0:04
Hello everyone and welcome back to conversations with Tyler.
Search this video
0:05
Today I'm here in San Francisco chatting with Michael Nielsen.
0:10
Michael is hard to introduce and also difficult to prepare for because he knows and has done so many different things.
0:17
different things. He's from Australia, has a PhD in physics, has written what is perhaps the best known text or co-authored it on quantum computing, is one of the leaders of the open science movement, has co-authored with Patrick
0:31
Collison on progress in science, has worked at Y Combinator, is an extraordinarily prolific writer, reader, commentator, tweeter, uh mentor to others, mentee, and many other things and currently is thinking about the fragility of civilization and much more. Michael, welcome. Thank you so much, Michael, welcome.
0:48
Thank you so much, Tyler.
0:51
So you were saying there should have been a metaculus on the opening question.
0:55
Why is the universe beautiful to human eyes? Is it selection? I have no idea.
0:59
I mean selection is a is a very attractive uh kind of an idea.
1:03
Um I I mean kind of to think not just instinctively.
1:10
Um but yeah, I I I don't know.
1:11
I Why why why are there simple rules?
1:13
Why do we have simple rules governing the universe?
1:15
In fact, why is simplicity uh and arguably truth somehow associated to to to to to beauty?
1:22
Uh uh physicists tend to assert that this is the case, but but I don't think anybody really knows the reason why.
1:26
How beautiful do we in fact think the universe is?
1:32
So people don't buy paintings of the universe.
1:33
Like people like you might, right?
1:35
But it's not Oh, I have a painting I've I have the the the Hubble Deep Field on my wall, of course.
1:40
But the most expensive paintings are not of the universe.
1:42
They're of people, they're of boating scenes, right?
1:43
I don't think that's really true.
1:45
I mean the James Webb Space Telescope was about I think was $10 billion.
1:48
Um and is arguably a a machine for for producing that kind of uh uh image.
1:56
Uh that's got to be one of the most important sort of image factories uh and most expensive image factories ever made.
2:01
So I'm not sure I I I buy that.
2:03
What's the most beautiful image of the universe?
2:11
The image of the uh we have a sort of a sequence of improved images of the 3° uh microwave uh background is I don't know.
2:22
Is it the most beautiful?
2:24
It's it maybe the most extraordinary.
2:24
It really is sort of a photograph of the universe as a whole.
2:28
Uh you can look at that and it says something about structure out in creation.
2:32
Why do the sounds of the universe not appeal to us so much? Right?
2:35
So it's beautiful visually, but orally, uh you know, we create very complicated things which we call music, which are beautiful.
2:45
Yeah, that's a great question.
2:45
I don't know why music is beautiful. Um Hm.
2:52
People have made attempts.
2:52
You know, there's things like sort of chirp sounds that might be produced near a black hole and sort of ideas like this.
2:59
And you're right, they they tend not to be all that all that beautiful.
3:03
The only ones that I can think of that sort of offhand um you know, it's being produced by evolution.
3:10
Birdsong is beautiful, but we're actually quite closely related to birds.
3:14
Uh so it's maybe not so surprising.
3:14
If I think about things like uh What's his name?
3:19
Run Run Sixsmith, I think his his name is a composer in Toronto, has made these musical pieces based on um the different periods in the solar system.
3:28
So the you the time that the Earth takes the the one year to go around the Sun, but also then Mars and Jupiter and all these.
3:35
And they are noticeably not particularly attractive uh musical pieces.
3:39
So so yeah, I I I that's a good question.
3:43
I wonder if the beauty of light isn't part of the reason for the beauty of the universe.
3:47
So as human beings, maybe we're evolved to be attracted to light.
3:51
It gives you an integrated theory of the beauty of the universe and beauty of paintings.
3:56
Vermeer, a great painter, he's very attractive to people because of how he uses light.
3:59
When you look at the universe, you're typically seeing signatures of light in many cases.
4:03
You look at the Milky Way, right? It's pretty strange.
4:07
I mean, we see in such a tiny, you know, band of wavelengths.
4:12
We're we're really not seeing almost anything.
4:14
We're not seeing into the infrared and the radio, we're not seeing into the ultraviolet and the the X-ray.
4:19
Um uh so a lot of what we view as as beautiful locally, I mean, it's it's got this sort of evolutionary explanation again.
4:27
Why the large-scale structure um is beautiful.
4:34
Uh okay, I can maybe attempt sort of a I partially believe this uh explanation um who who which is that we do seem to be sort of programmed to recognize and find attractive instinctively uh novelty which is associated to structure somehow.
4:51
Uh and so we we look and we see spiral galaxies or things like that.
4:56
It's reflecting something which is interesting.
4:58
Uh we don't necessarily know quite what, but but maybe there's kind of an evolutionary explanation for why that is attractive at least. But I don't know.
5:05
I can apply that explanation to your question about sound.
5:06
Um it's equally as good and uh unfortunately it just doesn't seem to hold there.
5:11
So I I'll be I'm not that confident.
5:16
Now you've written that in the first half of your life, you typically were the youngest person in your circle.
5:19
And that in the second half of your life, which is probably now, you're typically the oldest person in your circle.
5:27
How would you model that as a claim about you?
5:32
I hope I'm in the first 5% of my life.
5:34
Um but it's un- sadly unlikely.
5:37
now and you live to 100, right? Which is plausible. Which is plausible.
5:40
And you would now be in the second half of your life. Yeah.
5:42
Um I mean, I can give shallow reasons.
5:48
I can't give good reasons.
5:49
In The good reason in the first half was so much of the work I was doing was kind of uh new fields of science.
5:55
Um and those tend to be dominated essentially for sort of almost sunk cost reasons.
6:01
People who don't have any sunk costs uh tend to be younger.
6:06
They go into these fields uh and and and uh so this kind of early days of quantum computing, early days of open science, they were dominated by people in their 20s.
6:14
Um uh and then they'd sort of go off and become faculty members.
6:19
They'd be the youngest person on the faculty.
6:20
Now, maybe it's just because I found sort of San Francisco and it's such an interesting cultural institution or sort of achievement of civilization.
6:29
We've got this amplifier for 25-year-olds that that lets them uh uh make dreams in the world.
6:34
Uh and that's for me anyway, for a person with my personality, very attractive for many of the same reasons.
6:40
Well, let's say you had a theory of your collaborators.
6:42
And other than yes, they're smart, they work hard, but trying to pin down in as few dimensions as possible who's likely to become a collaborator of yours after taking into account the obvious, what's your theory of your own collaborators?
6:56
They're all extremely open to experience.
6:58
They're all extremely curious.
6:59
They're all extremely prosocial.
7:01
They're all extremely ambitious.
7:03
They're all extremely imaginative.
7:07
And you think that ends up pairing you with collaborators who are more different than you.
7:10
So a lot of collaborators are very similar and then other types are very different.
7:13
So almost always as well, I will select for somebody who has at least one very strong skill which I do not have.
7:19
And that's that's sort of enough diversity from my point of view.
7:23
And that may account for some of the age differences Yeah, sure. throughout your life.
7:27
But also then there's also just there's this like this this local selection effect.
7:30
When I live in the Bay Area, it's um there's a lot of really amazing 29-year-olds around. It's just incredible.
7:37
I was told to ask what's the influence of Simone Weil on you?
7:39
Oh, what an interesting question.
7:43
Um She's one of the maybe the best examples of sincerity that I know of.
7:50
The fact that she wrote what she wrote for herself.
7:53
She wasn't attempting to get published.
7:55
It was just this deep internal colloquy that was going on.
7:59
Um and it's reflected in every aspect of her life.
8:01
You know, she went off to fight in Spain at a time when women did not go off to fight in Spain.
8:07
Um she had all these Everything that she did, she did it at 500 miles an hour.
8:09
It was really She's she's remarkable as kind of a she's an extreme very extreme type of a human being in a way that I find very interesting.
8:21
If you and she collaborated, what would it be on?
8:24
I don't think Gosh, I'm sure she was a difficult person.
