a little bit more

Richard Craib (35:12.440)

like investing than kind of plumbing,

Lex Fridman (35:17.840)

like to go long a stock that's mispriced on one exchange

Lex Fridman (35:21.840)

and short on another exchange, that's just arbitrage.

Lex Fridman (35:26.120)

But what we're trying to do is really know something more

Richard Craib (35:30.000)

about the longer term future of the stock.

Lex Fridman (35:31.920)

Yeah, so from the patterns,

Richard Craib (35:33.320)

from these like periods of time series data,

Lex Fridman (35:37.000)

you're trying to understand something fundamental

Richard Craib (35:39.480)

about the stock, not like about deep value,

Lex Fridman (35:43.280)

about like it's big in the context of the market,

Richard Craib (35:46.360)

is it underpriced, overpriced, all that kind of stuff.

Lex Fridman (35:48.880)

So like, this is about investing.

Richard Craib (35:50.800)

It's not about like, just like you said,

Lex Fridman (35:52.840)

high frequency trading,

Richard Craib (35:54.640)

which I think is a fascinating open question

Lex Fridman (35:57.040)

from a machine learning perspective,

Lex Fridman (35:58.480)

but just to like sort of build on that.

Lex Fridman (36:00.880)

So you've anonymized the data

Lex Fridman (36:02.800)

and now you're giving away the data.

Lex Fridman (36:05.240)

And then now anyone can try to build algorithms

Richard Craib (36:12.000)

that make investing decisions on top of that data

Lex Fridman (36:14.760)

or predictions on the top of that data.

Richard Craib (36:16.440)

Exactly.

Lex Fridman (36:17.280)

And so that's, what does that look like?

Lex Fridman (36:21.400)

What's the goal of that?

Lex Fridman (36:22.240)

What are the underlying principles of that?

Lex Fridman (36:24.440)

So the first thing is,

Lex Fridman (36:26.320)

we could obviously model that data in house, right?

Richard Craib (36:29.080)

We can make an XGBoost model on the data

Lex Fridman (36:33.320)

and that would be quite good too.

Lex Fridman (36:36.200)

But what we're trying to do is by opening it up

Lex Fridman (36:40.240)

and letting anybody participate,

Richard Craib (36:43.040)

we can do quite a lot better than if we modeled it ourselves

Lex Fridman (36:47.560)

and a lot better on the stock market

Richard Craib (36:50.120)

doesn't need to be very much.

Lex Fridman (36:52.120)

Like it really matters the difference

Richard Craib (36:54.000)

between if you can make 10 and 12%

Lex Fridman (36:56.520)

in an equity market neutral hedge fund

Richard Craib (36:58.560)

because usually you're charging 2% fees.

Lex Fridman (37:02.800)

So if you can do 2% better,

Richard Craib (37:05.080)

that's like all your fees, it's worth it.

Lex Fridman (37:07.600)

So we're trying to make sure

Richard Craib (37:09.440)

that we always have the best possible model

Lex Fridman (37:11.360)

as new machine learning libraries come out,

Richard Craib (37:13.320)

new techniques come out,

Lex Fridman (37:14.960)

they get automatically synthesized.

Richard Craib (37:16.760)

Like if there's a great paper on supervised learning,

Lex Fridman (37:19.840)

someone on Numerai will figure out

Lex Fridman (37:21.840)

how to use it on Numerai's data.

Lex Fridman (37:23.480)

And is there an ensemble of models going on

Richard Craib (37:28.960)

or is it more towards kind of like one or two or three

Lex Fridman (37:33.440)

like best performing models?

Lex Fridman (37:35.360)

So the way we decide on how to weight

Lex Fridman (37:37.760)

all of the predictions together

Richard Craib (37:40.360)

is by how much the users are staking on them.

Lex Fridman (37:44.720)

How much of the cryptocurrency

Richard Craib (37:46.280)

that they're putting behind their models.

Lex Fridman (37:48.280)

So they're saying, I believe in my model.

Richard Craib (37:51.280)

You can trust me because I'm gonna put skin in the game.

Lex Fridman (37:55.320)

And so we can take the stake weighted predictions

Richard Craib (37:57.960)

from all our users, add those together,

Lex Fridman (38:01.040)

average those together,

Lex Fridman (38:02.240)

and that's a much better model than any one model

Lex Fridman (38:05.920)

in the sum because ensembling a lot of models together

Richard Craib (38:08.800)

is kind of the key thing you need to do in investing too.

