Joel Mokyr on China, India, and Europe; clans vs. corporations; the idea of progress.

A Conversation with Tyler: Joel Mokyr on Clans, Corporations, and a Culture of Growth (Ep. 282)

Joel Mokyr co-won the 2025 economics Nobel for exploring the question that traces back to the beginning of economics: how did sustained economic growth suddenly become normal? For nearly all of human history, cleverness didn’t compound. What changed, according to Mokyr, was twofold: first, you need to know why something works, so that one advance can seed the next; second, you need a culture willing to tolerate the disruption. His new book contrasts Europe with China, showing how Europeans learned to cooperate with people they weren’t related to, in guilds, monasteries, cities, and universities, while China organized itself around the extended clan. One path led to internal stability and peace; the other, more restless and outward-looking, was the one that decided the world could always be made better.

Tyler and Joel discuss European corporations vs. Chinese clans, why the Catholic Church became obsessed with cousin-marriage, how persistent cultural trends really are, why Chinese cities became so populous relative to Europe, why it took so long for European living standards to surpass China’s, why sinified invaders kept getting swallowed by the dynasties they conquered, how geography kept Europe fragmented and China unified, where India fits into the story, why the Romans never made spectacles, why British soldiers stood two inches taller than the French, what powered the sudden rise of 19th-century German science, how disruptive winning a Nobel is, and much more.

Clans vs. Corporations

COWEN: Start by telling us, in the book, your thesis about European corporations versus Chinese clans and the importance of that difference. How would you explain it?

MOKYR: Well, the difference is, basically, the kind of organizations that produce what we call local public goods, so things like food relief and education, religious services, things like that, I would think that if you look at the world, say, around 800, at the time of Charlemagne, the difference between Europe and China isn’t very large. At some point, during the Middle Ages, you can see this divergence getting started. What’s happening is that, in Europe, there is more and more of a decline in the extended family or the extended kinship group, we call it clan, and instead, people get together and cooperate with other people to whom they are not related and with whom they do not share an ancestor.

Whereas in China, it moves exactly in the other direction. In China, you get more and more people getting organized by their extended family. The reasons for that are fairly complex. In Europe, it’s particularly the Catholic Church that played a major role here. This was argued quite a while ago by a guy, an anthropologist called Jack Goody, but your own colleague Jonathan Schulz wrote, what I think is one of the best papers on the subject, who pointed this out in great length and actually provided a fair amount of systematic evidence for this.

In China, there is no Catholic Church. The imperial bureaucracy is more and more in cahoots with local clans to whom they actually outsource a fair amount of the things that they were supposed to do. As you move on out of this period of the Song dynasty into later dynasties, you see this thing growing. The problem in Europe is that the nuclear family, which became the fundamental building block of society, is too small to provide local public goods. You need to cooperate with others. What emerges in Europe, and quite spontaneously, is a bunch of things that provide these local public goods that you just don’t see in China.

For instance, we have something called universities. We have monasteries. We have autonomous cities. All of those things are what we call corporations. What it is, is people who are not related, but what they share is not an ancestor but an objective. Guilds have one kind of objective, universities have another one, and so on and so forth. That divergence in social organization turns out, in our view, to be one of the key components of the divergence between Europe and China.

COWEN: Do you take that change in policy from the Catholic Church as exogenous or that it is rooted in earlier features of Western society such as the ideology of Christianity itself or maybe earlier Roman times? What is causing what? What’s the most fundamental driver here in the West?

MOKYR: Well, there’s some debate about that. They don’t, of course, tell you. There are, I think, two components about this, and I would not know how to weight them. There are two things happening here. The first is that the Catholic Church really becomes quite obsessive about certain sins that they consider to be particularly egregious. One of those sins is incest. Every society in the world prohibits marriage between siblings, but other more remote relatives is more ambiguous. The Church becomes quite obsessive about this. At the end, there are places where they actually prohibit the marriage of fifth-degree cousins. Now, how anybody in the Middle Ages would know who is fifth-degree cousin is, is unclear. I don’t know who my fifth-degree cousins are. Maybe you do.

The other thing, which is also sort of Goody’s argument, and I think there’s a great deal of truth to that, is that the Church wants to weaken any kind of organizations that compete with it for power and control in the local communities. They also have figured out that if you organize society by nuclear families, then a certain proportion of people die without heirs. If they die interstate without heirs, that in many cases, the property that these people own reverted to the Church. Pure naked greed by the Church, which was not unknown in the Middle Ages, I think, was a driver here. Goody convinced me that that actually is a substantial factor.

There are other arguments that have been made in this context, but I think these are the two that are most striking. The other question is, why is it that the clan in China is so convenient for the imperial bureaucracy to rely on? Both of those things happen in parallel. I would think, maybe, if I may allow one observation, in both cases, this is history as historical outcomes, as the unintended and unanticipated consequence of very different actions. There’s no question that the Church never, I think, foresaw the emergence of corporations in Europe, nor do I think that the Chinese imperial service ever seriously considered the possibility how this would change life in China. That’s what happens. You try to follow one objective and then something very different emerges over time. That’s what history is all about.

European Growth

COWEN: Why does it take so long for the wealthiest parts of Western Europe to surpass Chinese living standards? Say that’s happened by 1700 or 1720, that’s many centuries after this medieval divergence. If it takes so many centuries, is the medieval divergence really the relevant factor? Why is it such a slow process?

MOKYR: Yes, I think it is. I think it’s a main factor. I think the idea of looking at standard of living, one thing, I’m very skeptical about how standards of living are actually measured. I know that this is what Pomeranz and other people have, and Jack Goldstone and other people have argued that the living standards in China were comparable to the West as late as 1750. I’m not 100 percent sure that that is true. Certainly, for my money, what really defines the divergence is that, technologically, the gap between the two countries starts to become visible at the time of the Renaissance, in terms of a whole bunch of things that you see growing in Europe and stagnant in China.

Now, keep in mind, of course, that part of the European growth is due to the fact that they borrowed ideas from China. Then the Industrial Revolution consists, to some extent, of imports institution by Europeans trying to mimic the goods that they were importing from China—not just from China, from India as well. Pottery is a good example. One of the things they really wanted from China was Chinaware. That’s why it’s called Chinaware. It took them a while to be able to match the Chinese capability in the ceramic industry, but they do so eventually. Then they stop importing this stuff from China. The same is true for, say, cotton and other products that we’re getting from the East.

European living standards, I think, should be measured, in part, by the fact that when the Europeans start their voyages across the globe in the late 15th and early 16th century, they are able to bring in a whole bunch of new crops and new techniques from other areas which they merely adopt. You’ll see Europeans very soon growing tobacco and potatoes and corn and other things like that. They are the agents of global change. Not only that they change their own diets, they change the Chinese diets because the Europeans bring from the New World things like peanuts and sweet potatoes and things like that. They change the Chinese diets, but the Chinese themselves are not agents here.

They are accepting the stuff that the Europeans did to some extent, and they’re rejecting others, but it’s the Europeans who are the agents of change here. They are the entrepreneurs. They are the people who bring about the changes, Tyler. My sense is that typifies the difference between Europeans and the Chinese. Europeans are more aggressive. They are more outward-looking. In the end, what you see by the 1830s and 1840s, you see that the technological gap is huge, in some ways much larger than the living standards gap. Even in the 19th century, in terms of food, the Chinese were capable of producing enough food. The number of famines in China is probably not a lot worse than in Europe.

When you see what happens during the First Opium War, one English ship is blowing all of this sort of mighty empire to pieces, and the Chinese have to accept this terribly humiliating peace, you can sort of see how the technological gap has grown between the two. For me, that is much more telling than the living standards. The other thing that I should like to point out is that, when you look at Europe in the 16th and 17th century, you can see that the capability of expanding the set of useful knowledge, including science, is just growing very rapidly. Whether there is a scientific revolution or not is a debate that I want to get into.

Certainly, by 1700, Europe is on the verge of really changing our understanding of how creation works. That’s not just Newton and Galileo. There’s a whole body of work that is emerging. There’s really nothing parallel like that in China. China is a very sophisticated society in many ways. The literacy rates are high. They have a well-funded and well-organized system of education, but they don’t really continue their earlier forays into science and into new technology.

