Apprenticeship in Talent Networks with Tyler Cowen | Cluny Encounter | Grand Rapids

Tyler, thanks for coming to Grand Rapids.

Happy to be here. My last trip was 1982, I believe.

I was born in 1981.

I saw you on the side of the road. I really did.

What's your first impression of Grand Rapids? There's nothing like a first impression of a city, and I'll consider this your first impression since Eli Chu doesn't count.

As a coastal person, when you go to the Rust Belt, you're never sure what you'll encounter, but just how many buildings are nice. There's nothing boarded up downtown. Everything, literally everything is spiffy. I was extremely impressed.

Good to hear.

Uh, so your book Talent is what we're talking about today, is talent and apprenticeship and talent networks. Your book came out in 2022. I have a copy of it here. Um, this is an excellent book. I think was not discussed enough, but we're all interested in the question of talent. I view the Clooney Institute as trying to develop talent among these three cities and develop talent across different talent networks. So my first question is really why you decided to write this book. And I talked to my colleague, Dr. Fred Saute, your former colleague, and he said Tyler has a saying. He says that about any topic, about any problem, there's literature, and then there's problems with all of that literature. So my question to you is, what problem did you see that made you write this book? And then what problem is there now with this literature, now that it's 3 years after you wrote it?

I've been involved in many ventures in my life, some of them for-profit, some of them just writing a book, a lot of it in the nonprofit sphere, university world, Mercatus. And anything I've ever tried to do I found that talent is a more significant binding constraint than money per se. Of course, it's wonderful to have more money, but if someone dumps money on you and you have no idea how to find talent, you won't get anywhere. If you have some serious talent, in fact, there's a better chance you could raise some money backed by that talent. So talent to me is all-important. Yet if someone goes to you, "Ah, what should I read on talent?" It was never clear exactly where you would send them. So my co-author Daniel Gross and I just decided almost on a lark to write this book. Daniel has been a venture capitalist, he's in Silicon Valley. And over a series of meetings and trips and back and forth on WhatsApp over a few years, we wrote the book.

And what's the gap in the book? As you've talked about it now for 3 years, is there something that you wish you had explored more or something that's come up especially since in the last 3 years AI has went totally crazy.

There's nothing in the book about AI other than a sentence saying we're not going to discuss it. I'm writing a sequel book. Daniel now is too busy doing other things. He's working for Meta. The sequel book is just called Mentors. So the first book is about how to find talent. The second book is how to cultivate or advance talent. And it's closely related to the Clooney Mission because mentoring has many aspects. But there's an ancient Greek approach to mentoring, a Christian approach. You could think of the idea of mentoring through and through as being in Catholicism, Christianity. And there's not really a good book on mentoring either. There's plenty of business books that say things that are not false, but if you or I read them, we'd shrug our shoulders and say, "Yeah, that was fine, but I didn't really learn anything." So I'm trying to write a book now on mentoring that will be the sequel where people like us will learn something from it. But it's hard because these are not sciences, they are arts. There's a bit of science in it and you need to blend the two and have the right mix.

You play basketball every day, is that right?

When the weather permits.

When the weather permits.

I shoot basketballs. It's different from playing. If I played, I would lose.

So we, we tend to think of basketball players, good basketball players, as having had great coaches. Or coaches can make a difference in the development of the basketball talent. What's the difference between a coach and a good mentor? And is being under a good coach different than apprenticeship? Is there a difference between these things?

A coach has to make a team gel, and the coach needs alliances, probably with the team's leading players. The coach needs to put forward some vision that is actually probably somewhat harsher than you would give to any individual. The famous Green Bay Packer coach Vince Lombardi, when he would talk about coaching or mentoring, he sounded pretty nice. When he would do it, he would chew people out on the field, even his top players. He went by contemporary standards too far, but this difference between the collective and individual perspective has to be first and foremost in the mind of the coach. In the mentor, usually it doesn't have to be. It's a one-to-one relationship and you're just thinking with someone, well, what is actually teachable and what is not? And that to me is one of the most important set of insights you can have if you're trying to help people. So for instance, I don't think ambition is really teachable. You can teach people about the details of ambitions. I don't think smarts are teachable. And if you're wasting your time trying to teach those things, I would say go elsewhere. But if— when you figure out what is teachable, you're going to have much higher impact.