8:27
Although her brother Andre Weil was a very great uh uh mathematician.
8:31
And you can sort of see in in some of the stories about the two of them that she must have quite liked a sort of scientist types.
8:37
Um maybe we would have found something to collaborate on.
8:42
Why is Charles Sanders Peirce still an important thinker?
8:44
I don't know enough about Peirce to be able to answer that question.
8:48
You and I were both fans of Olaf Stapledon who wrote the dual classics Last and First Men and Star Maker.
8:56
What's the biggest analytical mistake he made in those narratives?
9:00
So a lot of implausible things happen, right?
9:02
But those are too simple to point to.
9:03
Where is his understanding of the social world going wrong?
9:07
You know, he was both to certainly to some degree a socialist and certainly a pacifist.
9:11
Um Though in World War he switched out of pacifism.
9:16
As did as did many people.
9:16
Um uh I find myself as I read those books actually becoming a little bit more sympathetic.
9:23
I'm not a priori particularly sympathetic to them.
9:25
And and I I start to think um Like he has this very long view of of history.
9:33
I mean, much longer than most people who would say they have a long view of history.
9:35
And and I think he sort of he sees some of his pacifism in in that light.
9:40
Um It's kind of questions about, you know, what's actually good for a species or in fact not even a single species, but across multiple species.
9:45
Is it is it good to be pacifist?
9:50
Um and that's a really interesting point of view.
9:51
It's hard to reconcile with kind of a selfish gene uh kind of a point of view.
9:55
But of course, this is an ongoing problem in evolutionary biology.
9:57
It actually seems like, you know, group selection doesn't quite work, but but something at that level that has to be has to be a little bit true.
10:06
So, if you take that that seriously, um then then maybe his pacifism, which just seems like a sort of an outright mistake, maybe it's actually justifiable um in in some way.
10:17
Actually, I'm not answering what I I've entered the I've entered the inverse of your question, which is to justify the bits that I I a priori find most implausible.
10:24
But yeah, I I think those are mistakes.
10:25
I worry that he too quickly assumes collective action problems are solved, which is close to your answer.
10:31
So, he thinks the League of Nations can be effective for a long period of time, which I suspect was not really contingently possible.
10:40
And he has this Hegelian sense, what Hegel would call a national spirit.
10:42
For him is a civilizational or certain stage of man spirit that so shapes how people think.
10:51
And I hang out with a lot of economists.
10:53
I think that's much stronger than the economists believe.
10:56
Your overall view of the world and what's important, uh but I don't think it's nearly as strong as Stapleton believed.
11:03
So, the way in which collective spirit rules millions, billions, or trillions of beings, I feel he's overestimating the efficacy of that.
11:11
Yeah, I I The comment about the League of Nations is really interesting.
11:15
I think you there's this spirit at the time.
11:17
Lots of people wanted this idea to work.
11:20
Lots of his friends would have wanted it.
11:22
I I think it's a sort of a shallow kind of a mistake that he made there.
11:25
But your comment about collective action problems seems much more to the heart of it.
11:30
I I think he didn't really believe in them or actually sort of understand just how difficult they are to to solve, how how difficult it is to to supply public goods and and these kind of things.
11:43
It's It He always does away with it sort of narratively and you know, it's just it's assumed away uh without really a mechanism being given.
11:53
And and he assumes a lot away in those books.
11:55
Um but when the question when the problems are interesting, he usually doesn't and and that problem is interesting and he he still assumes it away.
12:02
I I'm not very sympathetic to that at all.
12:05
But I'm not sure how big a mistake League of Nations was.
12:07
So, clearly it didn't work and I just criticized him for it.
12:11
But if you think about 1815 up through the First World War, almost a century, you have an unprecedented degree of peace in much, not all of Europe.
12:22
And everyone has just lived through that and they maybe thought that was not possible.
12:25
And maybe that is itself still a bit of a mystery.
12:29
And then there's World War I and you feel you can get back to some version of what you had and the League of Nations appears to be the closest path to doing that.
12:36
And it might have been more plausible at the time.
12:38
I'm just saying this thing there's a gap between aspiration and what actually happened.
12:42
Um I mean with the League and then later with the United Nations.
12:44
I think all the the the you know, you had the hopes and then you had the what what actually happened.
12:51
There's a very large gap.
12:51
Although, I mean of course as prototypes, you know, over the next few centuries, uh maybe these things are terrific.
12:58
Maybe we learn a a lot from them.
12:58
Things like I don't know whether with the Montreal Protocol have been possible without the United Nations. Probably not.
13:06
I have a very concrete question for you.
13:08
And this is to clear up a confusion of mine.
13:10
So, I've asked experts in quantum computing, what's the status of quantum computing right now?
13:14
Some of them say we already have it.
13:16
Some of them say the others will tell you we already have it, but we don't.
13:19
Others will say we're on the verge of having it.
13:21
And there's two or three other answers I hear, often people who nominally would seem to know what they're talking about.
13:29
So, let me ask you, Michael, what is actually the status of quantum computing right now?
13:33
See, I I'm I'm the wrong person.
13:33
I am determinedly very agnostic about about this.
13:37
I I stopped I worked on the it from 1992 to 2007.
13:39
Um And actually I I I do keep up with friends.
13:44
In fact, I'm going to have have coffee after this with somebody who's who's still on the quantum train.
13:50
Um but I don't Impressive is it's very impressive progress each year.
13:59
It is an extremely difficult problem. It's not solved.
14:01
There's no way is it's definitely not solved.
14:03
Um But the fact that there's sort of order 100 cubit systems which you can apparently manipulate as you will um suggest to me we just wait.
14:16
It's it's it's it's going to happen.
14:16
Um we don't know what it will mean.
14:22
What's your maximum likelihood estimate for the first year when it will do something useful?
14:31
Useful to me or useful to civilization?
14:34
Um And the most interesting thing would be to discover that quantum mechanics was wrong from my point of view.
14:39
The other most interesting thing is probably discovery of new materials.
14:42
Um I I How would it discover new materials?
14:47
Just by being able to do simulations very very rapidly.
14:51
It's very hard to do simulations of of stuff down at the the quantum scale.
14:53
The ways that we have are pretty terrible and often produce wrong results.
14:57
Um the fact that we may actually have a very high throughput way of of doing lots and lots of simulations which give correct results.
15:05
You know, it's like being able to do a thousand times as many experiments as before.
15:08
Um that that will just speed things up.
15:10
In so far as there's anything to discover, I can't tell you what we'll discover.
15:14
Uh Will there be quantum money?
15:15
Will all money be quantum money in this world whenever it comes?
15:20
Uh I actually don't know.
15:20
I mean there's this old idea of Stephen Wiesner which he he called quantum money.
15:23
Um it's it's meant to be uncounterfeitable. I don't know.
15:28
But isn't everything else counterfeitable if quantum computing is up and running and thus you need a quantum money to protect against to sheer counterfeiting?
15:41
Most of the 19th century money, they were often counterfeit.
15:42
We don't know the exact percentage, but we believe it was quite high.
15:47
Yeah, well, it's still true in in the world today, never mind the 19th century as as yes, we we've talked about before.
15:54
Um Yeah, I I I I don't I won't be surprised if we end up with with systems like that.
16:03
It's hard to make it stable. That's the issue.
16:06
But my guess is that in the long run we actually will find ways of of making quantum systems surprisingly stable.
16:13
That's speculation on my part, but you if I come back in a hundred years time and that's true, we may just have quantum coherence everywhere.
16:21
Do you think that leads to a mass privatization of a lot of social activity?
16:24
So, something like AI, we're in San Francisco, the private sector does it.
16:28
No government is really close to doing it, right?
16:33
You have to pay high salaries, hire the most talented people.
16:36
So, if AI and quantum computing are done by the private sector, what is government in that world? I don't know.
16:43
I I mean, you it's an interesting fact, right, that work on nuclear weapons was actually nationalized. Right.
16:50
Um you know, I think 1948 or something like that.
16:52
Um so, I mean potentially that's just one answer, right?