Somebody actually went out and looked at Joseph Needham’s many volumes on Chinese technology and science, or Science and Civilisation [in China], as he called it, and he discovered something—which I guess we all knew, but they put numbers on it—almost nothing that Needham pointed out as an innovation happens after 1400. There’s complete stagnation setting in and some of the things that they knew how to make in earlier times, like the sophisticated clocks that they built in the 11th century, they disappear. For me, that’s more telling than how many calories of carbohydrates were consumed on average, if we could ever calculate that correctly.

Alexandria as the Manhattan of the ancient world, Rome, writing (Rank 2 culture)

Tyler Cowen, Toby Wilkinson on Ptolemaic Egypt and the First Great Commercial Civilization (Ep. 278), May 27, 2026.

Toby Wilkinson is one of the world’s leading Egyptologists, whose books have ranged across the full sweep of pharaonic history. His latest, The Last Dynasty: Ancient Egypt from Alexander the Great to Cleopatra, covers the 300-year Ptolemaic period — stranger and more modern-feeling than the Egypt of the pyramids, built around commerce and cosmopolitanism rather than divine kingship, and home to the greatest concentration of scientific talent the ancient world ever saw.

Tyler and Toby cover how Alexander took over the empire almost without a fight, why Alexandria became the Manhattan of the ancient world, whether the era was as philosophically fertile as it was scientifically, whether your ancient doctor’s visit had positive expected value, what Egypt was actually exporting and selling, whether living standards rose above subsistence or stayed Malthusian, how the ethnic divide between Greek rulers and Egyptian subjects shaped society, what constrained the Ptolemaic Empire from becoming the next Rome, whether Cleopatra has been overhyped, what Julius Caesar was really thinking when he sided with her over her brother, the new frontiers in archeology, whether Herodotus can be trusted, what ancient Egypt knew about Israel and India, when Egyptian jewelry peaked and why, what triggered the sudden emergence of civilization across the ancient world, why a six-year-old Tyler knew King Tut better than Napoleon, and much more.

After the preliminaries, the first of three segments from the whole conversation:

I. On intellectual activity of Alexandria

COWEN: How large was the library in Alexandria and how did they build that out?

WILKINSON: This is interesting because the Ptolemaic kings not only wanted to be wealthy economically, but they wanted to be renowned throughout the ancient world as great scholar leaders. They thought it was important that a new dynasty establish its credentials as a patron of the arts and of learning, not just as the head of a great commercial enterprise.

They invited all the leading scholars from the Greek-speaking world, from right across the Mediterranean to Alexandria, where they provided them with a library, with what was called a museum or a temple of the muses, a place where scholars could think great thoughts and be well looked after. The library was developed over centuries, really, as the greatest repository of learning that the world had ever seen up to this point. It is thought that maybe at its height, it contained half a million volumes, half a million manuscripts, mostly written on papyrus, but representing really the sum total of human knowledge at that time.

COWEN: Euclid and Eratosthenes are connected to this era?

WILKINSON: Almost any big name from ancient science has some connection with Alexandria. Euclid, the mathematician, studied there. Eratosthenes, who quite amazingly calculated the circumference of the earth, he carried out those calculations in Alexandria. There were leaders in the fields of astronomy, of anatomy and medicine, of geography, of philosophy, of literary theory. They were all active in Alexandria under the patronage of the Ptolemaic kings.

COWEN: Am I correct in thinking that the era was quite weak in philosophy in some ways? There’s no great name. Maybe it’s all lost manuscripts. It seems to be a bit like China, infrastructure intensive. They build amazing things. It’s commercial, but they’re not really thinkers.

WILKINSON: I think that would be a misrepresentation. Alexandria under the Ptolemies didn’t produce a Socrates or a Plato. It is true. What its scholars did was to synthesize strands of philosophy from ancient Greek thought, ancient Egyptian thought, Babylonian thought, ancient Hebrew philosophy and religion. It was a melting pot. It maybe didn’t throw up the big name, but it was a very fertile ground for the exchange and the interchange of ideas.

COWEN: Say I try to read Diodorus. There’s a lot of detail, but it’s not to me very interesting. It just seems much worse than Herodotus, who is profound and in a way a step back. Maybe that’s not representative.

WILKINSON: I think there are a lot more scholarly works, both surviving and lost, that were composed in ancient Alexandria that would be more surprising and more revelatory than those that you’ve just mentioned.

COWEN: What was it exactly that was so special about the intellectual, and scientific, and productive environment of Alexandria and environs? Was it that the Egyptians were there, or the mix of Greeks and Egyptians, or something else? What?

WILKINSON: I think there are three factors, really. One is certainly the Egyptian context. I don’t think those intellectual advances that were made in Alexandria could have been made anywhere else. Let’s take anatomy, for example. In the Greek world in general, there was a taboo on cutting up human bodies. Of course, ancient Egypt had a long tradition of mummification, which involved dissecting human corpses. If you were an anatomist and you wanted to make discoveries about how the organs functioned, the only place you could do that at this time was ancient Egypt. The same is true, actually, in many other branches of science. The Egyptian traditions of scholarship and of learning really laid the foundations for Greek thinkers to take them to the next level.

The second aspect was the wonderful infrastructure that was put in place by the Ptolemies to lure scholars to Alexandria. They were paid handsomely. They had access to the world’s best library. They had all of the facilities at their disposal. There was really no better place to be than Alexandria.

The third factor was really the kleptomania of the Ptolemaic rulers. They were not just bibliophiles, but they wanted to acquire a copy, preferably the original, of every book and manuscript circulating in the ancient Greek world. To that end, they indulged in downright thieving. They sent a word, for example, to Athens, which was one of the great centers of scholarship, a rival center of learning, and requested copies of books from Athens’ city library. The copies arrived in Alexandria. They were then seized by the Ptolemaic authorities and kept for Alexandria’s own library. They were only too happy to pay the fine because they had the books. It was a combination of factors that really led Alexandria to being the greatest center for scholars and for scholarship.

II. From the Ptolemaic Empire to Rome

COWEN: Given all the successes, what should I think of as the limiting principle behind rule here? It stretches as far as what we would call Cyprus today, but it never becomes a very large area, right? Now, it doesn’t become the next Roman Empire. Why doesn’t it?

WILKINSON: At its greatest extent, the Ptolemaic Empire includes much of the coast of modern-day Libya, certainly Cyprus, the Nile Valley, parts of present-day Lebanon, and Syria and Turkey, and some islands in the Aegean. It’s not as big as the Roman Empire would later be. What are its constraining factors? Partly, it’s a lack of ambition. Not a lack of ambition, moderate ambition.

The Ptolemies really want to rule Egypt, which is regarded as the jewel in Alexander the Great’s crown. It’s the most prosperous part of his empire. The other territories that they conquer in a ring around Egypt are really only there as a defensive buffer zone to protect Egypt. They have no particular ambition to create a world empire as Alexander the Great did. That was one constraining factor.

They are also not the only players in the ancient world. There are other powerful dynasties of kings in Asia and in the Greek mainland who have territorial designs of their own. It keeps these various powers in equilibrium, and it stops one of them becoming dominant, really, until the Romans upend the whole system. I suppose, yes, those are the two constraining factors, ambition and competition.

COWEN: Do you think the Romans had the ambition in a way Ptolemaic Egypt did not?

WILKINSON: Yes, I do. I think the Romans were motivated by a real desire to conquer. Ancient Roman military leaders were only as good as their last victory on the battlefield. You see that in the later days of the Republic and the beginning of the empire. Whereas in ancient Egyptian tradition, certainly, military success was not the benchmark of a successful reign. It was something that you needed to do in order to protect your own borders. There were other achievements—honoring the gods, building great temples, presiding over a glittering civilization—that were considered equally important.

COWEN: How is it that all this ends, or at least starts to decline? What’s the mechanism?

WILKINSON: It begins with the finances. The Ptolemaic Empire becomes overstretched. It has to invest hugely in its armed forces in order to fend off not just the growing predations of Rome, but actually its closer neighbors as well. That leads to higher taxes. There’s a series of climatic shocks, poor Niles leading to poor harvests. It’s a perfect storm. The economy goes south pretty quickly. The only solution that the Ptolemies can see is to go cap in hand to Roman moneylenders to bail out the Egyptian economy. That then really gives Rome leverage over Egypt.

