Latest / Faster, Please! — The Podcast / ✨⏩ My chat (+transcript) with ... economist Robin Hanson on AI, innovation, and economic reality
Transcript
- 0:06SPEAKER_01: Welcome to Faster Please, the podcast.
- 0:09SPEAKER_01: I'm your host, Jim Pethokoukis.
- 0:12SPEAKER_01: Several times a month,
- 0:13SPEAKER_01: I'll feature a lively conversation with a fascinating and provocative guest about
- 0:18SPEAKER_01: how to make the world a better place through scientific discovery,
- 0:21SPEAKER_01: technological innovation,
- 0:23SPEAKER_01: and economic growth.
- 0:24SPEAKER_01: You're also going to want to check out my Faster Please newsletter.
- 0:27SPEAKER_01: You're on Substack throughout the week for fresh essays,
- 0:30SPEAKER_01: Q&As,
- 0:31SPEAKER_01: and stories from around the internet and around the world.
- 0:36SPEAKER_01: Today on Faster Please,
- 0:37SPEAKER_01: the podcast,
- 0:38SPEAKER_01: I talk with economist Robin Hanson about how much technological change our society
- 0:44SPEAKER_01: will undergo in the foreseeable future,
- 0:46SPEAKER_01: what form we want that change to take,
- 0:49SPEAKER_01: and how much we can ever reasonably predict.
- 0:52SPEAKER_01: Hanson is an associate professor of economics at George Mason University.
- 0:56SPEAKER_01: He was formerly a research associate at the Future of Humanity Institute at Oxford
- 1:00SPEAKER_01: and is the author of the Overcoming Bias Substack.
- 1:03SPEAKER_01: In addition,
- 1:04SPEAKER_01: he's the author of the 2017 book,
- 1:06SPEAKER_01: The Elephant in the Brain,
- 1:08SPEAKER_01: Hidden Motives in Everyday Life,
- 1:11SPEAKER_01: as well as the 2016 book,
- 1:13SPEAKER_01: The Age of M,
- 1:15SPEAKER_01: Work,
- 1:15SPEAKER_01: Love,
- 1:15SPEAKER_01: and Life When Robots Rule the Earth.
- 1:18SPEAKER_01: Robin, welcome to the podcast.
- 1:20SPEAKER_01: Nice to be back.
- 1:22SPEAKER_01: Do you think that the tech advances of recent years,
- 1:26SPEAKER_01: you know,
- 1:27SPEAKER_01: obviously in AI and what we're seeing with reusable rockets or CRISPR,
- 1:34SPEAKER_01: different energy advances,
- 1:36SPEAKER_01: fusion perhaps,
- 1:37SPEAKER_01: even Ozempic.
- 1:40SPEAKER_01: Do you think that the collective cluster of these technologies has put humanity on
- 1:47SPEAKER_01: a different path than perhaps it was on 10 years ago?
- 1:51SPEAKER_00: That's a pretty big standard.
- 1:54SPEAKER_00: You know, as you know, the world has been growing exponentially for a very long time.
- 1:59SPEAKER_00: And new technologies have been appearing for a very long time.
- 2:02SPEAKER_00: And the economy doubles roughly every 15 or 20 years.
- 2:06SPEAKER_00: And that can't happen without a whole lot of technological change.
- 2:10SPEAKER_00: So most people don't notice just how much stuff is changing behind the scenes in
- 2:13SPEAKER_00: order for the economy to double every 15 or 20 years.
- 2:17SPEAKER_00: So to say that we're going more than that is really a high standard here.
- 2:23SPEAKER_00: So I don't think it meets that standard.
- 2:27SPEAKER_00: Maybe the standard it meets is to say people were worried about maybe a stagnation
- 2:32SPEAKER_00: or slowdown a decade or two ago.
- 2:35SPEAKER_00: And I think this might weaken your concerns about that.
- 2:38SPEAKER_00: I think you might say, well, we're still on target.
- 2:44SPEAKER_00: And innovation's clumpy.
- 2:45SPEAKER_00: There are some lumpy ones once in a while, lumpier innovations than usual.
- 2:50SPEAKER_00: And those, you know, boost higher than expected sometimes, lower than expected sometimes.
- 2:55SPEAKER_00: And maybe in the last 10 years, we've had a higher than expected clump.
- 3:00SPEAKER_00: The main thing that does is make you not doubt as much as you did when you had the
- 3:04SPEAKER_00: lower than expected clump in the previous 10 years or 20 years.
- 3:08SPEAKER_00: Because people had seen this long-term history and they thought, lately we're not seeing so much.
- 3:11SPEAKER_00: I wonder if this is done.
- 3:13SPEAKER_00: I wonder if we're running out.
- 3:14SPEAKER_00: And I think the last 10 years tells you, well, no.
- 3:18SPEAKER_00: We're kind of still on target.
- 3:19SPEAKER_00: We're still having big, important advances.