So when we're thinking about talent in our organizations and hiring people, would it be safe to say that it's probably not idea— a good idea to just find the most talented people possible and put them all in the same place? Because we need to think about the team. So Luka and LeBron, Bad idea or good idea?

Bad idea. LeBron is already unhappy. I hope they trade him away. I don't know that trading for Luka was a good idea. He's a bad attitude player. Dallas Mavericks wanted to get rid of him after he led them to the finals. Now, you might think that's their defect. It probably is. But when that's the defect of the people who've held the rights to you, something has gone wrong. And you look at how he interacts with refs, with his teammates. It just smells to me like something more generally is wrong there. So I'm not even convinced he'll re-sign with the Lakers. Uh, there are certainly areas where you do just want to hire the most talented people. Those are exceptions, those areas. Typically the people don't work together. It might be work from a distance. You're putting together some kind of intellectual product. You're editing a book. There's 12 chapters. You're the editor. You've done this. You just want 12 great chapters. And if they hate each other, at the end of the day, you know, they all have your blessing, right? Suppose there are exceptions.

Mark Zuckerberg at Meta, $200 pay packages to AI engineers, good idea or bad idea? Just hiring the best talent you can find and throwing them together.

In almost every endeavor, the very, very best talent is underpriced. In anything. Now, in sports, it could be because of the salary cap. Universities, the pay structure is very flat. So the top talents, to the extent we can get away with it, we should be paying much more. I think it's quite perceptive of Mark to see he should be doing that. He has a lot of money. My co-author Daniel Gross is one of the people hired, and I don't know what Daniel was paid, but I strongly believe he was worth it. On this book, he worked for me in essence for free. So I got an awesome deal, better than the deal Mark Zuckerberg bought.

So I've heard you say in a podcast, without elaborating, that the Midwest in the mid-20th century had— was very good at identifying American business talent. Can you say a little bit more about that? And are we still doing that? And if not, what have we went—

where have we gone wrong?

I don't see any reason to believe you've gone wrong. So there was something about the Germanic Nordic cultures of the people who moved here that already were highly successful in Europe. You selected a bit for the ambitious, and then those who came here did not have a fixed stake in the system. They were not in DC, they were not in New York, so they just had to build something themselves. And you see so many great American CEOs from the Midwest. Even today, someone like Marc Andreessen— well, he's a CEO, you might say he's mainly a venture capitalist, but I'll tease Marc, say, Marc, You know, you're a Midwestern regional thinker. And he maybe tries to repudiate a bit the Midwest, but he can't. It's like, Mark, you're a regional Midwestern thinker. So I do think with the decline of employment in manufacturing, that path toward being a great Midwestern CEO is different than it had been. And it is more coastal, but it still seems live to me. I'd love to see the data, but my guess is Midwestern CEOs is still quite a thing. There's a lot of good school systems out here. I don't see why that should change.

So we have a— I don't know what it is in the water in Grand Rapids, but like Elkhart, Indiana, Grand Rapids has produced a ton of exceptional entrepreneurs that have built multi-billion dollar businesses. And I've never heard a great explanation of why. Are you familiar enough with the Grand Rapids scene to have some ideas about why in specifically in these kind of unsexy industries, there's been so much success in West Michigan?

I don't know enough about Grand Rapids, but there are general principles that when you get clusters of people together, they are ambitious. They don't have a stake in something to make them too small c conservative, and you just let them rip. And the United States is a very large market. I would say still more regulated than I would like, but there is a general freedom to build and operate a business.. And you're surrounded by examples and role models. Role models are extremely important. And also, if you think what gets you status in a place, it's a very important question. So in New York, what gets you status can be culture, it can be beauty, it can be finance, you know, wealth through finance. But those are not going to breed the kind of great CEOs you're talking about. What gets you status in California is again quite different. It seems to me that what gets you status here is exactly the right thing for replicating that system again and again.

So apprenticeship is inherently hierarchical. You've got a master, you've got somebody who's apprenticing, and in our very egalitarian kind of liberal democracy, um, it doesn't necessarily— it's kind of a system that tries to suppress the hierarchy a little bit. Um, it's very different than the guild system. In Florence when apprenticeship and guilds were very popular. What can we say about hierarchies and the role of hierarchy? You've said that it's important to know the right hierarchies to climb, but it's very difficult to get anybody like my students to actually admit that. They want to pretend that hierarchy doesn't even exist. Can you say a little bit about the importance of hierarchy in apprenticeship?