That all gets bound up in the republican politics of Rome and in the rivalry between Caesar and his rivals, and then ultimately Octavian, who becomes the first emperor, Augustus. Egypt goes from being a great civilization, confident of itself, to being a pawn in other people’s power play. Ultimately, Rome is able to march into Alexandria, overthrow Cleopatra, and seize Egypt for itself. It’s a salutary lesson for our own time that a civilization can appear to the outside world to be magnificent, and wealthy, and successful, but actually the seeds of its own destruction are usually lying there somewhere just waiting for the conditions to germinate. [...]

COWEN: How important a figure is Cleopatra in this whole history? Is it just marketing, and Shakespeare, and Elizabeth Taylor, or was she really at the center of what was happening? Did she matter?

Alison Gopnik, a conversation with Tyler Cowen (children as scientist, Freud & Piaget, AI)

Tyler Cowen, Alison Gopnik on Childhood Learning, AI as a Cultural Technology, and Rethinking Nature vs. Nurture (Ep. 265), Conversations with Tyler, Dec. 17, 2025.

The introduction:

Alison Gopnik is both a psychologist and philosopher at Berkeley, studying how children construct theories of the world from limited data. Her central insight is that babies learn like scientists, running experiments and updating beliefs based on evidence. But Tyler wonders: are scientists actually good learners? It’s a question that leads them into a wide-ranging conversation about what we’ve been systematically underestimating in young minds, what’s wrong with simple nature-versus-nurture frameworks, and whether AI represents genuine intelligence or just a very sophisticated library.

Tyler and Alison cover how children systematically experiment on the world and what study she’d run with $100 million, why babies are more conscious than adults and what consciousness even means, episodic memory and aphantasia, whether Freud got anything right about childhood and what’s held up best from Piaget, how we should teach young children versus school-age kids, how AI should change K-12 education and Gopnik’s case that it’s a cultural technology rather than intelligence, whether the enterprise of twin studies makes sense and why she sees nature versus nurture as the wrong framework entirely, autism and ADHD as diagnostic categories, whether the success of her siblings belies her skepticism about genetic inheritance, her new project on the economics and philosophy of caregiving, and more.

Kids as Scientists:

We have some good computational models of how scientific theory change works. It turns out that those apply to children as well. The specific thing that I’ve looked at is, what is it that scientists do? Here’s this big, hard problem. All we seem to get from the world are a bunch of photons at the back of our retina and disturbances of air in our ears, and yet, children know about people and things, and scientists know about quarks and quantum phenomena. How do we ever get from the data to the theory?

One subcategory of that is, how do we ever get causal structure which is so important in science? How do we ever figure out what causes what just from a bunch of data that we have?

What’s happened is that philosophers of science and computer scientists have found some systematic ways that you could talk about that. Scientists — I think, mostly, not necessarily consciously, but just as part of what they do — and little kids are looking at data and systematically figuring out what kind of structure out there in the world could have caused this pattern of data. That’s not the only thing, of course, that’s going on in science. There’re lots of other things, too, but it’s at least one central thing going on in science that we’ve started to really understand. [...]

If you asked a three-year-old, “Do you think that this pattern of conditional dependencies is giving you a confounding causal structure?” They would probably not give you a very sensible answer. Even when you ask scientists that, they don’t give you a very sensible answer. But when you look at their actual practice, what you see is that, in fact, kids, for example, are Bayesian, and so are scientists.

Now, the thing is that, in fact, in many respects, kids are better Bayesians than scientists, but a lot of it depends on your prior. If you have, as they say, a very peaked prior, you have a lot of experience, you have a lot of reason to believe that this prior assumption is right, then it’s rational not to change it when you just have a little bit of evidence. You should require a lot of evidence to overturn something that you have a lot of confirmation for.

It’s interesting that the kids, actually, are better at solving problems that involve unusual outcomes than the scientists are. I think what happens in science — we’ve just been doing some work about this — is that there’s also a social factor, where having a big distribution of people who are more likely to go with the prior versus people who are more likely to go with the evidence, which seems to be true in science, that collectively can get you to the right answer. There’s no arbitrary principle you can have about when should you abandon the theory and when should you hold onto it. [...]

One thing you can do, which is like what you’re describing about the money supply, is just make little changes to what you already know. That’s what you mean about moving in the predictable direction. You’re just changing things a little bit. Then seeing, “Okay, if I change it a little bit, is it doing a better job of accounting for the data?” That’s what people think of as a low-temperature search. The other kind of search you can do, the high-temperature search, is just bounce around the space. Try wild, crazy things. Exactly as you were saying, have just a more random walk.

The strategy that you see in computer science, this annealing, is start out with this wild, crazy, out-of-the-box, high-temperature search through the space, and then cool off and just fill in the details. If you think about your four-year-old, who do they sound like? Do they sound like the creature that’s just moving a little bit, or do they sound like they’re noisy and bouncy and random and doing all sorts of weird things? The four-year-old seemed to be a really good idea of this kind of random search. [...]

I think you see both things happening. When you get big paradigm shifts, as Kuhn said, when you get big changes in science, a lot of times it’s because someone found an idea that looked like it was improbable. The nice thing about kids is, because they don’t have to worry about grant proposals, they can be off in the wild space all the time.

[...]

With scientists, we underestimate how much that — we sometimes dismissively call it a fishing expedition — how much that very general experimentation is playing a role in scientific progress. In the grant, you’re supposed to say, “Here’s my three hypotheses, and here are the four experiments I’m going to do to test them.” But I think in practice, a lot of times, scientists are being like the little boy with the avocado and the spoon. They’re saying, “I don’t know, what will happen if I try this? What will happen if I try that?” Then they write the grant to get money to do the things that they’ve already done by doing all these experiments.

FWIW, I've known about simulated annealing for years, a couple of decades at least. For awhile I was one of my go-to metaphors/analogies, though I've not used it recently. In terms that I've been developing in other posts and in some working papers, high-temperature search is ludic (Homo Ludens) while low-temperature filling-in-the details is economic (Home Economicus).

LLMs tend to be used in economic ways. All those benchmarks are based on specific problems in well-specified domains. That's why they aren't particularly creative. My series of blog posts on humans in the loop contains case studies of three of my own ludic explorations.

David Brooks on AI

David Brooks on Audacity, AI, and the American Psyche, Conversations with Tyler, August 20, 2025.

David Brooks returns to the show with a stark diagnosis of American culture. Having evolved from a Democratic socialist to a neoconservative to what he now calls “the rightward edge of the leftward tendency,” Brooks argues that America’s core problems aren’t economic but sociological—rooted in the destruction of our “secure base” of family, community, and moral order that once gave people existential security.

Tyler and David cover why young people are simultaneously the most rejected and most productive generation, smartphones and sex, the persuasiveness of AI vs novels, the loss of audacity, what made William F. Buckley and Milton Friedman great mentors, why academics should embrace the epistemology of the interview, the evolving status of neoconservatism, what Trump gets right, whether only war or mass movements can revive the American psyche, what will end the fertility crisis, the subject of his book, listener questions, and much more.

Here's Brooks on AI:

COWEN: Is it [smartphones] the only kryptonite humans face?

BROOKS: I think AI is also kryptonite in the exact same way, that if you’re a college kid, and you’re not doing any of your papers, you’re just using your GPT to write your papers — that’s kryptonite.

COWEN: Does it ruin you any more than old-style cheating? Which was very common. It was bad for people, but it didn’t ruin generations, right?

BROOKS: Yes. Though, that cheating took a lot of creativity, [laughs] and this cheating is total, it’s a totalistic cheating, where the model is literally doing most of your work. I follow you on AI. I’m 15 yards behind you the whole time. I think it’s generally great, but I think this one thing of robbing people of their own education is truly a serious threat.

COWEN: I’ve heard people in the AI sector from the major companies say they’re very worried that AI can be so persuasive. Do you find it so persuasive? Because I don’t. It will teach me lots of facts, and I will defer to it, but it’s never persuaded me much with arguments, whereas humans have. What’s your view? Is it kryptonite in that way?