16:54
Sort of contingently But that seems more of a brute force thing than what I see OpenAI has done.
17:01
I'm just saying that that's a that's a potential outcome. Yeah.
17:04
I think quite a plausible potential outcome.
17:06
I don't think it's likely, but but but it's it's not 99% unlikely either.
17:11
Um Yeah, I I mean that that's certainly it's kind of a very Neil Postman point of view.
17:18
You have this basically you you almost Larry Lessig, code is law.
17:23
You just keep building more and more sort of governance infrastructure into the technology and you're moving it out of the hands of the the the population and and into the into the technology and that seems to be certainly the story of the last hundred years and very likely the story of the next hundred years.
17:39
Is the status of linear algebra rising? That's a great question.
17:45
So, it's prominent probably has, yeah.
17:47
It's prominent in quantum, right? It's prominent in AI.
17:49
I Google is built on, you know, matrix multiplication.
17:53
Um you know, it's it's it's prominent for for a lot of reasons.
17:58
Well, what should we infer from that about the whole nature of the world?
18:01
So, if differential equations were rising in status to a similar degree, we might infer one set of things.
18:04
But linear algebra, you almost feel a bit more grounded, don't you? Uh I don't Well, yeah.
18:11
Because when I took that class, I felt I understood it.
18:16
Uh I never quite know what your status questions mean, Tyler.
18:18
I don't know what it means for something to rise in status.
18:22
Well, AI now seems more important than it did five years ago.
18:26
And matrix multiplication is a big part of that.
18:28
If quantum computing happens as you're predicting it will, well, that I think would also make matrix algebra rise in status.
18:33
Like, oh, this is a really important tool.
18:35
It's behind all our quantum money.
18:37
Do you mean it's going to have more money go to it, more power go to it, more glamour go to it?
18:45
Are people going to, you know, regard this as as sexy or what?
18:46
But you would revise your ideas about the fundamental nature of the universe.
18:51
Just like our current understanding of quantum mechanics, it might be incorrect, but at least in the short run, it seems like probability theory is somehow more important than Einstein might have thought.
19:05
And as you know, he famously asserted God is not playing dice with the universe, perhaps incorrectly.
19:10
I mean the people who remake this understanding are very good at at ignoring status. Um But others aren't. But others aren't.
19:21
Um I think I'm inclined to think uh I mean maybe I I don't care.
19:30
I actually just don't care that much.
19:30
Um Uh Yeah, I mean if you're searching for comparative advantage in doing creative work, you want to know where status is, but mostly so you can avoid it. Yeah, absolutely.
19:41
Be short status as Peter Thiel likes to say. it's well put, yeah.
19:45
Is there any chance Roger Penrose is right and the human brain is some kind of quantum computer?
19:50
I would love it if he was right.
19:50
I think the answer unfortunately is not really.
19:57
It's only possible that there's some very interesting structure in there that that that is quantum mechanical in some really interesting way.
20:03
I mean, lots of structure in there is quantum mechanical.
20:06
The reason why atoms are stable has to do with quantum mechanics.
20:09
Like all these sorts of things.
20:09
But like an interesting unsuspected way would be that would be terrific and I think it's not completely out of the question, but it probably doesn't make affect anything about consciousness or anything like that.
20:20
I'd be very surprised if that were the case.
20:23
How are we going to make progress toward a theory of quantum gravity, a general understanding of everything? We seem to be stuck.
20:29
Many people hate string theory.
20:31
Many people hate Everett many worlds.
20:33
Those seem to be two major contenders.
20:37
Where are we at and what's going to happen next?
20:38
One fun reason for working on quantum computing is you're trying to build the most largest scale fully quantum coherent systems that have ever been built.
20:50
Whenever you push on into a new regime like that, there's some chance that things break down.
20:55
If something was to break down there, that would be fantastic cuz we'd learn a lot. Cuz we'd learn a lot.
20:59
And and that's kind of what the problem in some ways in physics has been that the fundamental theories have been just too successful for the last 50 years.
21:07
Yeah, yes, you're right again is very attractive for sort of a few years, but over 50 or 60 years, it's terrible.
21:16
And and I think that's you know, certainly part of the the the issue with with quantum gravity.
21:22
Does it bother you that so many people hate string theory, think it's now low status, think it's not aesthetic, think it's unintuitive?
21:29
Does that carry any weight with you or do you want to be like short status again on this one?
21:35
I mean, there's the question of sort of inside and outside.
21:40
Well, there's a question of inside and outside the profession.
21:41
There's also the question of inside and outside the the group of people who know something and those two are not exactly the same group, but there's a lot of overlap.
21:49
Um So I mean, outside it affects funding a little bit.
21:54
Well, actually maybe quite a bit.
21:56
Um And so in that sense, it matters.
21:58
Um But internally, I think I'm more interested in the question of just how much diversity of opinion is there.
22:09
Are people pursuing lots of different ideas?
22:10
One of the things that I've I've noticed over many years is I just I find mathematicians when I talk to them, it's such a healthy culture because each mathematician is really loyal.
22:18
A lot of them are very unique.
22:20
They've got their own sort of particular path and their set of beliefs.
22:23
Physics theoretical physics often seems a just a little bit more monotone.
22:26
They can sum themselves up in a few words when they're talking to their professional colleagues and that that's not so healthy.
22:35
So I'm really I'm not so much interested in the question you asked.
22:38
I'm I'm much more interested in the question of how do you generate that kind of diversity?
22:42
And do you feel that that ultimately the final theory of a universe or metaverse ought to be simple?
22:48
Does that Who's declaring ought here?
22:52
But when someone presents a theory to you, do you ever say, "No, that's too complicated.
22:55
It might be an intermediate theory at some level, but it's not going to be the final theory."
23:00
Cuz I hear this from many people. A lack of satisfaction. You want surprise.
23:04
It's the same when I've you know, the little tiny pieces of economics I've I've learned.
23:08
When I hear about I don't know, Ricardian comparative advantage or something like this.
23:13
There's a just a nice little element of surprise that you're getting a free lunch somehow.
23:18
So more interested in that than I am maybe in the the question of simplicity.
23:22
What makes for physicists who age well?
23:27
Uh I spent quite a bit of time thinking about this as actually in my late 20s and and went to to look and see what seemed to distinguish older physicists who had aged well and and older physicists who had maybe gotten a bit too complacent.
23:42
As far as I could tell, having younger mentors was really the key.
23:46
And why is that important? I don't know.
23:50
I mean, there's an I I have theories.
23:51
It's an empirical This was an empirical observation.
23:54
Yeah, but what's your best theory?
23:56
What I think is probably the case.
23:59
It's almost a network effect.
23:59
Basically, if there is some slight sort of downhill slide and most of your friends, you know, are not quite at the edge anymore, that's going to infect you.
24:10
But if you still have mentors who are 25, 28, extremely active and they're active in the latest ways, you get to partake of the positive and so I think that's why it's very important.
24:24
Not just to have people who work for you.
24:26
Lots of 70-year-old physicists have have 23-year-old students, but actually to have 23-year-olds, 28-year-olds who who you really learn from and and you regard as your mentors.
24:37
And holding constant your degree of power and influence, what's the best way to attract younger mentors?
24:44
Find people whose work you admire and befriend them.
24:46
And you think that works pretty well? Yeah. And just being nice.
24:52
I'm not sure being nice is the right There's There's a difference a lot of younger mentors.
24:55
You're known famously for being very nice. Right?
25:01
So this is partly a question about your own self-awareness.
25:03
But has you being very nice helped you get more younger mentors?
25:06
Or are they attracted to other aspects of you?
25:09
I am extremely disagreeable, but in a polite way, I hope.
25:12
And a kind way, hopefully.
25:15
So you are very nice then.
25:17
Well, people often find people who are disagreeable actually quite difficult, but if you look at all of the the younger mentors I've had in the last say seven or eight years, they're all people who enjoy disagreement.
25:27
You know, they they say the thing that they think is obvious and you say, "Here's another way of looking at it."
25:32
And they're like, "Oh, you know, they want to engage."