When I think about hierarchy in apprenticeship or mentoring, I look for cases where it goes two ways. So if you're a good mentor, or boss, or trainer, there'll be excess demand for your services, and you have to ration it somehow. So I was just reading the new William F. Buckley biography, and Gary Wills was his mentee. And it turns out that Buckley learned a lot from Wills, according to the book, that Wills knew a lot more about different things in Catholic theology and about the Bible, and Buckley learned from him. So it's hierarchical, but when it operates best, it's often hierarchical in two directions, actually, that there's trade and exchange of ideas going both ways. So to be a really good mentor, learning what it means to be a good mentee and having the younger people partly be your mentors at the same time. Like if you want to learn about artificial intelligence, the chance that you should look for someone younger than you, depends how old you are, but you know, it's pretty high. So that's my starting point for thinking about apprenticeship. In which ways does the hierarchy run in both directions? And you need this good balance.

You described Silicon Valley as a diverse monoculture. It sounds like it's a good thing in your view. Can you tell people what a diverse monoculture is and why it can actually develop talent?

What's striking to me about Silicon Valley In America, it has our very smartest people. That might sound like high praise, but in a funny way it isn't. It's not clear to me they're our wisest people. They're our smartest people, and they're our most ambitious people. There are big disagreements about what you think of the values there and their wisdom and prudence. But I think the monoculture of Silicon Valley is to value the ambition and the smarts which are just through the roof in terms of how impressive it is. I don't think there's any group of anyone anywhere in the world that comes close to that. But then what people want to do with those abilities, that's what's quite diverse, and how they want to change the world. And for any point of view, you can find someone in Silicon Valley you hate, is one way of putting it, right? So it's always going to be controversial. People who are very wealthy are controversial. It becomes your audience sacrifices and so on. So I think it's one of the most fascinating cultures we have produced. It works extremely well. And especially now with artificial intelligence, which is an outgrowth of a lot of previous things that were made, including the internet, it's the closest human beings have come to producing magic. And I mean that quite literally. That's scary, but it's also fascinating. It's the most important story of our time. And they've done that. And they hadn't really done it at the time when I called them a diverse monoculture. So they're a way more impressive diverse monoculture than they were even when I said that.

So since you mentioned Girard, yeah, mimetic desire, uh, net positive, net negative for talent.

Role models are inescapable. One thing I've learned doing social science, I always see new papers about how important role models are. And it can be valuable that the role model in some way is like you. It doesn't have to be skin color or gender, though it can be. It could just be sort of attitude or how the person carries themself in public. But it's like choosing the right hierarchies to climb. If you want to be the best in the world at a computer game, I would say, well, you're probably very talented, but maybe think twice about that. You're going to copy someone, so choose that very wisely. And even when you're deciding whom to copy, you're going to look to copy, you know, other people who've copied others. So there's a kind of very difficult infinite regress in there. But I would say being aware of that is more useful than storming around and saying how independent-minded you are. You're not going to escape the Girardian logic, but you can turn it to your advantage.

Choose your models wisely.

Yes.

Is there value to being hyper-specialized? Is there value in the young Silicon Valley founder who knows AI inside and out and has religious literacy at a first-grade level because it can, you know, he's just totally focused on building the technology? Is there value in that, or is there harm in the hyper-specialization? I, I imagine the answer might be nuanced, but How do you see the— this strikes at the heart of Clooney's mission. Is it good to be siloed in some cases, or should we be trying to expose people in these different cities to new ways of thinking? Is multidisciplinary approaches— is there value in that?

First, I'm curious as to what is the first grade level of religious literacy? I had been first grade in a while. Back then it was pretty loud. But I think people in Silicon Valley, in different ways, they are looking for something. And they don't necessarily know what it is. I think there's actually two somewhat distinct groups. There are those who are Jewish in heritage, and when they're looking for something, it may not be God with a capital G they find, but they will be looking in Jewish directions. And I think that's more or less inevitable. The others are quite open and have understood that the way things were, were not working, and that there are new challenges, a lot of them coming from artificial intelligence. And you just mentioned to me earlier in our discussion, there's been some kind of document from the Vatican on AI, which I have yet to read, and it seems quite important, and you seem to think it was good. So no one really has good answers. What will the world look like? How should we behave? What is an ethical AI system? What's an ethical deployment of AI? So the notion that people will look to religion and Catholicism seems quite likely to me. And I've been saying now for the last 10 years, the most important thinkers of the next era will essentially be religious thinkers, whether they admit it or not. So there's a critical role for Clooney and many other groups to get people ideas that are useful and interesting and will develop those individuals in terms of what they're doing in the world. So there's never been a better time, I think.