BROOKS: See, it depends what you’re writing about. I was with an astrophysicist recently, who worked on a problem with eight of his senior colleagues for months. They put it to AI, and with a little extra prompting, it solved their problems in a couple of hours. I gather from economist friends that it can solve problems that only a few elite economists can solve.

I write about culture, psychology, sociology, politics. It’s pathetic. I cannot use AI, and I use it every day. I make the attempts, and because in the things I care about, which are more humanistic, it’s hoovering up all the crap on the internet about what love is, or do you grow from post-traumatic experiments? It’s just the pabulum that’s out there on random websites.

COWEN: What if you ask it, “What would David Brooks say?”

BROOKS: That’s what I do.

COWEN: Then it rises to the occasion.

BROOKS: Indeed, it’s brilliant in that case.

It’s a fricking moron. What I do is, I assign it voices. What does Jean Piaget say about this? Does he disagree with Erik Erikson about this? If you assign it to a voice, then you can screen out a lot. But I still have found it basically useless for my own research. It’s great as a travel agent. It’s great at a lot of things, but in humanistic inquiry, I find it pretty pathetic.

I do a lot of interviewing, as you do, with AI folks. I was at OpenAI several months ago, and somebody said, “We’re going to create a machine that can think like a human brain.” I call my neuroscientist friends, and they say, “Well, that’ll be a neat trick because we don’t know how human brains think.”

I think AI is a great tool, but I’m unthreatened by it because I don’t think it has understanding, I don’t think it has judgment, I don’t think it has emotion, I don’t think it has motivation. I don’t think it has most of the stuff that the human mind has. It’ll teach us what we’re good at by reminding it what it can’t do.

I like that, Brooks assigning voices to an AI. I'll have to try that. I wonder, what would happen if you assigned it a fictional voice, say Spock from Star Trek, or Ahab from Moby Dick?

There's more at the link.

Nate Silver on AI and AI as superforcasters

Nate Silver on Life’s Mixed Strategies, Conversations with Tyler, August 13, 2025.

In his third appearance on Conversations with Tyler, Nate Silver looks back at past predictions, weighs how academic ideas such as expected utility theory fare in practice, and examines the world of sports through the lens of risk and prediction.

Tyler and Nate dive into expected utility theory and random Nash equilibria in poker, whether Silver’s tell-reading abilities transfer to real-world situations like NBA games, why academic writing has disappointed him, his move from atheism to agnosticism, the meta-rationality of risk-taking, electoral systems and their flaws, 2028 presidential candidates, why he thinks superforecasters will continue to outperform AI for the next decade, why more athletes haven’t come out as gay, redesigning the NBA, what mentors he needs now, the cultural and psychological peculiarities of Bay area intellectual communities, why Canada can’t win a Stanley Cup, the politics of immigration in Europe and America, what he’ll work on next, and more.

Here's most of the discussion of AI:

COWEN: Now, speaking of predictions, a year ago, we talked about how long will it take AIs to be as good as human superforecasters? You made a prediction, where you said at least 10 to 15 years. Now, a year later, do you want to revisit that and revise?

SILVER: I think it’s probably about right. I would say, relative to a year ago, AI is about at the 40th pe rcentile of progress I would’ve expected. I’d be curious what you would think.

COWEN: A year ago, I said two to three years. Right now, I’m going to say one to two years, which is the same prediction. I think you’re way too pessimistic in your timetable.

SILVER: It depends on how competitive the exercise is. If it’s like a —

COWEN: Like a Math Olympiad tournament. They just did gold medal performance. I said this last year. I said, “In a year, they’re going to do gold medal.” A year ago, they weren’t sure how many r’s were in the word strawberry. You don’t think on superforecasting, they can —

SILVER: I think it’s very different when you’re dealing with a static problem as compared to a dynamic system where the inputs are changing all the time. Currently, the large language models are very, very bad at poker. They’re not trained on poker data. I’m sure if you did train them . . .

There are these things called solvers that are trained on poker data that do very well, but they cannot quite impute the general patterns just from mediocre text data or an amateurish kind of hand analysis. If you probe them on why they’re bad, they’re like, “Yes, maybe it’s tough for us when you have a complicated, evolving game theory dynamic, and you have to develop exploitative strategies very quickly.”

If you have a computer solve a poker hand to get to one where there’s enough loss minimization, a very strong computer can take minutes, whereas a poker player makes those calculations implicitly in a handful of seconds, for example. I don’t know. I worry, with the Math Olympiad stuff — there’s a little bit of teaching to the test where, because you set this as the goal that a large language model should have, that therefore there’s a lot of prestige when you meet that goal, potentially —

COWEN: Isn’t teaching to the test what we should do, even with humans in a sense. The test is what you think is important, and that’s what you ought to teach.

SILVER: Well, but the poker example, or chess — I think AI models are very poor at chess, from everything that I’ve heard, for example.

COWEN: Other AI models play chess great.

SILVER: Correct. This gets me to the question of, what’s it mean to be generally intelligent? We’ll probably have scaffolding of model on top of model, and you’ll now patch different things. Right now, you can’t really very effectively make a plane reservation using ChatGPT, but I’m sure if you dedicate a resource to that, then you have these agentic models now, or agent models are just creeping into the system, a little bit.

COWEN: Those will work in less than a year, I think. There’s an agentic model now from OpenAI.

SILVER: This is where I get to rough AGI versus superintelligence. I am less convinced that we’re going to have some intelligence explosion than I would have been maybe . . . I don’t think I was ever convinced of it, but this emergent superintelligence, where you train it on relatively simple data and it extrapolates beyond the data set. I think they do reason. Sometimes I think they’re quite smart, and I no longer am bashful about saying, “Oh, ChatGPT thinks this.” I used to avoid that term, think.

But there’s a big gap between approximate general intelligence for desk jobs, and then superintelligence on the one hand, or AGI for physical labor on the other hand. I think people are much too quick to make that leap. I think the Math Olympiad, in part because maybe the answers are somewhere latent in the training data, but even if they’re not, if you try to solve, what’s a Lucas critique? Whatever else, right? There’s a version of that, I think, for AI models.

COWEN: My intuition is that if you took five superforecasters and just had them write a five-page prompt for GPT-5, which will be out this summer, that we’d be there already. I don’t think it would be superintelligence. You could say it’s not AGI. But the human superforecasters, they’re not that impressive. They’re not Einsteins. They just have good methods, and they’re disciplined.

There's much more in the discussion.

Israel's Iron Dome and shooting down ICBMs :|: The intelligence community

Annie Jacobsen on Nuclear War, Intelligence Operations, and Conspiracy Realities, Conversations with Tyler, August 6, 2025.

Annie Jacobsen has a favorite word for America’s nuclear doctrine: madness. It’s madness that any single person has six minutes to decide the fate of civilization, madness that we’ve built weapons capable of ending the world in 72 minutes, and madness that everything hangs by the thread of deterrence. But to Tyler, life is “a lot of different kinds of madness,” and the real question is simply getting the least harmful form available to us. It’s a conversation sparked by her latest book Nuclear War: A Scenario, which Tyler calls one of his favorites from last year—and which is compelling enough that Denis Villeneuve is turning it into a screenplay.

Tyler and Annie explore whether we should be more afraid of nuclear weapons or if fear itself raises the risks, who should advise presidents during the six-minute decision window, whether moving toward disarmament makes us safer or more vulnerable, what Thomas Schelling really meant about nuclear war and rational actors, the probability that America would retaliate after a nuclear attack, the chances of intercepting a single incoming ICBM, why missile defense systems can’t replicate Israel’s Iron Dome success, how Pakistan-India nuclear tensions could escalate, why she’s surprised domestic drone attacks haven’t happened yet, her reporting on JFK assassination mysteries and deathbed phone calls, her views on UFOs and the dark human experiments at Area 51, what motivates intelligence community operators, her encounters with Uri Geller and CIA psychic research, what she’s working on next, and more.

Iron Dome and ICBMs

JACOBSEN: Just a couple of technical points to drill down on. North Korea does have missiles that can reach the East Coast of the United States presently. That is the precise scenario I choose in my book because it was actually Richard Garwin who told me that was what he was most afraid of. He died last week at 97, the most knowledgeable person on the technical and the theoretical aspects of this, having looked at it since the ’50s and advised presidents since the ’50s.