25:36
Some people get insulted or they get threatened or they get annoyed when you do that and those people are not going to be they're not going to be good collaborators.
25:44
They're not going to be a match.
25:46
As the years passed, do you think your probability for God existing is going up or down?
25:53
Do you mean which type of God are you referring to here?
25:56
Are you referring to like a Abrahamic God or or what?
26:00
No, no, not a particular religion, but some explanation that would seem to stand prior to and outside of what we call physics and would be mystical in some way.
26:12
it's been um That That hasn't changed since I was 7 years old.
26:17
But that's weird that it hasn't changed, right?
26:19
Why shouldn't you've learned a lot? Why shouldn't it change? In whichever direction.
26:23
suppose I learned I had explained to me um three basic theories of cosmology when I was seven, one of which was the Big Bang and and then there were two others which the steady state theory and a third whose name I don't even remember anymore.
26:36
Um and they leave some questions unanswered.
26:44
Yeah, why is there anything?
26:47
But as far as I can tell, we haven't made any progress on sort of the the the those things in the the 40 odd years since.
26:54
Um Yeah, it's it's frustrating actually that that's the case.
27:00
I mean, you I think you're correct to say, "You've learned a lot.
27:02
Why haven't you changed?"
27:04
And my response to that is I've learned a lot.
27:07
Gosh, it's really annoying that I haven't that it hasn't impacted that question more.
27:11
I think my P has gone up a modest amount over time.
27:14
So when I was say in my young 20s, I thought physics was going to make more progress than it has at fundamental theoretical levels.
27:23
And the fact that it hasn't, it nudges me a bit to wonder well, these other types of explanations that I was not so keen on, maybe they're a bit more important than I had thought.
27:35
That hasn't happened with you?
27:37
No, the thing that that that's an interesting It hasn't really.
27:39
I I think I just don't think 40 years is very long.
27:45
If it had been 100,000 years, I But it's all I've got in a sense.
27:47
I mean, I'm I'm going to have a bit more, but Hopefully, yes.
27:52
My opinion time span is going to be 40 plus something.
27:55
I think my appreciation for God has gone way up.
27:59
Like I you know, I I appreciate the construction of the religions far more than I did.
28:04
What what sort of notions of God do for people.
28:07
It's vastly more appreciative.
28:09
But my my probability, I don't think it's really changed.
28:12
And what about evolutionary frameworks where there's some Darwinian process, some kinds of universes within a broader metaverse, they reproduce at greater frequencies, that shapes the properties of what we live in.
28:23
Isn't that a kind of substitute for a good explanation and that rises in probability just a bit? No, I mean Why not?
28:34
Um Yeah, you're you're relabeling what you mean by by universe.
28:39
If you just sort of use a term that means everything that is, then then that hasn't changed.
28:44
Our model of what it might be has has potentially changed quite a bit over the last few decades.
28:49
But it the the But maybe there's a simple theory for the metaverse, but we can never ever see it.
28:56
It's like gnostic religion.
28:56
And then our own universe, there's not a simple theory, but we do know the parameter values we got are enough to drag it across the finish line and that takes some of the burden off physics in a way.
29:09
Just like well, the platypus.
29:09
It seems an unlikely creature, but it has in fact survived.
29:14
Okay, you're you're just sort of saying that The universe is like a platypus.
29:18
And being a good Australian, you Yeah, indeed.
29:21
appreciate the platypus, right?
29:22
the platypus quite a bit.
29:22
It's good for fooling visiting Americans about whether or not this animal can exist or not. Yes.
29:29
Uh I still don't find that compelling.
29:31
I think because there's always we've always known that there seem likely to be fairly contingent facts about the universe.
29:37
I mean, it shifts the level at which they are.
29:39
It's more interesting if the value of the fine structure constant is actually a contingent fact.
29:46
That that that is interesting or if you know, some of the couple other couple of constants are changing over time or sort of models like this, but it's not particularly um it's still not getting at the essential question from my point of view.
29:59
I mean, which is why is there something rather than nothing?
30:03
That I've long thought is an impossible question.
30:06
So, we might have theories of parameter values or be able to predict how things interact or what happened a long time ago.
30:13
But the Heideggerian question, I don't I don't think it's a meaningful question at all.
30:19
Because the the word why is already embedded in some context which it sends a self-undercutting query. Yeah. Now, open science.
30:27
Why do some fields have preprint platforms and others not?
30:31
Is there an actual regularity or is that random and path dependent?
30:35
I think a lot of that probably comes down to individuals.
30:37
Um, one of my favorite things I I years ago before they'd started to spread in biology, I would ask often ask uh physicists and biologists this question and why why is there preprints in in physics but not in biology?
30:50
And and the biologists would say, "Well, you know, biology is so much more competitive than than physics that we can't possibly bear to share our results um too early."
30:59
And the physicists would say, "Physics is so much more competitive than biology that we have to share them as rapidly as possible to get the word out."
31:07
Um But with COVID, didn't biology go the route of at least some parts of biology go the route of physics?
31:14
And it just seems like it's just a cultural problem.
31:15
You know, it turns out it's a little bit more like fashion or something like that.
31:18
It doesn't need to be solved.
31:20
Like if you look at what was done in the early days of the physics preprint server, you know, some very clever things actually things which are reflected in some of my favorite economists uh sort of some of their ideas um were done by Paul Ginsparg when he was starting up the preprint server. He went very narrow.
31:34
He didn't try and solve the the problem all across all the fields and he went and kind of twisted the arms at some level of some very high status, high profile physicists to say, you know, I would like you to use this service.
31:47
Send me your your your best best paper.
31:49
So, on the first day, what's his name?
31:51
Uh Andy Strominger who I think is at Harvard and then uh uh you know, it was on the preprint server and Ed Witten showed up very quickly.
31:58
So, these are very prominent people and you get this sort of just it's a tiny community, but then you can sort of you can you can agglomerate.
32:04
You can start to attach other other communities.
32:07
Um, but that was just you know, that's a very contingent fact about history.
32:11
It could have it could have happened in some you know, subdiscipline of of biology as well.
32:16
Why do so many crummy journals survive and they can be quite expensive?
32:18
You might also have to pay to publish in them.
32:23
They seem terrible that if a good piece were in them, the journal would not certify the piece.
32:26
If anything, the piece would help certify the journal.
32:30
Why can't we get out of that?
32:30
Yeah, I mean, there's a a complicated set of things going on.
32:35
One is that libraries pay not to individuals uh usually for subscriptions.
32:38
So, they're not actually really the the person getting the utility is not the same as the person as the person making the buying decision. That's always bad.
32:48
Um there's also the fact um that since the 1990s and the rise of the internet, um you know, we get economies of scale.
32:57
You don't subscribe libraries don't subscribe to individual journals for the most part.
33:00
They subscribe to all you know, these giant bundles which is actually a terrific idea at some level.
33:05
It's a way of passing on economies of scale in publication to the customers, but it does go some way to explaining why these crummy journals persist.
33:12
But you think in part libraries are inefficient at capturing rents for themselves.
33:16
That they get this budget, they spend it on bad journals.
33:21
It might be better for the world if they just took the money home and bought ice cream, right? Or did whatever. Yeah. Yeah.
33:25
There's many other things that could be done with that money.
33:28
Um, it's it's difficult for them to reason about.
33:30
Having talked to many librarians, they will do things like they will use impact factor.
33:34
I mean, that's the the the differentiator that they tend to use.
33:37
So, they'll try and get you know, all impact journal impact factor you know, whatever it is and above journals.
33:43
Um, is the the the that's the kind of the way they they seem to think.
33:47
They're just using a very imperfect proxy.
33:50
They understand as well as anybody that their proxy is imperfect, but they don't have anything better to better better better to do as far as I can tell.
33:57
So, it's a very um unfortunate situation.
34:02
And right now, how high are the marginal returns to greater openness?
34:05
So, put aside terrorists manufacturing new pathogens.
34:09
Put aside people figuring out how to make their own nuclear weapons, AI problems.