- 3:23SPEAKER_00: As we have for two centuries.
- 3:25SPEAKER_01: So...
- 3:27SPEAKER_01: people who are especially enthusiastic about the recent advances with AI,
- 3:33SPEAKER_01: would you tell them that their baseline should probably be informed by economic
- 3:40SPEAKER_01: history rather than science fiction?
- 3:43SPEAKER_00: By technical history, look, we have 70 odd years of history of AI.
- 3:49SPEAKER_00: I was an AI researcher full-time from 84 to 93.
- 3:53SPEAKER_00: If you look at the long sweep of AI history, we've had some pretty big advances.
- 3:59SPEAKER_00: We couldn't be where we are now with a lot of pretty big advances all along the way.
- 4:04SPEAKER_00: You just think about the very first digital computer in 1950 or something and all
- 4:08SPEAKER_00: the things we've seen,
- 4:09SPEAKER_00: we have made large advances and they haven't been completely smooth.
- 4:14SPEAKER_00: They've come in a bit of clumps.
- 4:16SPEAKER_00: I was enticed into the field in 1984 because of a recent set of clumps then.
- 4:22SPEAKER_00: And look for a century, roughly every 30 years, we've had a burst of concern about automation and AI.
- 4:29SPEAKER_00: And we've had big concern in the sense people said, are we almost there?
- 4:33SPEAKER_00: Are we about to have pretty much all jobs automated?
- 4:36SPEAKER_00: They said that in the 1930s.
- 4:37SPEAKER_00: They said it in the 1960s.
- 4:38SPEAKER_00: There was a presidential commission in the 1960s.
- 4:41SPEAKER_00: What if all the jobs get automated?
- 4:43SPEAKER_00: I jumped in in the 80s, late 80s, when there was a big burst there.
- 4:48SPEAKER_00: And I,
- 4:48SPEAKER_00: as a young graduate student,
- 4:49SPEAKER_00: said,
- 4:50SPEAKER_00: gee,
- 4:50SPEAKER_00: if I don't get in now,
- 4:51SPEAKER_00: it'll all be over soon because I heard all the jobs are going to be automated soon.
- 4:55SPEAKER_00: And now in the last decade or so, we've had another big burst.
- 4:59SPEAKER_00: And I think people who haven't seen that history, it feels to them like it felt to me in 1984.
- 5:04SPEAKER_00: wow, unprecedented advances.
- 5:07SPEAKER_00: Everybody's really excited.
- 5:08SPEAKER_00: Maybe we're almost there.
- 5:10SPEAKER_00: Maybe if I jump in now, I'll be part of the big push over the line to just automate everything.
- 5:14SPEAKER_00: That was exciting.
- 5:15SPEAKER_00: It was tempting.
- 5:17SPEAKER_00: I was naive and I was sucked in.
- 5:20SPEAKER_00: And we're now in another era like that.
- 5:23SPEAKER_00: And yes,
- 5:23SPEAKER_00: if you're young and you haven't seen the world for,
- 5:26SPEAKER_00: you know,
- 5:26SPEAKER_00: decades,
- 5:27SPEAKER_00: you might well believe that we're almost there.
- 5:30SPEAKER_00: We're just about to automate everything, but we're not.
- 5:35SPEAKER_01: I like that you mentioned the automation scare of the 60s.
- 5:39SPEAKER_01: And just going back and looking at that,
- 5:41SPEAKER_01: it really surprised me,
- 5:44SPEAKER_01: I guess,
- 5:45SPEAKER_01: how prevalent and widespread and how serious people took that.
- 5:50SPEAKER_01: I mean,
- 5:50SPEAKER_01: you can find speeches by Martin Luther King talking about how our society is going
- 5:58SPEAKER_01: to deal with the computerization of everything.
- 6:03SPEAKER_01: So it does seem to be a recurrent fear.
- 6:07SPEAKER_01: So at what point, what would you need to see to think it is different this time?
- 6:15SPEAKER_00: Well,
- 6:16SPEAKER_00: the obvious relevant parameter to be tracking is the percentage of world income
- 6:20SPEAKER_00: that goes to automation.
- 6:23SPEAKER_00: And that has been creeping up over the decades, but it's still less than 5%.
- 6:28SPEAKER_00: What is that statistic?
- 6:30SPEAKER_00: I would say if you look at the percentage of the economy that goes to computer
- 6:33SPEAKER_00: hardware and software or other mechanisms of automation,
- 6:37SPEAKER_00: you're still looking at less than 5% of the world economy.
- 6:42SPEAKER_00: So it's been creeping up.
- 6:43SPEAKER_00: Maybe decades ago it was 3%, you know, even 1% in 1960.
- 6:45SPEAKER_00: But it's creeping up slowly.
- 6:50SPEAKER_00: And obviously when that gets to be 80%,
- 6:53SPEAKER_00: game over,
- 6:54SPEAKER_00: you've got the economy has been replaced,
- 6:56SPEAKER_00: but that number is creeping up slowly and you can track it.