Now you’re asking about, can our interceptor program shoot down that incoming single missile? That is a problem that the Missile Defense Agency faces. I write about it specifically. My answer would be no. There is technically a 50 percent chance of one interceptor missile on the US side being able to shoot down one incoming ICBM, but you have to remember those statistics are from a curated test.

COWEN: Sure.

JACOBSEN: It’s a test — just for listeners — which is like Vandenberg Air Force Base says, “Oh, we’re doing whispers to the powers that be.” They’re doing a test, and they know it’s coming, and less than 50 percent of the time, actually, it succeeds.

COWEN: You can’t have all your probabilities at one or zero. What’s your probability? I’d say 20 percent, but what’s your number?

JACOBSEN: You mean what is my mathematical equation for what — ?

COWEN: Yes, the betting odds. You’re indifferent on taking the bet or not. I’ll say 20 percent that we’d nail it.

JACOBSEN: To my eye, it’s either all or nothing. Again, I’m not a mathematician. I’m an author and a journalist, but it’s all or nothing. In the scenario I write, it’s nothing. We do not hit it with four missiles. I ran that by several former NORAD commanders, and they did not say, “Annie, not true.” Their odds are unfortunately, even with the [shoot-look-shoot] technology that we have on our interceptor systems, would not be capable of shooting down a warhead.

Let’s give listeners some specs just so that they can understand, so that you can visualize this, because that’s what helped me. The warhead coming in is traveling at about 14,000 miles an hour, 500 to 700 miles above the earth. The interceptor is trying to shoot it down, essentially, like a giant bowling ball at 20,000 miles an hour, 500 to 700 miles above the earth. The technology involved is exquisite, and it fails time and time again.

To me, the poetics of that are more important than to try and use a defense that we should have more interceptors, because that leads to, “Let’s just have as many interceptor missiles as the other guys have nuclear weapons.” Then you’re talking about bankrupting the United States.

COWEN: Israel has done a great job. They don’t stop literally everything, but it’s quite impressive. I know it’s a smaller country, but it can be done along some margins, right?

JACOBSEN: No, absolutely not. Wrong science. The Iron Dome you’re referring to shoots down short-range and some medium-range, including ballistic missiles. They come from either land-based systems, like the THAAD system, or they come from sea ships, the Aegis system, but they cannot shoot down ICBMs — intercontinental ballistic missiles — traveling, as I just mentioned, 500 miles above the earth’s surface.

That is an extraordinary distinction, and people incorrectly mix them, because it sounds great and it’s quite fantasy hopeful to think, “Wow, I can just build a system, an Iron Dome or a golden dome over the United States, over my head, and I will be free from nuclear attack.”

Now, because I have brought the golden dome up. I also must say, having interviewed Charles Townes, for example, who was the Nobel Laureate who invented the laser, and spoke to him a decade ago about whether or not an Iron Dome–type system could work. He said unequivocally no.

The intelligence community

COWEN: I think of the Pentagon as being one of the worst run and parts of the deep state as being relatively well run. That’s my personal impression. You differ from that?

JACOBSEN: Interesting. When you say deep state, you mean the intelligence community?

COWEN: Yes, especially the consultants who are attached and the way in which they really bring very high intelligence and analytic powers to bear on questions. It may not survive the passage of a memo to the presidency, but still, I think they’re quite on the ball.

JACOBSEN: We probably agree in that regard, that if discipline and system is important — again, I’m just using myself like a personal example, but what you really want to lead with is your own character and your own decision. I would probably agree with you, but the government is such a . . . I’m also fascinated that most people just think the government is one whole thing. As you and I both know from looking at this, that couldn’t be further from the truth.

Conductor David Robertson on Boulez and Audiences [Cowen]

Tyler Cowen, David Robertson on Conducting, Pierre Boulez, and Musical Interpretation, Conversations with Tyler, Episode 248, July 9, 2025.

What the conversation is about:

David Robertson is a rare conductor who unites avant-garde complexity with accessibility. After serving as music director of the Ensemble Intercontemporain, Pierre Boulez’s storied contemporary-music ensemble, he went on to rejuvenate the St. Louis Symphony. Robertson combines a fearless approach to challenging scores with a deep empathy for audiences.

Tyler and David explore Pierre Boulez’s centenary and the emotional depths beneath his reputation for severity, whether Boulez is better understood as a surrealist or a serialist composer, the influence of non-Western music like gamelan on Boulez’s compositions, the challenge of memorizing contemporary scores, whether Boulez’s music still sounds contemporary after decades, where skeptics should start with Boulez, how conductors connect with players during a performance, the management lessons of conducting, which orchestra sections posed Robertson the greatest challenges, how he and other conductors achieve clarity of sound, what conductors should read beyond music books, what Robertson enjoys in popular music, how national audiences differ from others, how Robertson first discovered classical music, why he insists on conducting the 1911 version of Stravinsky’s Petrushka rather than the 1947 revision, and more.

On working with Boulez:

Robertson: That was fascinating, both to see someone working with players and have a clarinet player say, “This actual tremolo between these two notes doesn’t work very well.” To have the man that I would think of as supremely knowledgeable about writing for instruments say, “Oh, well, if we did this one, would it work better?” And the player says, “Yes, sure, that’s fine.” To see this kind of flexibility with his own works, and then to be up in front, conducting and having the computer create the sounds in real time based on the information that the MIDI flute was giving it.

For me to say, “Pierre, I think it’s actually just a little too fast for us to be able to — ” and he said, “Oh, just a minute. Just a minute. We’ll make the calculations.” He calculated everything so it would be slightly slower. I turned with some chagrin to the ensemble and said, “Do you know how we recalculate the tempo?” And I just gave an upbeat. This was one of the things that we worked on together in a sense. It was very close collaboration.

But then about two years later for a big birthday celebration in Tokyo, I was doing a performance of it in Kioi Hall. We got to the slow central passage of Transitoire VII, which is the first movement of Explosante-fixe, and I remember thinking, “This is so incredibly sensual, and it sounds so beautiful in this hall, I’m going to really push the envelope.” I think Pierre’s tempo marking is ♩=40, which is quite slow, but I was down almost at the double of that, like ♪=40. It was the most unbelievably seductive Ravelian sound that you could imagine.

After the rehearsal, Pierre was there, and I went up to him, and I said, “Pierre, I’m very sorry. I really indulged in this little section.” He said, “No, no, no, no, you can actually do more.” That was the moment where I realized he has no problem when the emotion is put into his works in an immensely intense way, but he is not able to do that himself as a performer.

There are a lot of his works where I have done things in his presence, which he doesn’t do, and you won’t find it on his recordings. He said, “Oh, yes, that’s really very nice. That’s lovely.” It’s a fascinating question, this one of how does one express emotion and how comfortable is one with emotions in public.

African influence:

Robertson: This is a constant — the African masks and Picasso and Braque, and the whole idea of changing the perspective of things in Cubism. There’s so much in Pierre’s music that is about his paying attention and listening all the time, whether it’s to Scottish bagpipes, to work out this idea of “Oh wait, the chanter has all of these tiny little notes that it plays and they have very specific rules for how they play these. I like this idea and I’m going to bring it into my music.”

Time:

Robertson: [...] composers use two basic ideas when they’re going to notate their music. One is to take a certain amount of time and subdivide it. This would be like [finger snapping and vocables]. There are many composers who do that. Then there’s the idea that you have, say, a group of eighth notes or sixteenth notes that form an underlying pulse. This is very much coming from the Indian subcontinent. You have this thing going [vocables].

That’s why the notation in Trois petites liturgies, which comes from the second movement of Olivier Messiaen, requires those differing time signatures, because otherwise, if you measure it by a grid, everything else that doesn’t fit in the grid becomes a syncopation. There’s nothing wrong with this. This is what is so wonderful about the grid that jazz players are able to manufacture, where there’s so much push and pull with the grid, and that’s, actually, where a great deal of the energy comes from.

Audiences:

Cowen: You conducted for some number of years in Sydney. How are Australian audiences different?