Lex Fridman (38:12.400)

Yeah, so you're putting,

Lex Fridman (38:14.520)

so there's a kind of duality from the user,

Lex Fridman (38:16.800)

from the perspective of a machine learning engineer

Richard Craib (38:19.360)

where it's both a competition, just a really interesting,

Lex Fridman (38:22.480)

difficult machine learning problem,

Lex Fridman (38:24.720)

and it's a way to invest algorithmically.

Lex Fridman (38:29.840)

So like, but the way to invest algorithmically

Richard Craib (38:33.800)

also is a way to put skin in the game

Lex Fridman (38:37.360)

that communicates to you that the quality of the algorithm

Lex Fridman (38:42.840)

and also forces you to really be serious

Lex Fridman (38:46.960)

about the models that you build.

Lex Fridman (38:49.440)

So it's like, everything just works nicely together.

Lex Fridman (38:52.320)

Like, I guess one way to say that

Richard Craib (38:54.640)

is the interests are aligned.

Lex Fridman (38:58.000)

Okay, so it's just like poker is not fun

Richard Craib (39:02.320)

when it's like for very low stakes.

Lex Fridman (39:04.760)

The higher the stakes,

Richard Craib (39:05.880)

the more the dynamics of the system

Lex Fridman (39:07.480)

starts playing out correctly.

Richard Craib (39:10.760)

Like as a small side note,

Lex Fridman (39:12.120)

is there something you can say about which kind,

Richard Craib (39:16.480)

looking at the big broad view of machine learning today

Lex Fridman (39:19.240)

or AI, what kind of algorithms seem to do good

Lex Fridman (39:24.640)

in these kinds of competitions at this time?

Lex Fridman (39:27.120)

Is there some universal thing you can say,

Richard Craib (39:29.960)

like neural networks suck,

Lex Fridman (39:32.040)

recurrent neural networks suck, transformers suck,

Richard Craib (39:34.480)

or they're awesome, like old school,

Lex Fridman (39:37.680)

sort of more basic kind of classifiers are better,

Richard Craib (39:40.360)

all that, is there some kind of conclusions

Lex Fridman (39:42.440)

so far that you can say?

Richard Craib (39:44.040)

There is definitely something pretty nice about tree models,

Lex Fridman (39:47.640)

like XGBoost,

Lex Fridman (39:50.160)

and they just seem to work pretty nicely

Lex Fridman (39:53.120)

on this type of data.

Lex Fridman (39:54.880)

So out of the box,

Lex Fridman (39:56.320)

if you're trying to come a hundredth

Richard Craib (39:59.480)

in the competition, in the tournament,

Lex Fridman (40:01.400)

maybe you would try to use that.

Lex Fridman (40:04.080)

But what's particularly interesting about the problem

Lex Fridman (40:09.240)

that not many people understand,

Richard Craib (40:11.800)

if you're familiar with machine learning,

Lex Fridman (40:14.680)

this typically will surprise you when you model our data.

Lex Fridman (40:18.400)

So one of the things that you look at in finance

Lex Fridman (40:23.920)

is you don't wanna be too exposed to any one risk.

Richard Craib (40:28.240)

Like, even if the best sector in the world

Lex Fridman (40:33.840)

to invest in over the last 10 years was tech,

Richard Craib (40:37.960)

does not mean you should put all of your money into tech.

Lex Fridman (40:40.760)

So if you train a model,

Richard Craib (40:42.880)

it would say, put all your money in tech, it's super good.

Lex Fridman (40:45.680)

But what you wanna do is actually be very careful

Richard Craib (40:49.040)

of how much of this exposure you have to certain features.

Lex Fridman (40:53.360)

So on Numeri, what a lot of people figure out is,

Richard Craib (40:58.400)

actually, if you train a model on this kind of data,

Lex Fridman (41:02.040)

you wanna somehow neutralize or minimize your exposure

Richard Craib (41:05.400)

to these certain features, which is unusual,

Lex Fridman (41:08.160)

because if you did train a stoplight

Richard Craib (41:11.840)

or stop street detection on computer vision,

Lex Fridman (41:18.080)

your favorite feature, let's say you have an auto encoder

Lex Fridman (41:21.720)

and it's figuring out, okay, it's gotta be red

Lex Fridman (41:23.320)

and it's gotta be white,

Richard Craib (41:25.040)

that's the last thing you wanna reduce your exposure to.

Lex Fridman (41:30.280)

Why would you reduce your exposure

Lex Fridman (41:31.440)

to the thing that's helping your model the most?

Lex Fridman (41:34.520)

And that's actually this counterintuitive thing

Richard Craib (41:36.360)

you have to do with machine learning on financial data.