Does AI play a role in accelerating their interest in religion and religious ways of thinking?

I think it absolutely does, and that's just getting started. So my best guess— and I stress the word guess— is that AI will give us a new world of oracles, a bit like what Plutarch wrote about in his long piece on oracles. That you'll have religious questions And you can go to your priest or rabbi, or you can go to the AI, and ask yourself honestly, who will give you a better answer? For the most part, not in every case, but O3 will give you a better answer. If you pay for O3 Pro, better answer yet, but O3 will be fine. So over time, people won't confuse the AI models with gods, but they will be intermediaries and treated very seriously as such. And I think that will slowly shift us, not quite toward a polytheistic understanding, but if you think of the forms of Catholicism, let's say, blend somewhat with the native religions of Latin America, which are as Catholic as any other form of Catholicism from my point of view. I think Catholicism in the United States will become more like that in some funny way. We should be ready for that, and we want that to be a good Catholicism, Christianity, relationship to God, and so on. But there's very little work that's been done that I see. LDS put out an early statement on AI that I thought was quite good. You mentioned the recent Catholic statement. So there's just so much to do. It's like the world's biggest problem and opportunity is how is AI going to reshape our vision of ourselves as humans, our relationship to God, gods, whatever, to each other. No one has any idea what to do. So that's like incredible opportunity to me.

Yeah.

How many people in the room knew that the Vatican issued a document on AI in the last 3 months?

Okay.

I think this is pretty— I see 4 hands.

Yeah.

There's, there's a few more over there, but it's not many. It's way less than 10% of the audience. And I, I would suspect this is a pretty Christian, maybe even Catholic audience. So that's a lot higher than the general population. I would, I guessed it was 1% or less.

Yeah.

When we talked earlier. So that, that in my mind is a problem and an enormous opportunity. What, what are the— so you mentioned that religious thinking will be important, but also that AI will play some role in this. Isn't it dangerous to have people asking AI fundamental metaphysical questions about human life, serious bioethical questions, what happens when you die?

Is—

are you worried about that at all? Well, aren't ideas dangerous? I mean that in a mostly positive sense, but I also mean there's true danger in it, right? What's the alternative? If you turn on NPR or read the New York Times— I do only one of those two things.

Wait a second, which one?

I read the New York Times. AI right now— it may change— but it's more objective and it's smarter than any other media source. Doesn't mean you're going to agree with all the answers. As it becomes more diverse, that story will become more complicated. But it's the most objective source we humans ever have built. And if you interrogate it about, say, the Bible, I find it is remarkably good. The Bible, biblical scholarship, different interpretations of passages, matters of translation. Because it's a text-based— the Bible having given rise to text-based religions— one of its great strengths. So I think it will help text-based religions in general, but it will also make them more different by attaching these oracles to them. And that's a real danger to navigate. But keep in mind, it's not just, well, we have a fixed vision now of how it should be and we need to impose that vision. Our own visions of how it should be will evolve with the technology, and we need to be open to what we can learn from it all, which will be a great amount.

So we're talking AI generally, but is there a specific AI that you found to be particularly good with these kinds of questions? Claude, ChatGPT, one of the other ones that's particularly good at the religious analysis?

When you say ChatGPT, there's many models in there. You know, you toggle down and you get very confused. There's 3, there's 4o. O3 right now is clearly the best of those along with O3-Prop, which is the one that costs a lot of money. And Claude is very, very good. Right now it's Claude 4. Those are the two that are best for textual analysis. The Gemini model from Google is fantastic, but it's, I would say, stronger for computational issues and numbers, which are not the topic tonight. There's other very good models, but I would start with those. If you want something zany and creative, Deepseek from China. If you want to really spice up your Christianity in a way that, yes, will be dangerous, ask Deepseek and tell him/her/it, whatever, to let it rip. You'll get some interesting ideas, definitely some dangerous ideas. It's just much less censored. Grok is its own story. Right now they're still trying to fix it. The different things that will say about Hitler. And I think they'll do that. Yeah, it's often very good, but we have to be careful at the moment.

Yeah, I find AI to be incredibly sycophantic. I'm writing a book right now.

This is far out, it's not sycophantic.