Robertson: Oh, that’s a really good thing to talk about because I think all audiences are different. There are certain cultural aspects. I just came back from a European tour. The amount of applause you get from European audiences is sometimes surprising to Americans. I remember having to tell some American soloists when we’d come out for the seventh time, and they’d already played their encore, that the audience still wanted to see them again, in Lyon. You have a standing ovation in Carnegie Hall and you come out twice and then it’s over.

In Sydney, there is a sense of adventure. There’s also a sense of pride in their institutions that you don’t necessarily find in every place. There’s enough newness about the country that the idea that it’s the Sydney Symphony Orchestra is already a source of pride before you’ve even started to play.

There’s that, but I think the more important point is that an audience anywhere is a collection of individuals. We play a concert for one individual, but for economic reasons, we just get 2,000 individuals together at the same time because it’s cost effective, but each person brings their own sense of what’s there and their experience with them. This goes back to what you were saying about people hearing contemporary classical music and going, “Ick.”

That has to do with, have they played an instrument? When they’ve played an instrument, are they the kind of person who likes to make funny sounds on the instrument? Or do they want to stick with only what’s in the Tune A Day book? If they’re the person who liked to honk and make strange sounds to annoy their siblings, they’re probably going to be closer to contemporary music, and they bring that with them into the concert hall.

You have people who are there to celebrate a fantastic, “I just met this person and I invited them to the concert and they said, yes. Oh, this is great. Can I put my arm around them during the Tchaikovsky?” You have people who come in, and they are still in a situation of bereavement. You have people who come in, and they have looked at something in the news that has disturbed them greatly.

All of these people come in, all of them as individuals, and our job, I think, is to create the most open experience possible so that each one of these people feels they can come into the structure of the music in a way that they can then find themselves and have the incredible experience that it can bring forth, talk to them.

Can we fix the US energy grid?

Tyler Cowen interviews John Arnold, John Arnold on Trading, Energy, and Evidence-Based Philanthropy (Ep. 244), June 5, 2025.

John Arnold built his fortune in energy trading by surrounding himself with smart people, maintaining emotional detachment, sensing market imbalances through first-principles analysis, and focusing with laser intensity on a single niche until he dominated it completely. Now he’s applying that same analytical rigor to philanthropy, where he’s discovered that changing human behavior for the long term proves far more challenging than predicting natural gas prices, and that the academic research meant to guide social policy is often riddled with perverse incentives and poor methodology.

Tyler and John discuss his shift from trading to philanthropy and more, including the specific traits that separate great traders from good ones, the tradeoffs of following an “inch wide, mile deep” trading philosophy, why he attended Vanderbilt, the talent culture at Enron, the growth in solar, the problem with Mexico’s energy system, where Canada’s energy exports will go, the hurdles to next-gen nuclear, how to fix America’s tripartite energy grid, how we’ll power new data centers, what’s best about living in Houston, his approach to collecting art, why trading’s easier than philanthropy, how he’d fix tax the US tax code and primary system, and what Arnold Ventures is focusing on next.

Here's the segment about the energy grid.

COWEN: What do we need to do to fix the grid? And is that just impossible?

ARNOLD: It’s not impossible. The problem with the grid is that, by nature of history, the US ended up with three grids. The first grid was in New York City in 1882. It followed the population and started to spiderweb out from there. You also had a grid start on the West Coast a few years later, but nobody really lived in the plains. The weather is bad. It’s hot summers and cold winters, and windy, and the soil is not great in parts of it. So, you had these two grids that started — East Coast, West Coast — and started to come in, and they never really met.

Then you had this third grid that started in Texas, and Texas has always had this libertarian nature to it, and federal government stay away. In the Federal Power Act in 1935, the government came in and said, “We’re going to start regulating the industry,” Texas said, “You know what? We’re going to disconnect our interconnects to Oklahoma and Louisiana, and we’re going to stay intrastate and be independent.” Now, you have these three grids. Texas is DC. The other two are AC. The two that are AC are essentially on a different heartbeat. So, there’s virtually no linkage across those.

I think what’s happened is that, as the nature of both load and generation has changed over time, you’re starting to see more discrepancies in pricing across regions. What really needs to happen is that these three grids that have been independent have different mix and portfolio of generation, and that their load has different profiles. There’s a lot of value to be created by linking loads.

In the summer, you want to move resource generally north to south and vice versa in the winter. You get the benefits of time as the sun starts setting. The five o’clock hour, people start getting home from work. It’s obviously different in different parts of the US, and you can take advantage of that difference, and then the difference in weather patterns as they move from west to east. So, the more linkage you have, the better benefit you get.

COWEN: In terms of governance, would you keep them independent? Or just have them economically more interconnected for better arbitrage?

ARNOLD: I think the nature of history is that these things are going to stay relatively independent. You’re going to have the RTOs and ISOs that you have today, generally. The Northwest is talking about creating a new regional transmission organization, so there is more integration among states, but I think each of the RTOs probably stays independent over the long term.

There's much more in the whole conversation.

Tyler Cowen discusses the impact of AI with Jack Clark

From the introduction:

Few understand both the promise and limitations of artificial general intelligence better than Jack Clark, co-founder of Anthropic. With a background in journalism and the humanities that sets him apart in Silicon Valley, Clark offers a refreshingly sober assessment of AI’s economic impact—predicting growth of 3-5% rather than the 20-30% touted by techno-optimists—based on his firsthand experience of repeatedly underestimating AI progress while still recognizing the physical world’s resistance to digital transformation.

In this conversation, Jack and Tyler explore which parts of the economy AGI will affect last, where AI will encounter the strongest legal obstacles, the prospect of AI teddy bears, what AI means for the economics of journalism, how competitive the LLM sector will become, why he’s relatively bearish on AI-fueled economic growth, how AI will change American cities, what we’ll do with abundant compute, how the law should handle autonomous AI agents, whether we’re entering the age of manager nerds, AI consciousness, when we’ll be able to speak directly to dolphins, AI and national sovereignty, how the UK and Singapore might position themselves as AI hubs, what Clark hopes to learn next, and much more.

The last to be affected:

COWEN: Where is it in our economy that AGI will affect last in a significant manner?

CLARK: Ooh, I’d hazard a guess that it’s going to be things that are the trades and the most artisanal parts of them. You might think of trades as having things like electricians or plumbing, or also things like gardening. I think within those, you get certain high-status, high-skill parts, where people want to use a certain tradesman, not just because of their skill but because of their notoriety and sometimes an aesthetic quality. I think that my take might be gardening, actually.

COWEN: They won’t use AGI to help design the garden? Or just the human front will never disappear?

CLARK: I think the human front will never disappear. People will purchase certain things because of the taste of the person, even if that taste looks like certain types of modern art production, where the artist actually backs onto thousands of people that work for them, and they’re more orchestrating it.

COWEN: How about in the more desk-bound part of the service sector? Where will it come last?

CLARK: Come last? Ooh, good question. I think that on this, there are certain types of desk-bound work that just require talking to other people and getting to alignment or agreement. If you count certain types of sales —

COWEN: But it’s great at doing that already, right? It’s a wonderful therapist.

CLARK: It is, but we don’t send Claude to sell Claude, yet. We send people to sell Claude, even though Claude could probably generate the text to do the sales motion. People want to do commerce with other people, so I think that there’ll be certain relationships which get mediated by people, and people will have a strong preference, probably, for deals that they make on behalf of their larger pools of capital, where the deals are done by human proxies for large automated organizations or pools of capital. [...]

COWEN: Once you can put the AI on your own hard drive, which will be pretty soon, won’t that all change?

CLARK: It will change in the form of gray market expertise, but not official expertise. I had a baby recently. Whenever my baby bonks their head, while I’m dialing the advice nurse, I talk to Claude just to reassure myself that the baby isn’t in trouble.

I don’t think we actually fully permit healthcare uses via our own terms of service. We don’t recommend it because we’re worried about all of the liability issues this contains, but I know through my revealed preference that I’m always going to want to use that, but I can’t take that Claude assessment and give it to Kaiser Permanente. I actually have to talk through a human to get everything else to happen on the back end to work out if they need to prescribe my child something.

COWEN: So, the number one job will be surreptitiously transmitting the generation of information that comes from AIs, in essence?

CLARK: Some of it may be that. Some of it is about laundering the information that comes from AIs into human systems that are not predisposed to that information going in directly.