Lex Fridman (41:38.600)

So reducing your exposure

Richard Craib (41:41.760)

would help you generalize the things that are...

Lex Fridman (41:44.040)

So basically, a financial data has a large amount

Richard Craib (41:48.600)

of patterns that appeared in the past

Lex Fridman (41:51.560)

and also a large amount of patterns

Richard Craib (41:53.240)

that have not appeared in the past.

Lex Fridman (41:55.040)

And so like in that sense,

Richard Craib (41:56.160)

you have to reduce the exposure to red lights,

Lex Fridman (41:59.880)

to the color red.

Richard Craib (42:02.440)

That's interesting, but how much of this is art

Lex Fridman (42:05.080)

and how much of it is science from your perspective so far

Richard Craib (42:09.080)

in terms of as you start to climb

Lex Fridman (42:11.080)

from the 100th position to the 95th in the competition?

Richard Craib (42:16.000)

Yeah, well, if you do make yourself super exposed

Lex Fridman (42:20.360)

to one or two features,

Richard Craib (42:23.200)

you can have a lot of volatility

Lex Fridman (42:25.680)

when you're playing Numerai.

Richard Craib (42:26.920)

You could maybe very rapidly rise to be high

Lex Fridman (42:31.200)

if you were getting lucky.

Richard Craib (42:33.000)

Yes.

Lex Fridman (42:34.240)

And that's a bit like the stock market.

Richard Craib (42:36.120)

Sure, take on massive risk exposure,

Lex Fridman (42:38.800)

put all your money into one stock

Lex Fridman (42:41.040)

and you might make 100%,

Lex Fridman (42:43.640)

but it doesn't in the long run work out very well.

Lex Fridman (42:47.680)

And so the best users are trying to stay high

Lex Fridman (42:53.520)

for as long as possible,

Richard Craib (42:55.920)

not necessarily try to be first for a little bit.

Lex Fridman (43:00.040)

So me, a developer, machine learning researcher,

Lex Fridman (43:04.840)

how do I, Lex Friedman, participate in this competition

Lex Fridman (43:07.920)

and how do others,

Richard Craib (43:09.400)

which I'm sure there'll be a lot of others

Lex Fridman (43:10.840)

interested in participating in this competition,

Lex Fridman (43:12.920)

what are, let's see, there's like a million questions,

Lex Fridman (43:15.720)

but like first one is how do I get started?

Richard Craib (43:19.640)

Well, you can go to numero.ai,

Lex Fridman (43:22.440)

sign up, download the data.

Lex Fridman (43:24.600)

And on the data is pretty small.

Lex Fridman (43:30.080)

In the data pack you download,

Richard Craib (43:32.000)

there's like an example script,

Lex Fridman (43:33.960)

Python script that just builds a XGBoost model

Richard Craib (43:37.840)

very quickly from the data.

Lex Fridman (43:40.720)

And so in a very short time,

Richard Craib (43:44.000)

you can have an example model.

Lex Fridman (43:46.080)

Is that a particular structure?

Lex Fridman (43:47.680)

Like what, is this model then submitted somewhere?

Lex Fridman (43:50.680)

So there needs to be some kind of structure

Richard Craib (43:52.520)

that communicates with some kind of API.

Lex Fridman (43:54.760)

Like how does the whole,

Lex Fridman (43:56.480)

how does your model, once you've built,

Lex Fridman (43:58.640)

once you create a little baby Frankenstein,

Lex Fridman (44:01.440)

how does it then live in the world?

Lex Fridman (44:02.760)

Okay, well, we want you to keep your baby Frankenstein

Richard Craib (44:05.120)

at home and take care of it.

Lex Fridman (44:06.880)

We don't want it.

Lex Fridman (44:07.920)

So you never upload your model to us.

Lex Fridman (44:10.760)

You always only giving us predictions.

Lex Fridman (44:14.680)

So we never see the code that wrote your model,

Lex Fridman (44:17.160)

which is pretty cool,

Richard Craib (44:18.160)

that our whole hedge fund is built from models

Lex Fridman (44:20.560)

where we've never, ever seen the code.

Lex Fridman (44:23.720)

But it's important for the users because it's their IP,

Lex Fridman (44:27.040)

why do they want to give it to us?

Richard Craib (44:28.000)

That's brilliant.

Lex Fridman (44:28.840)

So they've got it themselves,

Lex Fridman (44:31.040)

but they can basically almost like license

Lex Fridman (44:34.200)

the predictions from that model to us.

Richard Craib (44:36.720)

License the predictions, yeah.

Lex Fridman (44:38.680)

So. Think about it.