Very sycophantic. So if you ask it spiritual questions, it will make you think you're probably a saint, which can be incredibly dangerous if you're talking about moral issues.

You can put in your own prompt, you know. Sure, sure. And it will respond. So it's sufficiently sycophantic that it will abuse you if that's what you want. Yeah.

Yes.

Yeah. So to wrap up, is there such a thing— you know, typically when we talk about talent, we're talking about a certain kind of talent. What kinds of talent, in your view, are undervalued? Is there such a thing as spiritual talent, for instance?

There's definitely such a thing as spiritual talent. And I think there is often a calmness to it. It is not that showy. It is not necessarily correlated with the normal markers of worldly success. Uh, you see it when you get to know a person's network and how they are respected by others. So I think the more important that becomes, the more important networking becomes, The more we will need to assess other individuals, not by their SAT scores, their grades, how they performed on some test, or how many citations they have, but what kind of network they have and how that network relates to them. And that's quite intangible, but to judge talent, it's one of the most important skills you can learn is how to judge and learn about a person's network.

So Tyler Cowen, one of the many roles that Tyler has is running Emergent Ventures., which gives grants to talented young folks. And we've got two in the audience today, Matt and Michael, I believe. I'm not sure where they're sitting. There you go.

Um, welcome.

Uh, and Clooney also gives grants. Uh, we have one of our awardees, Sophia, in the audience who's, uh, building something related to AI for the allocation of labor. Um, the— in that, what is— how do you choose emergent ventures winners. I've applied for an Emergent Ventures grant, did not win one, by the way.

You have already emerged. That is your problem.

So you must get thousands of applications. And so there's two steps here. There's the application process and the screening, and you don't do interviews for Emergent Ventures.

No, of course everyone is— if I meet them, that's the equivalent of an interview, but everyone is interviewed. Who wins?

Okay.

Plenty of people are just no right off the bat.

How do you screen to get that many people through the interview process? What's a tip you can give those of us who receive a lot of applications for roles that we're looking for?

I think you have to do the key work in advance of getting the applications. You want to send out what is called the bat signal to get people, whether it's smart or energetic or ambitious, whatever qualities you're looking for, or spiritually important. Spiritual leaders, to get them to apply, but not to get too many other people to apply. So the signals you send out, they should in some key ways be quite obscure and at times difficult to interpret. You could say this about the Bible itself. It's incredible. It's an incredible mix of remarkable simplicity, some passages, and remarkable obscurity. And it's this amazing balance that drew in originally, you know, apostles of Jesus, uh, the Israelites with Moses earlier, but has drawn in so many remarkable people because it had that balance. And there's a lesson to be learned from that— not to be too simple, not to be too obscure, a mix of art and science. And to get that just right, because it's so much more about who comes to you because of what you radiate than how good you can be at choosing. It's a kind of arrogance to think you can ever be that good at choosing. I don't think you can. And by looking at a person's network, what you're really saying is, I realize my own humility about my own choices, but other people know something I don't. So you mentioned Matt, who's sitting in the front row. Big reason I chose him is I knew some people he had networked with, and I thought, well, if they find Matt interesting, I just barely met Matt, as he can vouch. Well, Matt must be a great pick, and he's been a great pick. So I was outsourcing that judgment in a sense.

Hmm.

So we have to put a bat signal out there, and it needs to, you know, not give away too much, but get the right people in the room. And judging by how mysterious you all seem to think this event was, I think we might have done something right.

And don't think you really know what you're doing. Yeah.

Yeah.

And like with you being to emerge, I've tried to focus on picking people who are not yet successful. I had an application once, the guy says, oh, I just raised $20 million and I'd love to add you to my cap table.

That's a no.

It was a good proposal, I think, but not going to have any chance.

Yeah.

I have two more questions for you and then we'll open it up to roughly 15 minutes of audience Q&A, if that's okay. My first, these are short rapid-fire questions. So the first one is, Tyler, you've said that you love to catch people off guard during interviews and ask them unstructured questions that they're not prepared for. One of your favorites being, what tabs do you currently have open on your browser right now? Nobody's expecting that question at all. However, I'm not gonna ask you that question, but you've now written this in a book and told everybody your question. So you can't really use it anymore.

And I don't use it anymore.

So what's, what's the new question? Um, you'll no longer be able to use this one, but you have any, any advice just to make, to quickly get to the heart of who a person is?