- 6:59SPEAKER_00: So I will just,
- 7:00SPEAKER_00: when you start seeing that number going up much faster or becoming a large number,
- 7:04SPEAKER_00: then that's the time to say,
- 7:05SPEAKER_00: okay,
- 7:06SPEAKER_00: it looks like we're close.
- 7:07SPEAKER_00: Maybe automation will in fact take over most jobs when it's getting most of world income.
- 7:14SPEAKER_01: If you're looking at,
- 7:16SPEAKER_01: economic statistics and you're looking at different forecasts,
- 7:19SPEAKER_01: whether by,
- 7:19SPEAKER_01: you know,
- 7:20SPEAKER_01: the Fed or CBO or Wall Street banks and the forecasts are,
- 7:24SPEAKER_01: well,
- 7:25SPEAKER_01: you know,
- 7:26SPEAKER_01: we expect maybe because of AI productivity growth to be,
- 7:30SPEAKER_01: you know,
- 7:31SPEAKER_01: 0.4 percentage points higher over this kind of time or it's,
- 7:35SPEAKER_01: Those kinds of numbers where we're talking about like a tenth of a point here.
- 7:40SPEAKER_01: That's not the kind of,
- 7:41SPEAKER_01: you know,
- 7:42SPEAKER_01: singularity emergent world that,
- 7:45SPEAKER_01: you know,
- 7:46SPEAKER_01: that some people think or hope or expect that we're on.
- 7:51SPEAKER_00: Absolutely.
- 7:52SPEAKER_00: And so, you know, you've got young, enthusiastic tech people, et cetera, and they're exaggerating.
- 7:57SPEAKER_00: They're, you know, the AI companies even, they're trying to push as big a dramatic image as they can.
- 8:02SPEAKER_00: And then all the stodgy,
- 8:04SPEAKER_00: conservative old folks,
- 8:06SPEAKER_00: they're afraid of seeing behind the times and not up with things and not getting it,
- 8:10SPEAKER_00: right?
- 8:11SPEAKER_00: That was the big phrase in the
- 8:12SPEAKER_00: Internet boom.
- 8:14SPEAKER_00: Who gets it that this is a new thing?
- 8:16SPEAKER_00: Now it would be hashtag, you know, team gets it.
- 8:20SPEAKER_00: Exactly.
- 8:20SPEAKER_00: Something like that.
- 8:21SPEAKER_00: Uh-huh.
- 8:22SPEAKER_00: Well, so they're trying to lean into it.
- 8:24SPEAKER_00: They're trying to like give it the best spin they can, but...
- 8:27SPEAKER_00: They have some self-respect.
- 8:29SPEAKER_00: So they're going to give you, wow, 0.4%.
- 8:31SPEAKER_00: They'll say, that's huge.
- 8:33SPEAKER_00: Wow, this is a really big thing.
- 8:34SPEAKER_00: Everybody should be into this.
- 8:36SPEAKER_00: But they can't go above 0.4% because they've got some common sense here.
- 8:41SPEAKER_00: So,
- 8:42SPEAKER_00: you know,
- 8:42SPEAKER_00: but we've even seen management consulting firms like over the last decade or so
- 8:46SPEAKER_00: make predictions that 10 years in the future,
- 8:48SPEAKER_00: half of all jobs will be automated.
- 8:50SPEAKER_00: So we've seen this long history of these really crazy extreme predictions happening.
- 8:54SPEAKER_00: into a decade, and none of those remotely happened, of course.
- 8:59SPEAKER_00: But people do want to be in with the latest thing.
- 9:01SPEAKER_00: Look, and this is obviously the latest round of technology.
- 9:05SPEAKER_00: It's impressive.
- 9:06SPEAKER_00: I'm proud to be a human, to have done part of the civilization to have done this.
- 9:11SPEAKER_00: And I like to try them out and see what I can do with them and think of where they could go.
- 9:15SPEAKER_00: That's all exciting and fun.
- 9:18SPEAKER_00: But we've seen that for 70 years.
- 9:20SPEAKER_00: New technologies,
- 9:21SPEAKER_00: we get excited,
- 9:22SPEAKER_00: we try them out,
- 9:23SPEAKER_00: we try to apply them,
- 9:24SPEAKER_00: and that's part of our progress.
- 9:28SPEAKER_01: And not to,
- 9:29SPEAKER_01: you know,
- 9:29SPEAKER_01: just talk just about AI,
- 9:30SPEAKER_01: but do you think AI is important enough that policymakers need to somehow guide the
- 9:37SPEAKER_01: technology to a certain outcome?