Lex Fridman (44:40.160)

What some users do is they set up a compute server

Lex Fridman (44:43.920)

and we call it Numeric Compute.

Richard Craib (44:45.120)

It's like a little AWS kind of image

Lex Fridman (44:47.280)

and you can automate this process.

Lex Fridman (44:50.160)

So we can ping you.

Lex Fridman (44:51.200)

We can be like, we need more predictions now.

Lex Fridman (44:53.000)

And then you send it to us.

Lex Fridman (44:55.520)

Okay, cool.

Lex Fridman (44:56.360)

So that's, is that described somewhere,

Lex Fridman (45:00.080)

like what the preferred is, the AWS,

Richard Craib (45:02.160)

or whether another cloud platform,

Lex Fridman (45:04.880)

is there, I mean, is there sort of specific technical things

Richard Craib (45:07.680)

you want to say that comes to mind

Lex Fridman (45:09.600)

that is a good path for getting started?

Lex Fridman (45:12.800)

So download the data, maybe play around,

Lex Fridman (45:16.040)

see if you can modify the basic algorithm provided

Richard Craib (45:22.160)

in the example.

Lex Fridman (45:25.760)

And then you, what, set up a little server on the AWS

Richard Craib (45:28.600)

that then runs this model and takes pings

Lex Fridman (45:31.840)

and then makes predictions.

Lex Fridman (45:34.120)

And so how does your own money actually come into play

Lex Fridman (45:37.720)

doing the stake of a cryptocurrency?

Richard Craib (45:41.240)

Yeah, so you don't have to stake.

Lex Fridman (45:44.120)

You can start without staking.

Lex Fridman (45:45.480)

And many users might try for months

Lex Fridman (45:49.040)

without staking anything at all

Lex Fridman (45:50.560)

to see if their model works on the real life data, right?

Lex Fridman (45:54.840)

And is not overfit.

Lex Fridman (45:57.080)

But then you can get Numeraire many different ways.

Lex Fridman (46:01.960)

You can buy it on, you can buy some on Coinbase.

Richard Craib (46:04.960)

You can buy some on Uniswap.

Lex Fridman (46:06.760)

You can buy some on Binance.

Lex Fridman (46:09.240)

So what did you say this is?

Lex Fridman (46:11.160)

How do you pronounce it?

Lex Fridman (46:12.080)

So this is the Numeraire cryptocurrency.

Lex Fridman (46:15.040)

Yeah, NMR.

Lex Fridman (46:16.440)

NMR, you just say NMR?

Lex Fridman (46:19.280)

It is technically called Numeraire.

Richard Craib (46:21.800)

Numeraire, I like it.

Lex Fridman (46:23.560)

Yeah, but NMR is simple.

Richard Craib (46:26.280)

NMR, Numeraire.

Lex Fridman (46:27.280)

Okay, so, and you could buy it basically anywhere.

Richard Craib (46:31.960)

Yeah, so it's a bit strange

Lex Fridman (46:33.080)

because sometimes people are like,

Lex Fridman (46:34.400)

is this like pay to play?

Lex Fridman (46:36.160)

Right.

Lex Fridman (46:37.000)

And it's like, yeah, you need to put some money down

Lex Fridman (46:40.800)

to show us you believe in your model.

Lex Fridman (46:42.600)

But weirdly, we're not selling you the,

Lex Fridman (46:45.160)

like you can't buy the cryptocurrency from us.

Richard Craib (46:48.240)

Right.

Lex Fridman (46:49.080)

It's like, it's also, we never,

Richard Craib (46:51.800)

if you do badly, we destroy your cryptocurrency.

Lex Fridman (46:57.440)

Okay, that's not good, right?

Richard Craib (46:58.520)

You don't want it to be destroyed.

Lex Fridman (47:00.000)

But what's good about it is it's also not coming to us.

Richard Craib (47:03.520)

Right.

Lex Fridman (47:04.360)

So it's not like we win when you lose or something,

Richard Craib (47:06.920)

like we're the house.

Lex Fridman (47:08.560)

Like we're definitely on the same team.

Richard Craib (47:10.360)

Yes.

Lex Fridman (47:11.200)

Helping us make a hedge fund that's never been done before.

Richard Craib (47:14.200)

Yeah, so again, interests are aligned.

Lex Fridman (47:15.760)

There's no, there's no tension there at all,

Richard Craib (47:18.360)

which is really fascinating.

Lex Fridman (47:19.920)

You're giving away everything

Lex Fridman (47:21.000)

and then the IP is owned by sort of the code.

Lex Fridman (47:24.400)

You never share the code.

Richard Craib (47:25.680)

That's fascinating.