I try to get people talking about something they have done recently or somewhere they have been and just engage them on something they really care about. I don't have to be an expert on the topic, but I feel I can't know zero about it. Often I'll use places because I've traveled a lot, I've been to many places. That doesn't always work, but most of the time it works. And just to see how the person thought about the encounters they had with the world, that in a way that's so much one of the most important things about a person. What kind of encounters do they have with the world? And how do they understand it? It's sufficiently weird. And you could say obscure that it's hard to prepare for. And I can always pick what particular encounter I'm going to ask them about. So they can't have it prepackaged in advance. It could be, what's the last movie you saw? Or who's the last person you spontaneously talked to on the street or in a restaurant? And, uh, I find that often works and it's different every time. So it's very hard for someone to gain that.

Hmm. Hmm.

So we have with us today two special local guests, Chen Shaki, who runs the West Michigan Construction Institute and is essentially training people in an apprenticeship to be able to go into construction. And George Patey, who's a former fighter pilot and is the Dean of Aviation at the West Michigan Aviation Academy. So, And I'd like to kick off the Q&A maybe first with you, Jen, and then with you, George. But I'd like to start by asking Tyler, what is the future of those, especially in a world where AI is encroaching on more and more of our labor?

What you're calling apprenticeship is making a massive comeback. So this is a big topic, but I would say our universities are often failing us. So you need to go somewhere else. I would say AI is changing rapidly. You can't learn that knowledge by picking up a book. You actually need to access other human beings who are doing it or building it or working with it. That too is a kind of apprenticeship. Our ability to contact other people, whether it's email or WhatsApp or Google or AI, whatever, has never been greater. So to connect with the right match for you,, much more powerful the world today, even than 10 years ago, never mind pre-internet. So all that's rising in importance. This is the new model. You made a mention of it's going away. It's not going away. It's just dominating everything. Like, what's your network? Who are you apprenticed with? Who are you teaching? What are your back-and-forth trades like Buckley and Gary Wills had, or like Sigmund Freud and Carl Jung had them, right? They learned from each other. They ended up kind of hating each other. And having a falling out, but they've learned a lot from each other in the meantime. And that is the model and will become the model more and more.

Will there be a return to the guilds or will we have to build a new guild system?

I don't know if it will be formal guilds, but I think we now have informal networks doing guild-like functions in a way where there are not all these oppressive legal requirements. You need to work for 7 years before you can whatever. Those probably never made sense. They certainly don't make sense today. So we're getting some of the better side of guilds, the quality instruction, and avoiding a lot of the problems. I think it's pretty great.

Beautiful.

Thank you. Thank you, Tyler.

Thank you all.

Yeah.

We've got time for a handful of questions. I'd love to start with Jen.

I'm the president of the West Michigan Construction Institute. We are a construction trade school based here in Grand Rapids, and we're 3 and a half years old. So we're still very much in startup phase of running a brand new school. Uh, we, we serve 3 kinds of students and 3 kinds of programs at our school. So we have a high school program. It's a career and technical education program. We teach our high school students in the carpentry and the electrical trades. We're a post-secondary trade school, so we teach adult students in 7 different trades. And then we also meet our construction industry professionals where they are through a portfolio of professional development programs. Our mission is rooted in lifelong learning, specifically in the construction industry. Tyler, I've been reading your book. I am about a third of the way through Talent. I bought it 3 weeks ago, and then I got married, so I'll ask for a little grace to not finish it quite yet. So I'm coming off the heels of that.

Congratulations.

Thank you. I've been intrigued by your curiosity about question asking. And as I look at our students, what I am so deeply curious about is, are high-performing trade students different than a traditional high-performing student in a traditional education system? And what I want to capture from them is who they are so that as we're recruiting and quote-unquote selling our program, we begin to find those right-fit students. So if you were me and sitting in my school with my hundreds of trade professional students in their apprenticeship programs, what kinds of questions would you ask my high performers to help you build out that profile of what a successful trade professional would be?