34:13
So, putting aside the very negative, just if the good stuff were more open, how much more rapidly would science progress?
34:23
Uh openness per se, I mean, that's a very weak word.
34:27
Uh you need to be sort of much more specific.
34:29
If you look at say um the culture around Jupiter notebooks in machine learning, I think having those very openly available and widely available really has driven a lot of progress.
34:37
You can just you write your Jupiter notebook with your experiment, you make it available to other people and and um that can really drive a lot of progress.
34:44
It's not the same as making your journal article open and available.
34:48
Um, it's a much more active kind of a uh material.
34:51
Like do I do I think that that is an important component in really significantly speeding up science? Yes.
34:58
Um, but it's not you know, it It's not going to be 2x.
35:02
Um Or I I think that there's much larger than 2x uh possible and this is a small this is a piece of of that. Yeah. Yeah.
35:09
But it's not 2x on its own.
35:13
No, it's it's actually it's two undefined uh a term.
35:16
You know, openness is always with respect to what platform, with respect to what set of institutions, with respect to what set of of norms.
35:24
Um, with the current sets of norms and institutions that we have, uh it buys you a little bit.
35:27
I don't think it buys you that much, but the norms and institutions they're going to change in response.
35:31
The way in which people work will change in response.
35:33
You I mean, the the Jupiter notebook example I gave is is I think a good good example of that.
35:39
Why are science textbooks so expensive? Is it marginal cost?
35:41
Is it third-party payment problems? Is it something else? I don't know.
35:47
Is it instructor lock-in because the notes are geared to text they've worked with for 15 years? Yeah.
35:54
Yeah, very few professors make that much money from the textbooks that they write, but they're often very protective.
35:59
I see people complaining on on on Twitter that they're not going to get the $400 check next year for their their textbook under a new open access policy.
36:08
Um, uh they're really up in arms about this $400.
36:14
Uh that that that's interesting.
36:17
It's hard for me to empathize with psychologically. Um I don't understand.
36:25
I mean, a priori I I wrote this uh neural nets uh uh textbook which I put online for free um and you know, that massively I mean, that's really made a really large difference to the impact which which it's had.
36:36
In fact, even if I just think purely financially, it wasn't I wasn't doing it thinking, "Oh, financially this will be better off."
36:43
But the greater impact has actually benefited me much more financially than any amount of royalties ever would.
36:48
You mean like giving talks or being invited? Yeah.
36:51
And just in general, people know you know, who you are, what you've done, what you're interested in and they're much more likely to provide all sorts of different opportunities including jobs.
37:00
Um so I think from the point of view of the authors, it really actually doesn't make that much sense.
37:06
Um, from the point of view of the publishers though, it it might might might make more sense.
37:10
They they can make a I mean, there's the textbook market is not huge, but it's it is multi-billions.
37:17
You have a well-known article with Patrick Collison on progress in science slowing down and it's published at a point say right before mRNA vaccines, right before GPT-4, other developments. Mhm. Right.
37:30
How well can we know the progress of science at any point in time?
37:32
Isn't there often an everything all at once effect?
37:36
And in fact, those years we were building up and investing in things that very suddenly then flourished.
37:43
Yeah, uh it's amusing to think about you know, different points at time which which you could write try and write the same article.
37:49
Um, you know, probably the the years before the Principia, uh they would have been you would have been able to to do the same thing.
38:03
There is some question about like what certain types of institution might possible. Actually, I don't know.
38:09
I I think really my my the higher to be in my response is going to be something like like I just think AI is is uh it it's not yet 100% clear, but I think it's very likely to drive a lot of scientific progress over the next few years.
38:22
Um, and that's just a case of you know, we're moving all of our so much of our cognition and eventually also the actuators, the the way we operate in the world um out into these devices where all of a sudden it becomes much more mutable and and hopefully improvable.
38:38
Do you think the private sector wages of scientists are a good proxy for progress in science?
38:46
Because if science if science is declining in value, you would think scientists would be paid less and less.
38:51
But over the last 40 years, mostly those wages haven't fallen.
38:54
Yeah, I I don't think it's a good I mean, you know, I just you know, Isaac Newton wasn't the the the richest person to ever live.
39:02
Um, but he probably did more for understand human understanding than than any other more of a market now.
39:08
If the Isaac Newton of today would probably be pretty wealthy.
39:11
Einstein could have been wealthy had he done more media, right?
39:18
But he wouldn't have been wealthy for what he did well.
39:19
Um you know, he would have been wealthy for what he did well. his fame.
39:24
He could have endorsed, you know, ski boots and other things. he what was it?
39:28
He he asked for $3,000 a year when he moved to the IAS and they gave him 15.
39:33
Um, so yeah, I think he wasn't very good at negotiating.
39:35
But just say that the wages for private sector sector pharma scientists, they seem to go up for quite a while.
39:42
When the drug pipeline seemed slow, should we have inferred from that, well, we're building up to some big things, some blockbusters, or not?
39:47
Yeah, this is a question for you, Tyler.
39:51
It's not a question for why I'm asking you.
39:56
Right, good good good good reason to do so.
39:58
Um I think I'm inclined to think I mean, there's always this interesting balance.
40:06
Actually, AI is a really interesting example at this point in time.
40:09
There's this theory, which has become widely believed by almost everybody, that scaling is very important.
40:14
Scaling is a very capital-friendly story.
40:16
So, it actually moves some of the power and the the negotiating power from individual researchers, I think, to centers of capital.
40:23
Um but it is just a story.
40:23
I I I think it's quite interesting that it it in some sense gives the individual researchers less negotiating power, whether or not this is going to, you know, eventually result in a diminished ability to build personal brands and then capture value from that. I I I I don't know.
40:46
I'm I'm really interested, actually, to see what will happen over the next few years.
40:48
It used to be that the big companies published a lot of papers very openly, and that is gradually going away.
40:54
And as that goes away, it damages the individual researchers because they're not able to build their brand brands publicly in that way.
41:01
They're not as easily able to say, I am the person who did But this is a small city. Doesn't everyone know?
41:06
I I had dinner with a bunch of AI researchers last night.
41:10
They all seemed to know each other's relative importance.
41:13
Yeah, I mean, I And you look at their salaries.
41:16
So, there's rumors They're doing fine.
41:19
researchers can be offered 5 to 10 million dollars a year.
41:21
Those must be some of the highest science salaries ever.
41:25
And you're saying AI is such a big thing.
41:27
You seem to be coming down on the side of the wages, predicting something.
41:33
I think actually, I mean, there all I'm, yeah, I I I think so much money is is is going in, and it's going in on the basis of of brand to some extent.
41:40
We've hired such and such a person who did such and such a thing.
41:46
And and if that can make your your company valuation go up by a few hundred million dollars, then, you know, offering them another extra million dollars a year makes sense.
41:53
But I I don't have a grand, you know, this is all local storytelling.
41:56
It's not grand, you know, theory of what's actually going on.
42:04
I I I haven't thought it through in enough detail to have any confidence there.
42:08
And now you're working on what I think you call the vulnerable world hypothesis. Yes? This what Nick Bostrom. Yeah, that's his term.
42:14
What do you think is the cost at which a nuclear weapon could destroy a city?
42:20
So, if that costs only $50,000, it seems to me the world's in big big trouble pretty quickly.
42:26
What's that cost level where you get very very nervous?
42:30
So, if it's 10 billion dollars, maybe things are fairly safe.
42:32
If it's 50K, we're done for. What's the threshold?
42:37
It's the cost and expertise.
42:39
Let's say the expertise is comparable.
42:41
So, somebody who has 50K is probably able to get the expertise as well. Right. Yeah, that's not great.
42:50
Um That's my fundamental worry.
42:51
I I With or without AI, that just that cost becomes low.
42:55
Honestly, I I mean, there's some question about what exactly you mean by nuclear weapon. Like, is it portable?
42:59
Is it um uh It would render a mid-sized city uninhabitable Okay.
43:05
for a few decades at the very least.
43:07
Sort of a multi-megaton, but but not It would make the headlines.