Lex Fridman (47:27.120)

So since I have you here and you said a hundred,

Richard Craib (47:31.000)

I didn't ask out of how many, so we'll just,

Lex Fridman (47:34.960)

but if I then once you get started

Lex Fridman (47:37.960)

and you find this interesting, how do you then win

Lex Fridman (47:43.640)

or do well, but also how do you potentially try to win

Richard Craib (47:47.280)

if this is something you want to take on seriously

Lex Fridman (47:49.880)

from the machine learning perspective,

Lex Fridman (47:51.200)

not from a financial perspective?

Lex Fridman (47:53.320)

Yeah, I think that first of all,

Richard Craib (47:55.760)

you would want to talk to the community.

Lex Fridman (47:57.240)

People are pretty open.

Richard Craib (47:58.880)

We give out really interesting scripts and ideas

Lex Fridman (48:01.960)

for things you might want to try.

Richard Craib (48:03.600)

And, but you're also going to need a lot of compute probably.

Lex Fridman (48:09.520)

And so some of the best users are, you know,

Richard Craib (48:12.400)

actually the very first time someone won on Numerai,

Lex Fridman (48:15.400)

I would, I wrote them a personal email.

Richard Craib (48:17.080)

It's like, you know, you've won some money.

Lex Fridman (48:18.880)

We're so excited to give you $300.

Lex Fridman (48:21.560)

And then they said, I spend way more on the compute,

Lex Fridman (48:26.040)

but.

Lex Fridman (48:26.880)

So this is fundamentally a machine learning problem first,

Lex Fridman (48:29.200)

I think is this is one of the exciting things.

Richard Craib (48:31.640)

I don't know if we'll, in how many ways we can approach this,

Lex Fridman (48:34.400)

but really this is less about kind of no offense,

Lex Fridman (48:40.200)

but like finance people, finance minded people,

Lex Fridman (48:43.320)

they're also, I'm sure great people,

Lex Fridman (48:45.400)

but it feels like from the community that I've experienced,

Lex Fridman (48:49.640)

these are people who see finance

Richard Craib (48:52.120)

as a fascinating problem space, source of data,

Lex Fridman (48:57.520)

but ultimately they're machine learning people or AI people,

Richard Craib (49:01.240)

which is a very different kind of flavor of community.

Lex Fridman (49:03.520)

And I mean, I should say to that,

Richard Craib (49:07.560)

I'd love to participate in this and I will participate in this

Lex Fridman (49:10.640)

and I'd love to hear from other people.

Richard Craib (49:12.840)

If you're listening to this,

Lex Fridman (49:13.720)

if you're a machine learning person,

Richard Craib (49:15.520)

you should participate in it and tell me,

Lex Fridman (49:17.640)

give me some hints how I can do well at this thing.

Richard Craib (49:21.320)

Cause this boomer, I'm not sure I still got it,

Lex Fridman (49:24.200)

but cause some of it is, it's like a Kaggle competitions.

Richard Craib (49:28.360)

Like some of it is certainly set of ideas,

Lex Fridman (49:33.280)

like research ideas, like fundamental innovation,

Lex Fridman (49:37.400)

but I'm sure some of it is like deeply understanding,

Lex Fridman (49:40.000)

getting like an intuition about the data.

Lex Fridman (49:42.680)

And then like a lot of it will be like figuring out

Lex Fridman (49:45.760)

like what works, like tricks.

Richard Craib (49:47.760)

I mean, you could argue most of deep learning research

Lex Fridman (49:50.040)

is just tricks on top of tricks,

Lex Fridman (49:51.520)

but there's some of it is just the art

Lex Fridman (49:55.880)

of getting to know how to work in a really difficult

Richard Craib (49:58.800)

machine learning problem.

Lex Fridman (50:00.560)

And I think what's important,

Richard Craib (50:02.000)

the important difference with something

Lex Fridman (50:03.520)

like a Kaggle competition,

Richard Craib (50:04.720)

where they'll set up this kind of toy problem

Lex Fridman (50:08.320)

and then there will be an out of sample test,

Richard Craib (50:10.640)

like, Hey, you did well out of sample.

Lex Fridman (50:12.320)

And this is like, okay, cool.

Lex Fridman (50:14.920)

But what's cool with Numeri is the out of sample

Lex Fridman (50:19.120)

is the real life stock market.

Richard Craib (50:22.080)

We don't even know,

Lex Fridman (50:23.400)

like we don't know the onset of the problem.

Richard Craib (50:25.640)

We don't, like you'll have to find out live.