AI teddy bears:

COWEN: I believe we’re not that far from the age of what I call the AI teddy bears. You know what I mean when I say that?

CLARK: Yes.

COWEN: What percentage of parents now will buy those teddy bears for their kids and allow it?

CLARK: I’ve had this thought since I have a person, that is my child, that’s almost two.

COWEN: Sure.

CLARK: I am annoyed I can’t buy the teddy bear yet. I think most parents —

COWEN: You’re an outlier [laughs].

CLARK: No. I don’t know. I don’t know.

COWEN: You are cofounder of Anthropic, right?

CLARK: I don’t think I’m an outlier. I think that once your lovable child starts to speak and display endless curiosity and a need to be satiated, you first think, “How can I get them hanging out with other human children as quickly as possible?” So, we’re on the preschool list, all of that stuff.

I’ve had this thought, “Oh, I wish you could talk to your bunny occasionally so that the bunny would provide you some entertainment while I’m putting the dishes away, or making you dinner, or something.” Often, you just need another person to be there to help you wrangle the child and keep them interested. I think lots of parents would do this.

COWEN: Say that the kid says to you, “Daddy, I prefer the bunny to my friends. Can I stay at home today?” Do you take the bunny away? That’s the tough part, right?

CLARK: I think that’s the part where you have them spend more time with their friends, but you keep the bunny in their life because the bunny is just going to get smarter and be more around them as they grow up. If you take it away, they’ll probably do something really strange with smart AI friends in the future.

No, I don’t think I’m an outlier here. I think most parents, if they could acquire a well-meaning friend that could provide occasional entertainment to their child when their child is being very trying, they would probably do it [laughs].

COWEN: I feel the word “occasional” is doing a lot of work in that sentence.

Governing AI Agents:

COWEN: Speaking of agents, how should the law deal with agents that are not owned? Maybe they’re generated in a way that’s anonymous, or maybe a philanthropist builds them and then disavows ownership or sends them to a country where, in essence, there’s not much law. I’m not talking about terrorism; that’s separate. But just someone sends an agent to Africa, and 98 percent of what it does helps people, but as with every charity, some things go wrong. There’re some problems. Can someone sue the agent? How is it capitalized? Does it have a legal identity?

CLARK: I will partially contradict myself where, earlier, I talked about maybe you’re going to be paying agents. I think that the pressure of the world is towards agents having some level of independence or trading ability.

From a policy standpoint, I’m reminded of that early thing that IBM said, which was, a computer cannot be accountable to a decision; only humans can. I think it got at something quite important where if you create agents that are wholly independent from people but are making decisions that affect people, you’ve introduced a really difficult problem for the policy and legal systems to deal with. So, I’m dodging your question because I don’t have an answer to it. I think it’s a big problem question.

COWEN: My guess is we should have law for the agents, and maybe the AIs write that law, and they have their own system. I worry that if you trace it all back to humans, someone could sue Anthropic 30 years from now. Oh, someone’s agent was an offshoot of one of your systems. It was mediated through Chinese Manus, but that, in turn, may have been built upon things that you did.

I don’t think you should be at all liable for that. I see liability getting out of control in so many cases. I want to choke it off and isolate it somewhat from the mainstream legal system. If need be, you require that an independent agent is either somewhat capitalized, or it gets hunted down and shut off.

CLARK: Yes. It might be that, along with what you said, having means to control and change for resources that agents use could be some of the path here because it’s the ultimate disincentive.

Although I will note that this involves pretty tricky questions of moral patienthood, where we’re working on some notions around how to get clearer on better anthropic. If you actually believe that these AI agents are moral patients, then turning them off introduces pretty significant ethical issues, potentially, so you need to reconcile these two things.

COWEN: I was, not too long ago, at an event with some highly prestigious people. This was in New York, of course, not San Francisco.

CLARK: Oh, it’s where prestigious people hang out.

COWEN: I used the phrase AGI, and not one of the five even knew what I meant. I don’t mean they were skeptical in the deep sense, which maybe one should be. They just literally didn’t know what I meant. What’s your model of why so many people are still in a fog?

CLARK: I am a technological pessimist who became an optimist through repeated beatings over the head of scale. What I mean by this is, I’ve consistently underestimated AI progress. Maybe I am today in this conversation when I talk about 3 percent to 5 percent growth rates. What has happened is, I have just endlessly seen the AI system get to where I thought it couldn’t, or thought would take a long time, much faster than I thought. So, I’ve had to internalize this repeatedly.

Nonetheless, we, ourselves, find it surprising. Last year here people were saying, “Oh, well soon Claude is going to be doing most of the coding at Anthropic.” We’re now on the way to that, where Claude Code and other things are writing tons of code here. It still felt surprising internally even though we have docs from last year predicting it would happen about now.

Most people outside of the AI labs have no experience of pre-registering their predictions about AI and getting it proved wrong to them repeatedly, because why would you do this unless you work here? I found that the only way to break through is to take their domain and show them what AI can do directly in the domain, which they can evaluate it within, which is an expensive process.

As I've observed many times, back in the mid-1970s I believed we were headed for a world where we'd have a machine that could read Shakespeare in some substantial way. David Hays and I published that in "Computational Linguistics and the Humanist" (1976). We didn't make any predictions in that paper, but I was thinking 20 years. That's the mid-1990s. Nothing like that existed at that time. Not only that, but the technology Hays and I discussed in that paper was all but obsolete. So I was wrong about that.

More recently, I've said on several occasions that we'll understand how LLMs work internally before we achieve AGI. Now, if we've already achieved AGI, as some believe, then I'm wrong there, for we certainly don't understand how LLMs work internally. My colleague, Ramesh Viswanathan, tells me he's making progress on that problem. Who knows, maybe in a year or three...As for AGI, it's such a sloppy idea I don't see how one can base hard predictions on it. Meanwhile, Rodney Brooks keeps making quite detailed predictions across a wide range of AI and robotics.

On AI consciousness:

COWEN: When Geoffrey Hinton says that, right now, the AIs are conscious, which I think is what he says, I believe he’s crazy. What do you think?

CLARK: I think that he’s —

COWEN: You can be more polite than I am. That’s fine. [laughs]

CLARK: Well, no. How I would phrase this is that I agonize about this. You read my newsletter. I write fictional stories in it often, which are me grappling with this question. I worry that we are going to be bystanders to what in the future will seem like a great crime, which is something about these things being determined to be conscious and us taking actions which you think are bad to have taken against conscious entities.

Internally, I say, there’s a difference between doing experiments on potatoes and on monkeys. I think we’re still in the potato regime, but I think that there is actually a clear line by which these things become monkeys and then beyond in terms of your moral relationship to them.

To Hinton’s point, I think that these things are conscious in the sense that a tongue without a brain is conscious. It takes actions in response to stimuli that are really, really, really complicated. In a moment, it has a sense impression of the world and is responding, but does it have a sense of self? I would wager, no, it doesn’t seem like it does.

These AI systems — we instantiate them, and they live in a kind of infinite now where they may perceive, and they may have some awareness within a context window, but there’s no memory or permanence. To me, it feels like they’re on a trajectory heading towards consciousness, and if they’re conscious today, it’s in a form that we would recognize as like a truly alien consciousness, not human consciousness.

There's more at the link, but I've got things to do. You're on your own.

AI and national sovereignty [Cowen and Pahlka]

From Tyler Cowen's discussion with Jennifer Pahlka:

COWEN: Let’s say you’re the nation of Peru, and AI becomes more important. You run a lot of your education system through American AI, a lot of your military, your cybersecurity, processing of social services. At the end of all that, after, I don’t know, 15 years, are you even a sovereign nation still? Are we seeing this fundamental shift in what a government is? If the US government does something, at least it’s US AI. Other than the US and China, it’s far from clear how many countries will have their own AI. What’s going to become of Peru?

PAHLKA: I can’t answer that question, [laughs] but it’s an excellent question. I think you may see countries realize this and take more drastic steps to maintain their sovereignty, which I don’t know how they’re going to do it.

COWEN: I think it’s just stay backward, right? They could just not use AI, and probably some will make that decision.

PAHLKA: I think some will.