43:09
It would make the headlines.
43:11
A very small nuclear bomb will make the headlines if detonated.
43:16
Um Uh uh Yeah, I mean, if we're getting down to hundreds of thousands of dollars, um uh Uh I The thing is, the the issue is the issue is the issue is, right, nuclear weapons are terrible, but they're not civilization-threatening directly.
43:30
Um Sure, but enough of the If enough of these go off, life as we know it is over.
43:35
They can certainly be destabilizing.
43:37
It'd be like the fall of the Roman Empire, maybe worse.
43:39
Yeah, that's that's what it it starts to to to to seem like.
43:45
But in what year do you think that the cost will be low enough that that happens?
43:50
I mean, at this point, I don't I don't have a good sense.
43:52
Um I suppose I'm actually more concerned about other threats.
43:54
But if this is a is the obvious thing.
43:56
I wouldn't call it a certain threat, but if you simply think technology will advance Here's the thing.
44:02
Like, the nuclear nonproliferation treaty means that it, you know, we actually have a lot of controls of the ability to produce fissile material.
44:08
But that's like the League of Nations, right? Ultimately? Yeah, I mean, it is.
44:11
It's it's really coming out of that same sort of, we have this cartel of, whatever it is, 10 countries or something.
44:18
It'll fail when the cost is low.
44:19
It will eventually fail, um but that's many decades away.
44:25
But you have reason to think we're going to last a thousand years in a civilized state.
44:29
Not every person dead, but I think getting off planet Earth and establishing a civilization elsewhere is very very important. Yeah.
44:39
Very hard for economic reasons, but utterly crucial.
44:43
Robots in a sense make it harder because you could send robots to Mars to do whatever might be economically useful there.
44:49
Means you never work hard on having humans do it. Yeah, that's true.
44:53
Um I mean, we're pretty curious.
44:55
Uh uh But the robot will take perfect footage.
45:01
Whatever is there, the robot will send back to us.
45:03
You'll have your, whatever is the current version of Apple Vision Pro on. Right?
45:06
It will seem very realistic. You're an economist. I'm a romantic, I think.
45:12
Might be Might be the difference.
45:16
But we'd have to settle them at scale.
45:17
So, 20 people on Mars limping along.
45:17
Oh, we're talking about like a million people, not not 20 people.
45:21
You know, you want But if we can do a million, we can do a billion, I would think. Sure, sure, sure.
45:26
I'm I'm saying, you want to get, you know, it's still not going to be It's not going to be self, what's the right term?
45:31
It's not going to be an autarky or whatever the the right term is.
45:34
It's not going to be completely self sustaining, yeah.
45:37
sustaining, but at a million people, you know, it's doing a lot of the the It has a lot of the civilizational infrastructure.
45:44
And so, I think that's the the right sort of scale.
45:46
Casey Handmer has a nice book, I I think it's, you know, how to build a city of a million people on Mars or something like that, which I think is it's way too optimistic in many of its assumptions, but but but he's got the right He's got the right scale.
45:59
Economics aside, what's the main scientific constraint that has to be overcome? Is it gravity?
46:03
Is it effective radiation on the human body? Is it water?
46:07
I mean, to some extent, we're not going to know until we go.
46:09
There was this great experiment done a few years ago where there was a There was a pair of twins.
46:13
One went up into space for a year, the other one stayed on Earth.
46:18
And so, there there was the first time we actually got to, you know, do a controlled, well, somewhat controlled study where we we see what the impact of being in space for a long period of time does to human body.
46:26
And I mean, they just they discovered so many things.
46:30
This is still below the Van Allen belts as well.
46:32
Um so, we don't I mean, we just don't we just don't know the answer to those questions.
46:36
What's going to, you know, there's a whole bunch of problems.
46:39
The regolith on on on Mars is terrible for human beings.
46:43
I'm sure that the low gravity is going to be bad for them.
46:45
There's What else is there's there's shortages of There's nitrogen, fortunately, which we don't really have on the moon.
46:52
Um So, you're making me think civilization as we know it won't last a thousand years.
46:58
Yeah, I I think it's that that's a No, I I I also have a lot of, I guess, faith in in long-run economic growth.
47:06
Basically, at the moment, for us to go to Mars is, you know, very very very expensive given the return or to establish a permanent human presence in space.
47:15
Um if we continue to have economic growth, relative cost is just going to keep going down.
47:20
Um at some point, it's it's actually not going to be that difficult.
47:22
Does the vulnerable world mean near universal surveillance?
47:28
Unfortunately, I think probably yes.
47:32
Doesn't that then become the great point of vulnerability? Oh, absolutely.
47:36
Um Yeah, it's If you could ban universal surveillance from here on out, forever, would you press that button?
47:42
No, I I mean, I think, you know, the history of justice, to some extent.
47:45
So, about the term surveillance, it's funny, you know, it it's it has negative connotations.
47:51
People, you know, they think of Bentham and the Panopticon and the Gulag Archipelago and and and and the Stasi and and all these things.
48:00
But in fact, our ability to supply justice is dependent upon having a good understanding of what has occurred in the present and in the past.
48:07
So, to the Maybe it needs opaqueness as well.
48:09
It's this optimal mix of surveillance and opaqueness.
48:16
That you actually have some latitude to break certain laws, to misbehave, that keeps the system stable, limits the abuses of power, limits how much power the powerful have over us.
48:30
Yeah, that there has to be some sort of Madisonian point of view where you're you're you're you're bringing the the powerful institutions into conflict with each other.
48:39
We do that very imperfectly at the moment.
48:40
You know, ideas like search warrants and and and things like this, they're supposed to be checks and balances.
48:46
But it it seems like um the organizations which do the surveillance are too powerful.
48:52
They don't have a significant strong enough uh checks on them.
48:57
I don't know whether, you know, just as a practical matter, the United States is capable of doing this well.
49:04
Um I'd be much more comfortable if it was another country's.
49:08
In a strong AI future, where do the economies of scale lie?
49:14
Say, within your lifetime, not 500 years from now.
49:18
I'm not sure I'm I'm not sure I'm I'm not sure what what you're what Well, we're we're all trying to figure out how AI will shape the future, right?
49:25
So, one model is everything is supplied competitively.
49:29
Maybe a bit like fast food today.
49:31
I suspect that's not true, but it could be true.
49:33
There's the oligopoly model.
49:35
There's the one company races ahead of the others, and then its own AI does R&D at an accelerated pace, and they stay ahead forever.
49:43
Or there's one country, you want company, you want something controls all the chips.
49:48
Uh where do you see the monopoly power evolving?
49:51
Because it's essential, I think, to predictions of the model.
49:57
I mean, when I talk to people who know much more than I do, uh you know, they all pointed at ASML uh as having been surprisingly hard to chip away at.
50:05
Um you know, just the ability to not so much do the lithography, but do the lithography at scale, uh which seems to be very very helpful.
50:12
So, we should be long Netherlands. Probably, yeah. Yeah.
50:16
That would be an amazing conclusion, wouldn't be an amazing conclusion, but a return to the Dutch Renaissance.
50:21
So, it's like agriculture and lithography.
50:25
And drawing on 17th century strengths. maybe. That's right. Yeah, yeah, yeah.
50:29
Maybe the return of Vermeer. Yeah, yeah.
50:33
What do you think of the Netherlands as a country? Oh, I love it.
50:35
I I've I've never spent a Well, actually, I I spent a month there.
50:39
Who am I Who am I Who am I kidding? That's a lot of time.
50:41
That's a a fair amount of time in in Leiden, actually.
50:45
Uh you know, it's a lovely place.
50:46
It has um many problems, uh of course.
50:49
I mean, their altitude is not great.
50:51
Um in some ways, actually, in some ways, it's good.
50:56
Um yeah, they're an interesting sort of a test case.
50:58
They show what a strong, determined civilization can do in response to nature. Yeah.
51:02
Um uh so, the flatness of it.
51:05
I like the water being everywhere.
51:07
Not everywhere, but most parts, at least, of western Netherlands. Mhm.