Lex Fridman (50:28.320)

And so we've had users who've like submitted every week

Richard Craib (50:31.640)

for like four years because it's kind of,

Lex Fridman (50:38.280)

we say it's the hardest data science problem

Lex Fridman (50:40.640)

on the planet, right?

Lex Fridman (50:41.840)

And it sounds maybe sounds like maybe

Lex Fridman (50:44.080)

but too much for like a marketing thing,

Lex Fridman (50:45.600)

but it's the hardest because it's the stock market.

Richard Craib (50:48.840)

It's like literally there are like billions of dollars

Lex Fridman (50:51.640)

at stake and like no one's like letting it be inefficient

Richard Craib (50:55.480)

on purpose.

Lex Fridman (50:56.680)

So if you can find something that works at Numeri,

Richard Craib (50:58.760)

you really have something that is like working

Lex Fridman (51:01.760)

on the real stock market.

Richard Craib (51:03.680)

Yeah, because there's like humans involved

Lex Fridman (51:05.720)

in the stock market.

Richard Craib (51:06.560)

I mean, you could argue there might be harder data sets

Lex Fridman (51:09.920)

like maybe predicting the weather,

Richard Craib (51:11.320)

all those kinds of things.

Lex Fridman (51:12.280)

But the fundamental statement here is, which I like,

Richard Craib (51:16.080)

I was thinking like,

Lex Fridman (51:16.920)

is this really the hardest data science problem?

Lex Fridman (51:19.520)

And you start thinking about that,

Lex Fridman (51:21.160)

but ultimately it also boils down to a problem

Richard Craib (51:24.880)

where the data is accessible.

Lex Fridman (51:26.720)

It's made accessible, made really easy and efficient

Richard Craib (51:31.320)

at like submitting algorithms.

Lex Fridman (51:33.960)

So it's not just, you know,

Richard Craib (51:35.560)

it's not about the data being out there, like the weather.

Lex Fridman (51:38.120)

It's about making the data super accessible,

Richard Craib (51:40.480)

making the ability of community around it.

Lex Fridman (51:43.040)

Like this is what ImageNet did.

Richard Craib (51:45.200)

Exactly.

Lex Fridman (51:46.040)

Like it's not just, there's always images.

Richard Craib (51:49.080)

The point is you aggregate them together.

Lex Fridman (51:51.280)

You give it a little title.

Richard Craib (51:52.680)

This is a community and that was one of the hardest,

Lex Fridman (51:56.920)

right, for a time.

Lex Fridman (51:58.960)

And most important data science problems in the world

Lex Fridman (52:03.960)

because it was accessible, because it was made sort of,

Richard Craib (52:08.680)

like there was mechanisms by which like standards

Lex Fridman (52:12.320)

and mechanisms by which you judge your performance,

Richard Craib (52:14.040)

all those kinds of things.

Lex Fridman (52:14.880)

And Numerize actually step up from that.

Richard Craib (52:17.360)

Is there something more you can say about why

Lex Fridman (52:19.680)

from your perspective it's the hardest problem in the world?

Richard Craib (52:24.560)

I mean, you said it's connected to the market.

Lex Fridman (52:26.560)

So if you could find a pattern in the market,

Richard Craib (52:29.320)

that's a really difficult thing to do

Lex Fridman (52:31.200)

because a lot of people are trying to do it.

Richard Craib (52:33.320)

Exactly.

Lex Fridman (52:34.160)

But there's also the biggest one is

Richard Craib (52:37.200)

it's non stationary time series.

Lex Fridman (52:40.480)

We've tried to regularize the data

Lex Fridman (52:42.680)

so you can find patterns by doing certain things

Lex Fridman (52:46.920)

to the features and the target.

Lex Fridman (52:48.600)

But ultimately you're in a space where you don't,

Lex Fridman (52:51.880)

there's no guarantees that the out of sample distributions

Richard Craib (52:55.720)

will conform to any of the training data.

Lex Fridman (52:59.680)

And every single era, which we call on the website,

Richard Craib (53:04.800)

like every single era in the data,

Lex Fridman (53:07.040)

which is like sort of showing you the order of the time.

Richard Craib (53:12.640)

Even the training data has the same dislocations.

Lex Fridman (53:16.320)

And so, yeah, and then there's so many things

Richard Craib (53:22.360)

that you might wanna try.

Lex Fridman (53:25.880)

There's unlimited possible number of models, right?

Lex Fridman (53:30.320)

And so by having it be open,

Lex Fridman (53:37.520)

we can at least search that space.

Richard Craib (53:40.600)

Zooming back out to the philosophical,

Lex Fridman (53:42.440)

you said that Numerai is very much like Wall Street Bets.