- 9:39SPEAKER_01: Deron Acemoglu,
- 9:40SPEAKER_01: who,
- 9:41SPEAKER_01: one of the Nobel Prize winners,
- 9:43SPEAKER_01: has for quite some time and certainly recently said that AI
- 9:48SPEAKER_01: That this technology needs to be guided by policymakers so that it helps people,
- 9:55SPEAKER_01: it helps workers,
- 9:56SPEAKER_01: it creates new tasks,
- 9:57SPEAKER_01: it creates new things for them to do,
- 9:59SPEAKER_01: not automate away their jobs or automate a bunch of tasks.
- 10:04SPEAKER_01: Do you think that there's something special in this technology that we need to
- 10:08SPEAKER_01: guide it to some sort of outcome?
- 10:11SPEAKER_00: I think those sort of people would say that about any new technology that seemed
- 10:17SPEAKER_00: like it was going to be important.
- 10:18SPEAKER_00: They are not actually distinguishing AI from other technologies.
- 10:22SPEAKER_00: This is just what they say about everything.
- 10:24SPEAKER_00: It could be technology X. We must guide it to the outcome that I have already determined.
- 10:29SPEAKER_00: As long as you've said this X is new,
- 10:31SPEAKER_00: X is exciting,
- 10:32SPEAKER_00: a lot of things seem to depend on X,
- 10:34SPEAKER_00: then their answer would be we need to guide it.
- 10:36SPEAKER_00: It wouldn't really matter what the details of X were.
- 10:39SPEAKER_00: That's just how they think about society and technology.
- 10:42SPEAKER_00: I don't see anything distinctive about this per se in that sense.
- 10:47SPEAKER_00: I mean, other than the fact that, look, in the long run, it's huge.
- 10:52SPEAKER_00: Well,
- 10:52SPEAKER_00: look,
- 10:52SPEAKER_00: space in the long run is huge because obviously in the long run,
- 10:56SPEAKER_00: almost everything will be in space.
- 10:57SPEAKER_00: So clearly, eventually, space will be the vast majority of everything, right?
- 11:02SPEAKER_00: That doesn't mean we need to guide space now or to do anything different about it per se.
- 11:07SPEAKER_00: At the moment, space is pretty small.
- 11:09SPEAKER_00: And it's pretty pedestrian, but it's exciting.
- 11:12SPEAKER_00: I'm glad it's doing the same for AI.
- 11:13SPEAKER_00: At the moment, AI is pretty small, minor.
- 11:17SPEAKER_00: AI is not remotely threatening to cause harm in our world today.
- 11:20SPEAKER_00: If you look about harmful technologies, this is way down the scale.
- 11:24SPEAKER_00: The demonstrated harms of AI in the last 10 years are minuscule compared to things like,
- 11:31SPEAKER_00: I don't know,
- 11:31SPEAKER_00: you know,
- 11:32SPEAKER_00: construction equipment or drugs or,
- 11:35SPEAKER_00: you know,
- 11:36SPEAKER_00: Even television, really.
- 11:38SPEAKER_00: I mean, this is small.
- 11:40SPEAKER_01: Ladders for climbing up on your roof to clean out the gutters?
- 11:44SPEAKER_01: That's a very dangerous technology.
- 11:46SPEAKER_00: Yeah, somebody should be looking into that.
- 11:47SPEAKER_00: We should be guiding the ladder industry to make sure they don't cause harm in the world, okay?
- 11:52SPEAKER_01: I'm not sure how much confidence we should ever have on long-term economic forecasts.
- 11:56SPEAKER_01: But have you seen any reason to think that they might be sort of less reliable than
- 12:02SPEAKER_01: they always have been,
- 12:03SPEAKER_01: that we might be approaching some sort of change that,
- 12:08SPEAKER_01: boy,
- 12:09SPEAKER_01: those 50-year forecasts of entitlement spending might be all wrong because the
- 12:14SPEAKER_01: economy is going to be growing so much faster or longevity is going to be
- 12:19SPEAKER_01: increasing so much faster?
- 12:21SPEAKER_00: The highest level thing to say is,
- 12:23SPEAKER_00: It was just a little over two centuries ago when the world saw this enormous revolution.
- 12:29SPEAKER_00: Previously, the world had been doubling roughly every thousand years.
- 12:33SPEAKER_00: And that had been going on for maybe 10,000 years.
- 12:36SPEAKER_00: And then within the space of a century, we switched to doubling roughly every 15 or 20 years.
- 12:42SPEAKER_00: That's a factor of 60 increase in the growth rate.
- 12:46SPEAKER_00: And it happened after a previous transition from foraging to farming, roughly 10 doublings before.
- 12:53SPEAKER_00: So you might say,
- 12:54SPEAKER_00: look,
- 12:55SPEAKER_00: we can't trust these forecasts more than,
- 12:58SPEAKER_00: these trends to continue maybe more than 10 doublings,
- 13:00SPEAKER_00: and then who knows what might happen.
- 13:02SPEAKER_00: So you could just say, well, that's 200 years, say, if you double every 20 years.
- 13:06SPEAKER_00: You could say, look, we just can't trust these forecasts more than 200 years now.