51:11
I find that very attractive. Mhm.
51:13
The there's this stereotype I sometimes encounter people sort of view it as being a little orderly uh uh I've heard people say it's dull, but actually, so I think some of the most interesting experiences of my life were were were there.
51:28
I went to a um What's it a It's like a jamboree in the field so a week What was it?
51:36
5 days long, called Hacking at Random in 2011, where some people from Anonymous spoke, a whole bunch of cryptographers spoke.
51:42
It was really sort of hacker culture.
51:46
Um and it was just intellectually wild in the most interesting way, and it really it grew out of Dutch hacker culture.
51:51
I think there's there's a lot of Yeah, that that spirit of the the Dutch Renaissance is still uh still still visible.
51:59
How would you describe the quality of those conversations? What were they like?
52:04
Different than the ones in San Francisco? Oh, yeah.
52:05
And how different How are they different?
52:17
They're not captured by capital to the same extent.
52:20
Um Conversations in San Francisco, particularly with younger people, tend to be extremely idealistic and and often very prosocial. Mhm.
52:28
Um but then later, there's this sort of negotiation that goes on, where they need access to capital to make their dreams come true.
52:36
A certain amount of compromise is is made, although they they also often keep a lot of their original sort of prosocial and idealistic character.
52:42
In the Netherlands, uh in those particular events, um there's been less of that.
52:48
They also had less access to capital. Yeah.
52:52
If someone's going to travel to the Netherlands, they have a tech background, Yeah.
52:55
like what should they do, or what advice do you have for them?
52:58
How should they try to learn more from the Netherlands?
53:01
years since I've been there.
53:01
Um so, I'm not I'm not the right person But since you have good memory through spaced repetition, Yeah, uh maybe.
53:06
Um I I I mean, I yeah, I I I love going to the museums in in in Amsterdam.
53:12
Just personally, I Rembrandt is uh maybe my my my my favorite painter. That that's a Yeah.
53:21
It's hard It's hard, actually, to think of anything else when I think of the Netherlands, yeah, as the other than Rembrandt's late self-portraits, which I think are some of the most extraordinary things ever done.
53:30
Let's say we all had better memories.
53:33
How big is the social gain there?
53:33
Is there any social gain at all?
53:35
So, you've been an advocate of spaced repetition for improving your memory.
53:39
It works for medical students.
53:41
It probably works for languages.
53:44
Uh but are there social gains, especially with AI coming?
53:46
I'm not actually I wouldn't say I was an advocate. Uh uh But you do it. I do it.
53:53
And you teach other people how to do it.
53:55
benefits from it, and some other people get benefits from it, and I'm very enthusiastic if they do.
54:01
And if they don't, lots of people try it, and they're like, "This isn't working for me."
54:03
And I'm like, "I'll stop doing it."
54:04
Um same as, you know, if you listen to Bach and don't like it, uh stop listening to Bach.
54:09
Um uh uh you know, to what extent do I think Yeah, there are these interesting There's a a long sequence of of of papers sort of trying to elucidate the connection between um deep sort of practical expertise uh and the role of of memory.
54:25
I suppose most famously, uh people like Herb Simon and and Anders Ericsson and people like this um have have tried to understand what relationship, if any. It's a little bit murky.
54:37
They all make very strong claims, you know, about you An expert is somebody who's acquired sort of 50,000 chunks of information and things like this. I They're nice stories.
54:47
Um uh they certainly seem to be uh borne out, but I don't know what the causal thing is.
54:52
If I talk to you about economics, you can tell me Well, actually, not Yeah, not just understanding number of things about economics, but about a lot of different things. But I I don't know.
54:59
Are you an expert because you know those things, or is it downstream of something uh you know, is it really downstream of something else?
55:05
Um I'm sure it's part of it.
55:08
Uh if you had a magic memory, uh it it it it might might help you um a little bit, but I suspect, actually, it's downstream of something else.
55:16
Your determination, um curiosity, something like that.
55:22
There's some evidence that students learn better when they take notes of what's being said.
55:28
Do you feel there's something for some people with memory a bit similar, that until they have memorized it, it's less real for them?
55:34
I don't Just mean that they remember it more.
55:35
But the initial impact, somehow, is created or defined by the later act of memorization.
55:41
Like people who take trips, and until they photograph something, they don't feel they've seen it. Yeah.
55:45
In fact, they probably didn't They They very likely didn't see it. Perhaps, yeah.
55:50
Um it's only It's part of the reason why I take photos.
55:53
Uh uh I will look more closely.
55:53
Um that seems to be uh I mean, it's part of the reason I will take notes.
55:58
Um it's also part It is part of the reason why I I I do spaced repetition.
56:02
I It provides me with another way of paying attention to the world.
56:06
At some margin, Those things are very valuable, right?
56:09
Like any any general-purpose strategy you have which will cause you to pay attention to the world um is incredibly valuable.
56:15
Um and so, I collect things like that.
56:18
And you know, why why did I say yes to coming on the podcast?
56:22
A huge part of it is because I know it's going to make me pay attention in different ways.
56:27
I know that you're going to ask me questions that nobody else is going to ask me.
56:32
And so, for me, like the reason I do spaced repetition and the reason I I will will come on a a podcast with somebody like you who asks very interesting questions, they it's kind of the same.
56:42
But in some other ways, you're a fan of non-legibility, as am I, at other margins.
56:45
And there's some tension.
56:48
Because when you take the photo, when you remember something, when you write it down, there's less legibility.
56:52
And how There's no tension at all there.
56:54
You're constantly expanding the legible, and and when you do that, there's this sort of penumbra of illegibility that surrounds it.
57:01
That moves, but it gives you access to to to to those other spaces.
57:05
You know, I you you go to the Netherlands, you travel in in in in general, you make more of of the world's culture, you make people much more legible to yourself, but that that expands what you're able to to to to to see as well, at the at the edge of that.
57:22
Um that's part of the reason for doing it.
57:24
That's part of the reason for for wanting to make things legible.
57:25
What's underrated about travel, other than that?
57:28
My god, almost everything.
57:31
Um you denied saying this, but but somebody once said to me that travel is the only education, and it's really stayed with me um as expressing some deep truth.
57:39
I think mostly just the the world is so incredibly deep.
57:42
Um There's whatever, there's 8 billion people um uh It's the only way to see that depth and breadth.
57:49
Yeah, and it's just unbelievable.
57:49
I mean, you pick almost a random person anywhere, uh uh and you you could spend a great year with them uh uh just learning things, and and you can't, unfortunately, do that 8 billion times.
58:01
But uh but but but but but yes, it's very very underrated.
58:05
It's been very underrated in my life.
58:07
I haven't traveled nearly as much as I should.
58:09
Why is the Was it the Midway aircraft carrier?
58:13
Why is that so interesting? There are many reasons.
58:17
That's in San Diego, right? It is, yeah.
58:18
So, for about 10 years, it was probably the most dangerous object in the world.
58:22
Um it carried nuclear weapons, I believe.
58:25
Um um Having 5,000 people on it, basically, you have Yeah, people have been making objects like that for Well, centuries, many centuries.
58:37
They keep getting better at it.
58:39
There's so much sort of built knowledge built into that environment.
58:43
It expresses so much very deep expertise.
58:46
And then you have four, four and a half thousand people all completely dedicated to a single purpose.
58:52
They all care an enormous amount about this purpose.
58:55
They don't suffer like Large organizations have all kinds of bloat and all kinds of problems.
59:01
So many of those problems have gone away there, but actually because, you know, when you're on a boat, it's not so easy to empire build.
59:07
Like there's real reasons to trim fat.
59:09
Um you have amazing sort of unity of purpose.
59:14
What the captain or what the admiral There was both on on the boat say, that that goes.
59:17
You know, you're not arguing about what the right corporate strategy is.
59:21
So, you have incredible clarity.
59:21
You have incredible belief in this purpose.
59:25
It actually is a high purpose.
59:25
In their case, you know, talking to some of the sailors, you know, they felt very strongly that they were protecting a civilization that they cared about a great deal.