Richard Craib (53:46.560)

Is there, I think it'd be interesting

Lex Fridman (53:51.120)

to dig in why you think so.

Richard Craib (53:52.480)

I think you're speaking to the distributed nature of the two

Lex Fridman (53:56.240)

and the power of the people nature of the two.

Lex Fridman (53:59.680)

So maybe can you speak to the similarities

Lex Fridman (54:02.280)

and the differences and in which way is Numerai more powerful

Lex Fridman (54:06.080)

in which way is Wall Street Bets more powerful?

Lex Fridman (54:09.400)

Yeah, this is why the Wall Street Bets story

Richard Craib (54:11.000)

is so interesting to me because it's like,

Lex Fridman (54:12.800)

feels like we're connected.

Lex Fridman (54:15.120)

And looking at how,

Lex Fridman (54:16.880)

just looking at the form of Wall Street Bets,

Richard Craib (54:19.480)

I was talking earlier about how,

Lex Fridman (54:21.480)

how can you make credible claims?

Richard Craib (54:23.480)

You're anonymous.

Lex Fridman (54:24.360)

Okay, well, maybe you can take a screenshot.

Richard Craib (54:26.920)

Or maybe you can upvote someone.

Lex Fridman (54:29.120)

Maybe you can have karma on Reddit.

Lex Fridman (54:31.280)

And those kinds of things make this emerging thing possible.

Lex Fridman (54:35.640)

Numerai, it didn't work at all when we started.

Richard Craib (54:40.120)

It didn't work at all.

Lex Fridman (54:41.680)

Why?

Richard Craib (54:42.560)

People made multiple accounts.

Lex Fridman (54:43.880)

They made really random models

Lex Fridman (54:45.720)

and hoped they would get lucky.

Lex Fridman (54:47.120)

And some of them did.

Richard Craib (54:48.520)

Yes.

Lex Fridman (54:49.440)

Staking was our solution to,

Lex Fridman (54:53.200)

could we make it so that we could trust?

Lex Fridman (54:56.800)

We could know which model people believed in the most.

Lex Fridman (55:00.120)

And we could weight models that had high stake more

Lex Fridman (55:04.600)

and effectively coordinate this group of people

Richard Craib (55:07.840)

to be like, well, actually there's no incentive

Lex Fridman (55:10.160)

to creating bot accounts anymore.

Richard Craib (55:12.560)

Either I stake my accounts,

Lex Fridman (55:14.640)

in which case I should believe in them

Richard Craib (55:16.080)

because I could lose my stake or I don't.

Lex Fridman (55:18.720)

And that's a very powerful thing

Richard Craib (55:20.840)

that having a negative incentive

Lex Fridman (55:23.120)

and a positive incentive can make things a lot better.

Lex Fridman (55:26.840)

And staking is like this,

Lex Fridman (55:28.240)

is this really nice like key thing about blockchain.

Richard Craib (55:31.560)

It's like something special you can do

Lex Fridman (55:33.760)

where they're not even trusting us

Richard Craib (55:35.880)

with their stake in some ways.

Lex Fridman (55:37.560)

They're trusting the blockchain, right?

Lex Fridman (55:40.040)

So the incentives, like you say,

Lex Fridman (55:42.240)

it's about making these perfect incentives

Lex Fridman (55:44.360)

so that you can have coordination to solve one problem.

Lex Fridman (55:47.920)

And nowadays I sleep easy

Richard Craib (55:52.120)

because I have less money in my own hedge fund

Lex Fridman (55:56.680)

than our users are staking on their models.

Richard Craib (56:00.640)

That's powerful.

Lex Fridman (56:01.680)

In some sense, from a human psychology perspective,

Lex Fridman (56:04.280)

it's fascinating that the WallStreetBets worked at all, right?

Lex Fridman (56:08.960)

That amidst that chaos, emerging behavior,

Richard Craib (56:12.920)

like behavior that made sense emerged.

Lex Fridman (56:15.720)

It would be fascinating to think if numerized style staking

Lex Fridman (56:20.160)

could then be transferred to places like Reddit, you know?

Lex Fridman (56:24.880)

And not necessarily for financial investments,

Lex Fridman (56:27.080)

but I wish sometimes people would have to stake something

Lex Fridman (56:34.360)

in the comments they make on the internet.

Richard Craib (56:36.800)

Yeah.

Lex Fridman (56:37.840)

That's the problem with anonymity is like,

Richard Craib (56:40.480)

anonymity is freedom and power

Lex Fridman (56:42.200)

that you don't have to, you can speak your mind,

Lex Fridman (56:44.960)

but it's too easy to just be shitty.