- 13:10SPEAKER_00: Look at what's happened in the past.
- 13:13SPEAKER_00: After that many doublings, big changes happened.
- 13:15SPEAKER_00: And you might say, therefore, expect on that sort of timescale something else big to happen.
- 13:20SPEAKER_00: That's not crazy to say.
- 13:22SPEAKER_00: That's not very specific.
- 13:24SPEAKER_00: And then if you say,
- 13:25SPEAKER_00: well,
- 13:25SPEAKER_00: what is the thing people most often speculate could be the cause of a big change?
- 13:29SPEAKER_00: They do say AI.
- 13:31SPEAKER_00: And then we actually have a concrete reason to think AI would change the growth rate of the economy.
- 13:37SPEAKER_00: So that is the fact that at the moment we make most stuff in factories and
- 13:42SPEAKER_00: factories typically push out from the factory as much value as the factory itself
- 13:48SPEAKER_00: embodies in economic terms in a few months.
- 13:52SPEAKER_00: So that says if you could have factories make factories, the economy could double every few months.
- 13:58SPEAKER_00: The reason we can't now is we have humans in the factories and factories don't double them.
- 14:02SPEAKER_00: But if you could make AIs in factories,
- 14:04SPEAKER_00: then the AIs made factories that made more AIs,
- 14:09SPEAKER_00: that could double every few months.
- 14:10SPEAKER_00: So the world economy could plausibly double every few months when AIs dominated the economy.
- 14:16SPEAKER_00: And that's
- 14:18SPEAKER_00: of the magnitude doubling every few months versus doubling every 20 years.
- 14:23SPEAKER_00: That's a magnitude similar to the magnitude we saw before from farming to industry.
- 14:29SPEAKER_00: And so that fits together as saying sometime in the next few centuries,
- 14:33SPEAKER_00: expect a transition that might increase the growth rate of the economy by a factor
- 14:37SPEAKER_00: of 100.
- 14:38SPEAKER_00: Now that's an abstract thing in the long-term.
- 14:41SPEAKER_00: It's not the next 10 years or 20 or something.
- 14:43SPEAKER_00: It's saying that, look, economic modes only last so long.
- 14:48SPEAKER_00: Something should come up eventually.
- 14:49SPEAKER_00: And this is our best guess of a thing that could come up.
- 14:52SPEAKER_00: So it's not crazy.
- 14:54SPEAKER_01: Are you a fertility rate worrier?
- 14:57SPEAKER_00: I am.
- 14:57SPEAKER_00: And in fact,
- 15:00SPEAKER_00: I think we have a limited deadline to develop,
- 15:03SPEAKER_00: say,
- 15:03SPEAKER_00: human level AI,
- 15:05SPEAKER_00: before which we won't for a long pause because falling fertility really threatens
- 15:11SPEAKER_00: innovation rates.
- 15:13SPEAKER_00: So this is something we economists understand that I think most other people don't.
- 15:18SPEAKER_00: You might have thought that a falling population could be easily compensated by a
- 15:22SPEAKER_00: growing economy and that we would still have,
- 15:24SPEAKER_00: you know,
- 15:25SPEAKER_00: rapid and fast innovation because we would just have a bigger economy with a lower population.
- 15:30SPEAKER_00: But apparently that's not true.
- 15:33SPEAKER_00: If the population falls, the best models say innovation rates would fall.
- 15:39SPEAKER_00: even faster.
- 15:40SPEAKER_00: So say populations roughly predicted to peak in say three decades and then start to fall.
- 15:46SPEAKER_00: And if it falls,
- 15:47SPEAKER_00: it would fall roughly a factor of two every generation or two,
- 15:53SPEAKER_00: depending on which populations dominate.
- 15:56SPEAKER_00: And then if it fell by say a factor of 10, the innovation rate would fall by more than a factor of 10.
- 16:03SPEAKER_00: And that means just a slower rate of new technologies.
- 16:07SPEAKER_00: And of course, also a reduction of the scale of the world economy.
- 16:11SPEAKER_00: And I think that plausibly also has the side effect of a loss in liberality.
- 16:18SPEAKER_00: I don't think people realize how much it was innovation and competition that drove
- 16:22SPEAKER_00: much of the world to become liberal because the winning nations in the world were
- 16:26SPEAKER_00: liberal and the rest were afraid of falling too far behind.
- 16:30SPEAKER_00: But when innovation goes away,
- 16:33SPEAKER_00: they won't be so eager to be liberal to be innovative because innovation just won't be a thing.
- 16:36SPEAKER_00: And so much of the world will just become a lot less liberal, I think.
- 16:41SPEAKER_00: And there's also the risk that basically computers are very durable technology in principle.
- 16:47SPEAKER_00: So typically we don't make them that durable because every two years, you know, they get twice as good.
- 16:52SPEAKER_00: But when innovation goes away, they won't get good very fast.
- 16:57SPEAKER_00: And then you'll be much more tempted to just make very durable computers.