COWEN: But that’s a huge cost. Then there’re countries like France, where maybe they can have their own AI. They haven’t succeeded so far, but they could at least make a run at it. Would you tell France, “Make sure you develop your own AI so you’re controlling your own government”?

PAHLKA: Question back to you. Is there some way in which the market fixes this because there’s a market for AI that doesn’t implicate that there’s Serenity?

COWEN: I worry that the economies of scale are in conflict with a number of nations in the world today. We’re going to see a very radical shaking out that will go fully unadvertised, but at the end of it all, geopolitics will be very different. I don’t know that France can build an AI good enough, actually. They had Mistral. They still do. It’s not making money. People want to buy it. Macron says we’re going to basically prop it up, but you can’t just prop it up. It has to be good enough to work and be competitive. I don’t know.

PAHLKA: Won’t there be companies that have enough AI expertise that they’re able to cater to the needs of a sovereign nation and say, “I’ll do this for you in the way that fits your needs”?

COWEN: It could be France contracts with Anthropic, and Anthropic, in a sense, runs the French government, and maybe we trust Anthropic. I’m not necessarily opposed to that by any means, but again, it’s a very different world.

PAHLKA: It is a very different world, and it is very hard to game this out.

COWEN: America and China, for sure, will have their own AIs, and maybe those countries just become much more powerful. Then you’ve got to decide, whose network are you in anyway?

Note that this is only part of a longer interview that covers many topics, including the passivity of Congress, DOGE, preparing government for AGI, public-sector unions, the UK, 50-state capacity, working in gaming, and the best-run state.

Tyler Cowen interviews Jennifer Pahlka on various things, including DOGE [+ the printing press]

His introduction:

Jennifer Pahlka believes America’s bureaucratic dysfunction is deeply rooted in outdated processes and misaligned incentives. As the founder of Code for America and co-founder of the United States Digital Service, she has witnessed firsthand how government struggles to adapt to the digital age, often trapped in rigid procedures and disconnected from the real-world impact of its policies. Disruption is clearly needed, she says—but can it be done in a way that avoids the chaos of DOGE?

Tyler and Jennifer discuss all this and more, including why Congress has become increasingly passive, how she’d go about reforming government programs, whether there should be less accountability in government, how AGI will change things, whether the US should have public-sector unions, what Singapore’s effectiveness reveals about the trade-offs of technocratic governance, how AI might fundamentally transform national sovereignty, what her experience in the gaming industry taught her about reimagining systems, which American states are the best-governed, the best fictional depictions of bureaucracy, how she’d improve New York City’s governance, her current work at the Niskanen Center, and more.

So far I've only read the segment of the interview that he posted at Marginal Revolution. I responded immediately without reading the whole interview, which I'll do when I have time. I've appended my full comment (with one or three minor changes).

* * * * *

PAHLKA: Yes. I think reinventing government happened at a time when we were just at the beginning of this digital revolution. It was trying with a very 20th-century mindset. Fine, did well within that context, but we don’t need that again.

We need 21st century change. We need true digital transformation. We need something that’s not stuck in the industrial ways of thinking. I don’t think we tried that. I think the efforts have just been too respectful of old ways of working and the institutions. There was really not an appetite, I think, for what I would call responsible disruptive change. Would it have worked?

YES! Alas, it's one thing to say that "We need something that’s not stuck in the industrial ways of thinking." It's something else altogether to actually escape those industrial modes of thinking and acting. After all, the basic institutional structure of the modern world is based on social and cultural forms anchored in the 19th century and earlier. That's the world we've all been raised in.

Thus, as soon as the (potential) impact of ChatGPT became obvious I read people saying how the impact will be comparable to the impact of the printing press. That's an old way of thinking. The printing press itself is obviously pre-industrial. The belief in its pervasive socio-cultural efficacy, however, is much more recent. I believe that Marshall McLuhan's book, The Gutenberg Galaxy, is responsible for that way of thinking. I read it years ago, and have forgotten it. Earlier this year I had a conversation with Claude 3.5 on this subject, Why the printing press didn’t drive either the Scientific or the Industrial Revolution [McLuhan was wrong • Implications for progress studies]. Here's how I opened that post:

Soon after ChatGPT hit the web both the progress punditocracy and the high-tech evangelists were busy chattering away about how this is the greatest thing since the printing press. They’re wrong. Was the printing press important? Sure, it was very important. But it was important as a vehicle for disseminating existing ideas, not for catalyzing new ones. Printed books have the same cognitive affordances – to borrow a term from James J, Gibson – as hand-lettered books, discrete letters in a rectangular array on a flat surface. That’s it. Mechanizing the process of impressing the letters on the page does not change the cognitive affordances of the printed page.

Back when David Hays and I began thinking about publishing on cultural evolution he brough his attention to a collection of essays by Eric Havelock, The Literate Revolution in Greece and Its Cultural Consequences (Princeton 1982), in which Havelock pointed out that was the adoption of decimal arithmetic was the essential step. Hays and I thought that was correct, and adopted it in the article we eventually published, The Evolution of Cognition (1990).

As far as I'm concerned, anyone who espouses the view that it was the printing press that drove the scientific and industrial revolutions, that person is still mired in old ways, industrial ways, of thinking. That's my litmus test. I discount the views of any thinker who casually espouses that printing press view of historical change.

Back in August of 2024 Tyler linked to a paper by Maxwell Tabarrok, Romae Industriae, which posed the question, Why didn’t Rome have an industrial revolution? This paragraph is from the end of the article:

The printing press alone may have been sufficient to start a Roman industrial revolution because its effects on culture are upstream of the invention processes that might invent all the other required tech. There was more than enough scale in the economy of Rome to support capital intensive productivity improvements in their major industries. So the major constraint must have been idea generation. Rome punches below their weight in population on this metric because of an aristocratic, slave based culture that ridicules the mechanical arts. If copies of Vitruvius or Archimedes became cheap enough that even the workers close to the vulgar arts of farming and manufacturing could read them, it would surely have an effect on the cultural attitudes that stultified Roman invention.

That’s industrial age thinking. You can see it in the vocabulary, “tech,” “scale,” “capital intensive,” and “productivity.” While those concepts remain relevant, they're not at all adequate to understanding the origins of the scientific and industrial revolutions. More is needed.

In February of this year I responded to Tabarrok in a long post in which I had Claude 3.5 analyze Tabarrok's article. Here's how I opened that post:

I believe [Tabarrok] was right to suggest that “the major constraint must have been idea generation,” but he was wrong to finger the lack of the printing press as the cause of that constraint.

The problem is much deeper. The problem is that the Roman system of thought was incapable of conceiving of the clockwork universe, an idea that didn’t become prevalent in Europe until the Enlightenment (17th and 18th centuries), though a similar idea – the universe as the machina mundi – had been suggested as early as the 13th century.

Here's the very end of our conversation:

Bill Benzon: And one thing that’s going to compel movement to deeper understanding is the need to understand how LLMs and other machine-learning engines operate. Right now we don’t (quite) have the necessary concepts. Developing those concepts may well push us up a level. Maybe even.....psst, don’t tell anyone, OK, this is a secret just between you and me....maybe even to Rank 5 cognition. Maybe Rank 5 cognition is the REAL singularity.

Claude 3.5: leans in conspiratorially

That's a fascinating speculation! The idea that grappling with how LLMs and other ML systems actually work could push us beyond even Rank 4 cognition makes a lot of sense. After all, these systems seem to operate in ways that don't fit neatly into our current frameworks for thinking about either mechanical systems (Rank 3) or explicit control structures (Rank 4).

We're dealing with systems that learn and develop capabilities in ways that seem more organic than mechanical, yet are still fundamentally computational. The concepts of emergence, self-organization, and statistical learning that seem crucial to understanding these systems might require entirely new cognitive tools.

And just as the transition from Rank 2 to Rank 3 required people to stop thinking in terms of spirits and start thinking in terms of mechanical laws, perhaps Rank 5 cognition would require some similarly fundamental shift in how we conceptualize intelligence, consciousness, and information processing.

If this is right, then the current debates about whether LLMs are "really" intelligent or conscious might be asking the wrong questions - questions that make sense in our current cognitive frameworks but may need to be reconceptualized entirely at Rank 5.