59:35
You know, they they speak with so much pride about it.
59:37
So, it's almost sort of the perfect floating civilization in some regards.
59:41
It's just immensely interesting.
59:46
Here's something you once wrote, and I quote, "The great talent identifiers I know or know of all seem very idiosyncratic.
59:52
They're rather like Michelin chefs."
59:56
This is getting us back to the tension between the opaque and the legible.
59:57
Why do you think that's true, that they're idiosyncratic?
1:00:04
Partly it's just because the boundaries of knowledge at any given time tend to be idiosyncratic almost by definition.
1:00:09
Like they haven't been commoditized yet.
1:00:11
There is a best, you know, person at making superconducting circuits in the world. There is a best person.
1:00:16
You you know, you're I think you Thomas Schelling I think was your PhD supervisor.
1:00:20
You know, when you read Schelling, you realize that that some of the things he did he did very well, and it must have been remarkable to talk to him.
1:00:27
But he's a little bit illegible. That's right.
1:00:30
Even though he's a very very was when you would speak to him as well. I'll bet he was, yeah.
1:00:34
But that's that's the part of the value, you you're like, oh, you know, this person is actually out on the edge of civilization.
1:00:40
And and I think that that that um uh you people who are good at identifying people who are able to expand boundaries like that like that they need to have some sense of of of that edge.
1:00:53
And here's a question you wanted me to ask you, quote, "You initially were skeptical of Emergent Ventures, but you've changed your mind and become enthusiastic about it.
1:01:00
What caused the switch, and what would you change about Emergent Ventures?"
1:01:06
I mean, the biggest single thing is just empirical.
1:01:08
I've met a bunch of EV grantees.
1:01:09
I've I've encouraged a bunch of people to apply, some of whom you've you've given grants to. Uh and they're great.
1:01:14
Um also they they were not that you have not funded the people you haven't funded in the way I might have expected.
1:01:20
I was funding I was talking to somebody.
1:01:22
I I've actually forgotten who it is, and um yeah, they clearly had some sort of socialist quite anti-libertarian ideas, and you've given them a a a grant, and I thought Might have been a mistake, of course.
1:01:34
So typically Tyler, like he's trying to figure out do they actually believe in the this idea?
1:01:38
Do they actually really care?
1:01:40
And then you you don't mind that that that you know, they're they're not they're certainly not coming to the Mercatus Center to you know, carry forward the libertarian flag.
1:01:50
Um uh uh uh and somehow I think seeing so many people who are doing very worthwhile things which have no or very little institutional chance of support being amplified.
1:02:15
That I care about a great deal.
1:02:19
And I mean, that's that seems like EV is one of the the places that is doing that the the the best that really yeah.
1:02:26
What do you think we can do to attract more non-legible but excellent people?
1:02:33
Find other people like yourself.
1:02:36
Well, they're going to be like you in this abstract way, but actually very unlike you in in in in other ways.
1:02:42
I I think about I don't know, people like Stewart Brand and people like that in the the sort of in the past who've just been wonderful at at talent identification, but they're not identifying the same kind of talent as you.
1:02:51
To some extent they they become human marketplaces as well.
1:02:55
Like they're actually they're at a crossroads.
1:02:57
They're connecting people to opportunity. Right.
1:02:59
Um and that's a very special type of a person.
1:03:05
Maybe met five to 10 people in my life who seem like that. But really strong.
1:03:10
Here's something else you wrote, and I quote, "I internalized a lot of Ivan Illich, John Holt, A. S.
1:03:17
Neill, and Paulo Freire as a kid." What did you mean?
1:03:21
You were talking I think in the context of agency, but how did that shape you?
1:03:27
At the time as a 12 or 13-year-old, it mostly probably made me insufferable to my parents because I couldn't you know, I hated school already.
1:03:34
But that's a good thing, right?
1:03:35
It gave me you know, a real way of expressing that, and they they dealt with me very patiently.
1:03:42
Um I think over the long term the most important of those was I for many years I would have said Illich. Maybe still say Illich.
1:03:52
And basically you his point is about the question it's about the question of what's the relationship between human beings and institutions, and how paternalistic are those institutions towards the humans.
1:04:07
So, in Deschooling Society, he really makes the point that in fact schools do not treat children as as human to some extent at all.
1:04:14
Denies them the most basic kinds of agency.
1:04:17
And just thinking about that kind of a relationship, what what relationship should our institutions have to individuals was very very important to me as a teenager and then sort of I mean, through my entire life.
1:04:28
Is that why you didn't join OpenAI in 2015? No.
1:04:36
And do you regret that decision?
1:04:39
Um yeah, yeah, I said I did consider uh go going as they were getting started.
1:04:44
It would have been an an interesting life choice.
1:04:47
I had a lot of doubts about the wisdom of pursuing artificial general intelligence.
1:04:55
Um which wouldn't I mean, not at all resolved then. They were just fears. Yeah.
1:04:59
Um so, that was part of the reason.
1:05:00
Honestly, really the main reason at that point though it's it's this point about comparative advantage.
1:05:03
It was like, oh, AI is happening.
1:05:05
It's become very fashionable.
1:05:06
Um you know, if you kind of wake up in the morning and it turns out that some institution is mad keen to pay you to do whatever it is you're doing, you should actually think about whether or not you're in the right line of business.
1:05:17
Um uh I tend to think unfortunately for creative work there's an anti-correlation between how valuable what you're doing is and and and what you're being paid often.
1:05:28
Well, it's kind of the it's the the anti economist point of view. It's not it's not right.
1:05:33
Um but I like there's a simple model in which that is right.
1:05:38
Actually, a very simple very much kind of an economist model.
1:05:39
Um so, I have felt for more than a decade that AI was you know, it's it become sort of around 2011, 2012 when I decided to write I started writing my book about neural nets.
1:05:51
It had become sort of this unstoppable force in the world or very difficult to to stop.
1:05:55
It was clear it was going to attract more and more capital, more and more people.
1:05:59
And so, also to some extent I just felt like I I should go and do something else if that's the case.
1:06:06
In the next 20 years, what do you think your comparative advantage will be? Tell you after 20 years.
1:06:16
But you face a year coming up now, right?
1:06:18
Yeah, the problem is There's a step up before you.
1:06:19
Probably is because it's very helpful for motivational reasons to have answers to that, but your answers never turn out to be what the the correct answer was.
1:06:29
Probably my ability to write.
1:06:32
And your Twitter biography says, and I quote, "Searching for the numinous."
1:06:36
What does that mean with respect to you?
1:06:50
Just trying to find the deepest possible experiences in the world, in people and things and ideas and places.
1:06:57
And final question, what do you think it is that you will learn about next?
1:07:03
You love to answer ask this question.
1:07:04
I've heard you ask it before.
1:07:13
I will learn more much more deeply about religion than I have in the past.
1:07:16
And that involves travel, going to church, reading books, talking to people, all of those things?
1:07:23
all of those things, yeah.
1:07:23
I'm going to see the the Hagia Sophia in Istanbul in the near future. That's gorgeous. Yeah.
1:07:29
Wanted to see for almost all my adult life, and I have never been.
1:07:32
Have you been to Amritsar? No.
1:07:36
That to me is the most religious feeling site I've ever visited.
1:07:39
So, I would recommend going there.
1:07:42
And it's not a hard trip in any way at all.
1:07:44
I've never been to India. You must go to Amritsar. Yeah.
1:07:45
And the old cliche something like when it comes to religion every Indian's a millionaire.
1:07:51
It's not really true, but I still think India is the best place to go to think about religion.
1:07:57
Yeah, that makes a lot of sense. Yes.
1:07:58
Yeah, I'd love to go to Jerusalem as well.
1:08:00
I think for somewhat similar reasons.
1:08:02
But there you think about tension. Yeah.
1:08:06
The religious aspects of tension, but you think more I think more about tension than about religion per se. Mhm.
1:08:12
And it's very useful for that.
1:08:14
Michael Nielsen, thank you very much. Thanks so much, Tyler.