- 17:01SPEAKER_00: And the first generation that makes very durable computers that last hundreds of
- 17:04SPEAKER_00: years say,
- 17:05SPEAKER_00: well,
- 17:06SPEAKER_00: the next generation won't want to buy new computers.
- 17:08SPEAKER_00: They'll just use the old durable ones as the economy is shrinking.
- 17:11SPEAKER_00: And then the industry that manufactures computers might just go away.
- 17:16SPEAKER_00: And then it could be a long time before people felt a need to rediscover those
- 17:20SPEAKER_00: technologies or difficulty.
- 17:21SPEAKER_00: And so,
- 17:22SPEAKER_00: and I think the larger scale story is there's no obvious process that would prevent
- 17:27SPEAKER_00: this continued decline
- 17:29SPEAKER_00: Because there's no level at which when you get that,
- 17:32SPEAKER_00: some process kicks in and that makes us say,
- 17:34SPEAKER_00: oh,
- 17:34SPEAKER_00: we need to increase the population.
- 17:36SPEAKER_00: But the most likely scenario is just that the Amish and Haredim and other insular
- 17:41SPEAKER_00: fertile subgroups who have been doubling every 20 years for a century will just
- 17:45SPEAKER_00: keep doing that and then come to dominate the world,
- 17:48SPEAKER_00: much like Christians took over the Roman Empire.
- 17:50SPEAKER_00: They took it over by doubling every 20 years for three centuries.
- 17:55SPEAKER_00: And that's my default future.
- 17:57SPEAKER_00: And then if we don't get AI or colonized space before this decline,
- 18:02SPEAKER_00: which I've estimated would be roughly 70 years worth more of progress at previous rates,
- 18:07SPEAKER_00: then we don't get it again until the Amish and Redem not only just take over the world,
- 18:11SPEAKER_00: but rediscover a taste for technology and economic growth.
- 18:16SPEAKER_00: And then eventually all of the great stuff could happen.
- 18:19SPEAKER_00: But that could be many centuries later.
- 18:22SPEAKER_01: Yes, this does not sound like an issue that can be fundamentally altered by tweaking the tax code.
- 18:32SPEAKER_01: You'd have to make a large turn to the dial, really turn that dial.
- 18:35SPEAKER_00: And people are uncomfortable with larger than small tweaks,
- 18:38SPEAKER_00: of course,
- 18:39SPEAKER_00: but we're not in an era that's at all eager for vast changes in policy.
- 18:43SPEAKER_00: We are in a pretty conservative era that just wants to tweak things.
- 18:47SPEAKER_00: And so, yeah, Greeks won't do it.
- 18:49SPEAKER_01: We can't do things like daylight savings time, which some people want to change.
- 18:52SPEAKER_01: You mentioned about like this window.
- 18:54SPEAKER_01: Elon Musk has talked about a window, a window for expansion into space.
- 18:59SPEAKER_01: And this is a couple of years ago.
- 19:00SPEAKER_01: He said the windows closed before.
- 19:02SPEAKER_01: Don't it's open now.
- 19:04SPEAKER_01: Don't assume it will always be open.
- 19:08SPEAKER_01: Is that right?
- 19:09SPEAKER_01: Like, why would I mean, why would it close?
- 19:10SPEAKER_01: You know, because of higher interest rates, because the Amish don't want to go to space.
- 19:15SPEAKER_01: Why?
- 19:15SPEAKER_01: Why would the window close?
- 19:17SPEAKER_00: There's a demand for space stuff mostly at the moment to service Earth.
- 19:21SPEAKER_00: Like, you know, the internet circling the Earth, say, is Elon's big project to fund his spaceships.
- 19:29SPEAKER_00: And there's also demand for, you know, satellites to do surveillance of the Earth, etc.
- 19:35SPEAKER_00: As the Earth economy shrinks, the demand for that stuff will shrink.
- 19:40SPEAKER_00: So, you know, at some point they won't be able to afford fixed costs.
- 19:45SPEAKER_00: So a big question is about, as you know, marginal cost versus fixed costs.
- 19:50SPEAKER_00: How much is the fixed cost just to have this capacity to send stuff into space
- 19:54SPEAKER_00: versus the marginal cost of adding each new rocket?
- 19:58SPEAKER_00: If it's dominated by marginal costs and they make the rockets cheaper,
- 20:01SPEAKER_00: OK,
- 20:01SPEAKER_00: they can just do fewer rockets less often and they can still send satellites up
- 20:05SPEAKER_00: into space.
- 20:06SPEAKER_00: But if you're thinking of something where there's a key scale that you need to get
- 20:10SPEAKER_00: past even to support this industry,
- 20:12SPEAKER_00: then there's a different thing.
- 20:13SPEAKER_00: So I think like thinking about a Mars economy or even a moon economy or a solar
- 20:17SPEAKER_00: system economy,
- 20:18SPEAKER_00: you're looking at a scale thing.