Lex Fridman (56:47.760)

Yeah, exactly.

Lex Fridman (56:49.840)

So this, I mean, you're making me realize

Lex Fridman (56:52.360)

from a profoundly philosophical aspect, numerized staking

Richard Craib (56:58.360)

is a really clean way to solve this problem.

Lex Fridman (57:01.280)

It's a really beautiful way.

Richard Craib (57:02.280)

Of course, it only with Numerai currently works

Lex Fridman (57:05.360)

for a very particular problem, right?

Richard Craib (57:07.760)

Not for human interaction on the internet,

Lex Fridman (57:10.280)

but that's fascinating.

Richard Craib (57:11.560)

Yeah, there's nothing to stop people.

Lex Fridman (57:13.520)

In fact, we've open sourced the code we use for staking

Richard Craib (57:17.280)

in a protocol we call Erasure.

Lex Fridman (57:19.960)

And if Reddit wanted to, they could even use that code

Richard Craib (57:23.560)

to enable staking on our Wall Street pets.

Lex Fridman (57:29.040)

And they're actually researching now,

Richard Craib (57:30.600)

they've had some Ethereum grants

Lex Fridman (57:32.440)

on how could they have more crypto stuff in there

Lex Fridman (57:37.200)

in Ethereum, because wouldn't that be interesting?

Lex Fridman (57:40.360)

Like, imagine you could, instead of seeing a screenshot,

Richard Craib (57:43.920)

like, guys, I promise I will not sell my GameStop.

Lex Fridman (57:49.360)

We're just going to go huge.

Richard Craib (57:50.720)

We're not going to sell at all.

Lex Fridman (57:53.800)

And here is a smart contract, which no one in the world,

Richard Craib (57:58.480)

including me, can undo, that says,

Lex Fridman (58:02.480)

I have staked millions against this claim.

Richard Craib (58:08.200)

That's powerful.

Lex Fridman (58:09.200)

And then what could you do?

Lex Fridman (58:11.240)

And of course, it doesn't have to be millions.

Lex Fridman (58:12.720)

It could be just a very small amount,

Lex Fridman (58:14.560)

but then just a huge number of users doing that kind of stake.

Lex Fridman (58:17.480)

Exactly.

Richard Craib (58:20.160)

That could change the internet.

Lex Fridman (58:21.920)

It would change, and then Wall Street.

Richard Craib (58:23.960)

It would change Wall Street.

Lex Fridman (58:25.240)

They would never have been able to,

Richard Craib (58:26.800)

they would still be short squeezing one day

Lex Fridman (58:29.520)

after the next, every single hedge fund collapsing.

Richard Craib (58:32.680)

If we look into the future, do you think it's possible

Lex Fridman (58:35.360)

that numerae style infrastructure,

Richard Craib (58:39.080)

where AI systems backed by humans are doing the trading,

Lex Fridman (58:44.760)

is what the entirety of the stock market is,

Richard Craib (58:48.200)

or the entirety of the economy, is

Lex Fridman (58:50.240)

run by basically this army of AI systems

Lex Fridman (58:54.240)

with high level human supervision?

Lex Fridman (58:57.360)

Yeah, the thing is that some of them could be bad actors.

Lex Fridman (59:02.640)

Some of the humans?

Lex Fridman (59:03.400)

No, well, these systems could be tricky.

Lex Fridman (59:06.080)

So actually, I once met a hedge fund manager,

Lex Fridman (59:08.800)

and this is kind of interesting.

Richard Craib (59:10.040)

He said, very famous one, and he said,

Lex Fridman (59:14.280)

we can see, sometimes we can see things in the market

Richard Craib (59:17.600)

where we know we can make money, but it will mess shit up.

Lex Fridman (59:22.760)

We know we can make money, but it will mess things up.

Lex Fridman (59:25.360)

And we choose not to do those things.

Lex Fridman (59:28.200)

And on the one hand, maybe this is like, oh, you're

Richard Craib (59:30.160)

being super arrogant.

Lex Fridman (59:31.920)

Of course you can't do this, but maybe he can.

Lex Fridman (59:35.240)

And maybe he really isn't doing things

Lex Fridman (59:38.280)

he knows he could do, but would be pretty bad.

Richard Craib (59:43.560)

Would the Reddit army have that kind of morality or concern

Lex Fridman (59:51.760)

for what they're doing?

Richard Craib (59:53.680)

Probably not, based on what we've seen.

Lex Fridman (59:55.560)

The madness of crowds.

Richard Craib (59:57.600)

There'll be one person that says, hey, maybe,