- 20:20SPEAKER_00: That thing needs to be big enough to be self-sustaining and economically cost
- 20:25SPEAKER_00: effective or it's just not going to work.
- 20:28SPEAKER_00: So I think,
- 20:29SPEAKER_00: unfortunately,
- 20:30SPEAKER_00: we've got a limited window to try to jumpstart a space economy before the Earth
- 20:37SPEAKER_00: economy shrinks.
- 20:39SPEAKER_00: and isn't getting much value from a space economy.
- 20:42SPEAKER_00: Space economy needs to be big enough just to support itself, etc.
- 20:46SPEAKER_00: And that's a problem because it's the same humans in space who are down here on
- 20:48SPEAKER_00: Earth are going to have the same fertility problems up there unless they somehow
- 20:52SPEAKER_00: figure out a way to make a very different culture.
- 20:53SPEAKER_00: So a lot of people just assume,
- 20:55SPEAKER_00: oh,
- 20:55SPEAKER_00: you could have a very different culture on Mars and so they could solve...
- 20:59SPEAKER_00: our cultural problems just by being different but i'm not seeing that i think they
- 21:03SPEAKER_00: would just have a very strong interconnection with earth culture because they are
- 21:06SPEAKER_00: going to have just a rapid bandwidth stuff back and forth and their fertility
- 21:11SPEAKER_00: culture and all sorts of other culture will be tied closely to earth culture so i'm
- 21:16SPEAKER_00: not seeing how a mars colony really solves earth cultural problems
- 21:22SPEAKER_01: The average person is aware that these things,
- 21:24SPEAKER_01: whether it's betting markets or these online consensus prediction markets,
- 21:28SPEAKER_01: like that they exist,
- 21:30SPEAKER_01: that you can bet on presidential races and you can make predictions about a
- 21:37SPEAKER_01: superconductor breakthrough or something like that,
- 21:39SPEAKER_01: or about when we're going to get AGI.
- 21:41SPEAKER_01: It seems to me,
- 21:42SPEAKER_01: it seems like they have to some degree,
- 21:45SPEAKER_01: like broken through like the filter and people are aware that they're out there.
- 21:50SPEAKER_01: Have they come of age?
- 21:53SPEAKER_00: Well, in this presidential election, there's a lot of discussion that points to them.
- 21:59SPEAKER_00: And people were pretty open to that until Trump started to be favored and people said,
- 22:03SPEAKER_00: no,
- 22:03SPEAKER_00: no,
- 22:03SPEAKER_00: that can't be right.
- 22:04SPEAKER_00: There must be a lot of whales out there manipulating because it couldn't be Trump's winning.
- 22:08SPEAKER_00: So, you know, the openness to these things often depends on what their message is.
- 22:14SPEAKER_00: But honestly, the big value here isn't going to be betting on elections.
- 22:18SPEAKER_00: It's going to be organizations using them to make organization decisions.
- 22:21SPEAKER_00: And that process is being explored.
- 22:24SPEAKER_00: So 25 years ago, I invented this concept of decision markets using organizations.
- 22:29SPEAKER_00: And now in the last year, I've actually seen substantial experimentation with them.
- 22:32SPEAKER_00: And so I'm excited to see where that goes.
- 22:36SPEAKER_00: And I'm hopeful there, but that's not so much about the presidential markets.
- 22:40SPEAKER_00: I don't know if you know that roughly a century ago,
- 22:43SPEAKER_00: there was more money bet in presidential betting markets than in stock markets at
- 22:47SPEAKER_00: the time.
- 22:48SPEAKER_00: Betting markets were very big then,
- 22:49SPEAKER_00: and then they declined primarily because scientific polling was declared a more
- 22:55SPEAKER_00: scientific approach to estimating elections than betting markets.
- 22:58SPEAKER_00: And all the respectable people wanted to report on scientific polls.
- 23:03SPEAKER_00: And then, of course, the stock market became much, much bigger.
- 23:05SPEAKER_00: So,
- 23:06SPEAKER_00: you know,
- 23:07SPEAKER_00: the interest in presidential markets will wax and wane,
- 23:10SPEAKER_00: but there's actually not that much social value in having a better estimate of
- 23:13SPEAKER_00: who's going to win an election.
- 23:14SPEAKER_00: It doesn't really tell you who to vote for.
- 23:17SPEAKER_00: So there are other markets that would be much more socially valuable,
- 23:21SPEAKER_00: like predicting the consequences of who's elected as president.
- 23:25SPEAKER_00: We don't really have much markets on those, but maybe we will next time around.
- 23:28SPEAKER_00: But there is a lot of experimentation going in organizational prediction markets at
- 23:34SPEAKER_00: the moment compared to say 10 years ago.
- 23:36SPEAKER_00: And I'm excited about those experiments.
- 23:39SPEAKER_01: I want a culture that when
- 23:43SPEAKER_01: What are these new nuclear reactor or these nuclear reactors are restarting or
- 23:47SPEAKER_01: these new like small modular reactors when there's some sort of leak and or when a
- 23:54SPEAKER_01: or when a new SpaceX starship,
- 23:57SPEAKER_01: when some astronaut gets killed,
- 24:00SPEAKER_01: that we just don't collapse as a society,
- 24:02SPEAKER_01: that we're like,
- 24:03SPEAKER_01: well,
- 24:03SPEAKER_01: things happen.
- 24:04SPEAKER_01: We're going to keep moving forward.
- 24:06SPEAKER_01: One, do you think we have that kind of culture?
- 24:09SPEAKER_01: If not, how do we get it?
- 24:10SPEAKER_01: If at all, is that possible?
- 24:13SPEAKER_00: Well, that's the question.
- 24:14SPEAKER_00: Why has our society become so much more safety-oriented in the last half century?
- 24:19SPEAKER_00: And certainly one huge sign of it is the way we way over-regulated nuclear energy.
- 24:25SPEAKER_00: but we've also now been over regulating you know even kids going to school
- 24:29SPEAKER_00: apparently they can't just take their bikes to school anymore they have to go on a
- 24:33SPEAKER_00: bus because that's safer and just in a whole bunch of ways we are just vastly more
- 24:38SPEAKER_00: safety oriented and that seems to be a pretty broad cultural trend it's not in just
- 24:43SPEAKER_00: a particular areas and it's not in just in particular countries um so i've been
- 24:47SPEAKER_00: thinking a lot about long-term cultural trends and trying to understand them
- 24:51SPEAKER_00: The basic story,
- 24:52SPEAKER_00: I think,
- 24:52SPEAKER_00: is we don't have a good reason to believe long-term cultural trends are actually
- 24:55SPEAKER_00: healthy when they are sort of shared trends of norms and status markers that
- 25:01SPEAKER_00: everybody shares.
- 25:02SPEAKER_00: Cultural things that can vary within cultures,
- 25:04SPEAKER_00: like different technologies and firm cultures,
- 25:06SPEAKER_00: those were doing great.
- 25:07SPEAKER_00: We have great evolution of those things, and that's why we're having all these great technologies.
- 25:11SPEAKER_00: But things like safetyism is more of a shared cultural norm.
- 25:17SPEAKER_00: And I don't think we just, we just don't have good reasons to think those changes are healthy.
- 25:22SPEAKER_00: and they don't fix themselves.
- 25:24SPEAKER_00: So this is just another example of something that's going wrong.
- 25:27SPEAKER_01: They don't fix themselves because if you have sort of a strong,
- 25:32SPEAKER_01: very widely shared cultural norm,
- 25:34SPEAKER_01: and someone has a different idea,
- 25:36SPEAKER_01: they need to be prepared to pay a price,
- 25:38SPEAKER_01: and most of us aren't prepared to pay that price.
- 25:40SPEAKER_00: Right.
- 25:41SPEAKER_00: For example, we have 200 countries.
- 25:43SPEAKER_00: If they were actually independent experiments and had just had different cultures
- 25:46SPEAKER_00: going different directions,
- 25:47SPEAKER_00: then I'd feel great that,
- 25:48SPEAKER_00: okay...
- 25:50SPEAKER_00: The cultures that choose too much safety,
- 25:51SPEAKER_00: they will lose out to the others,
- 25:53SPEAKER_00: and eventually it'll be worn out.
- 25:55SPEAKER_00: If we had a healthy cultural evolution competition among even nations, this would be fine.
- 26:01SPEAKER_00: The problem is we have this global culture, a monoculture really, that enforces everybody.
- 26:06SPEAKER_00: So remember COVID, of course.
- 26:09SPEAKER_00: At the beginning of COVID, all the usual public health efforts said all the usual things.
- 26:13SPEAKER_00: And then world elites got together and talked about it.
- 26:16SPEAKER_00: And a month later, they said, no, that's all wrong.
- 26:18SPEAKER_00: We have a whole different thing to do.
- 26:20SPEAKER_00: Travel restrictions are good.
- 26:22SPEAKER_00: Masks are good.
- 26:23SPEAKER_00: Distancing is good.
- 26:24SPEAKER_00: And then the entire world did it the same way.
- 26:26SPEAKER_00: And there was strong pressure on any deviation,
- 26:28SPEAKER_00: even Sweden,
- 26:29SPEAKER_00: that would dare to deviate from the global consensus.
- 26:32SPEAKER_00: And if you look about many kinds of regulation,
- 26:35SPEAKER_00: It's very little deviation worldwide.
- 26:37SPEAKER_00: We don't have 200 or even 100 independent policy experiments.
- 26:41SPEAKER_00: We basically have a main global civilization that does it the same and maybe one or
- 26:46SPEAKER_00: two deviants that are allowed to have somewhat different behavior but pay a price
- 26:51SPEAKER_00: for it.