I've never met your students, of course, but one place you could start is just to ask people, "How ambitious are you?" This is a Peter Thiel question. It sounds absurdly simple and that anyone could lie or BS their way through it, but in fact, I found when I ask people, I get, for the most part, remarkably honest answers. It's very hard to fake. But you can't listen for the answer itself. So there's a lot of very ambitious people. You ask them, "How ambitious are you?" They're kind of so ambitious, their own level of ambition doesn't impress themselves that much. "Eh, you know." But the degree of passion and detail they respond with is how you judge their actual level of ambition. And to get them thinking, like, "How ambitious am I?" And what does that mean? Uh, we don't do that enough as a society. And you can go through school, whatever, they'll teach you 10,000 things, maybe some of them are useful, but they don't quite teach you how to ask yourself that question and eventually answer it. So that would be one place to start. And then another question I like to ask people is your group, like your 3 to 5 closest associates, that you're in touch with almost every day. Like, how good is that group now for helping you? And what are you doing to improve that group? And just to get people thinking about that explicitly. Uh, again, super simple. Everyone vaguely knows it matters a lot, but not that many people actually wake up every day and say, what can I do to improve that group and make it bigger, better, whatever else, more diverse over time? I would start maybe in those two places.

Thank you, Jen.

Hi, my name is George Peh. I oversee the aviation programs at a charter high school at the airport called West Michigan Aviation Academy. We're a tuition-free high school. And, you know, our students have the opportunity to get a Michigan Merit Curriculum graduation certificate, but we're primarily focused on the trades within engineering and aviation. For example, students can earn and work in their aerospace engineering, biomedical engineering, computer science. On our aviation side, we can become private pilots, we can become commercial drone license owners, and then also build-a-plane students that work towards A&P mechanics. We've grown from 80 kids in 2010 with Mr. Dick DeVos's foundation of the school to over 620 kids and over 100-kid waiting lists. My question for you is kind of interesting because with the, the AI component in aviation in particular, I think— I hope it goes without saying that like to be an effective pilot, decision-making is very important and critical thinking. And so when we were kids, you know, we looked over and if we needed to cheat, We looked over at Johnny's and looked, right? And then as they got a little bit older, then the internet happened. We were click and paste, cut and paste. And now what we're seeing is that my question to you is that like, do you think that AI is kind of going to this point where it's robbing our talent of critical thinking and decision-making ability and how that's gonna impact us in the future? Like how do we, in our educational institutions where we're working with these students that, They need to understand the tool, but yet is it robbing them of some of these essential skills to be successful out there?

I think of AI as like books or many other media, which would include YouTube, the internet. You can use it to make yourself stupider or smarter. Right now, our educational institutions, they see it as a threat, a threat to be kept out. Students can cheat. I understand why they feel that way, but it also means they're not doing anything to teach the students how to use AI to make themselves smarter. So AI, for instance, is a great way of learning how to ask great questions. Uh, we need to restructure our schools so the faculty themselves are using AI to make themselves smarter, and only then can they teach it to the students. We're in a funny position where often the students, even quite young ones, know more about AI than their teachers, and they could be teaching the teachers rather than vice versa. That's not the case in most areas, right? Certainly not in flying a plane. So we have a lot of work to do, and faculty can be resistant, or just they're quite busy, have their own lives, whatever. Uh, I think it will take a long time before we make that adjustment, but the sooner we start the better it's going to go. And we're not going to succeed in keeping it out any more than we've kept the internet out of people's lives. I know plenty of people who have been made stupider because of the internet. That's bad. Rather than try to ban the internet, again, we need to build a better internet, put more good content on it, teach people how to use the internet better. That's a tough battle, and we're going to need to do the same with AI.

So, um, somewhat actually piggybacking off the last question but in a different context, thinking about our public intellectuals. So I'm personally a big fan of Jordan Peterson. I followed his career going back to late 2017, 2018. But he said something interesting about his most recent book, Will You Wrestle with God, on the Joe Rogan Show. He said that he'd had somebody build an AI trained on Peterson's own material, and then Peterson brought it exegetical questions about the Old Testament and was like, you know, can you answer this question because I'm not sure I understand this passage, what do you think, asking the AI. And so of course the AI is going to give Peterson an answer that sounds A lot like what Peterson would say. But like, and so then he said a lot of his book was just written that way, basically kind of like co-written with this, his own AI. And thinking, man, I didn't— like, my motivation to read the book went way down after I heard that interview, you know. It sounded a lot like, you know, Augustine talks about this, this incurvatus in se, where you're like curved in on yourself and you're not engaging with the world out there. And It would have been great to hear that Peterson was engaging with, you know, people, commentaries, like the wisdom of the past. So a little bit similar to the last question, is there a threat there as well?

Jordan made at least one big mistake. First, at the current stage of development, it's a big mistake to train an AI on your own writings. It actually makes the AI stupider for the most part. So 2 years ago, I thought of training an AI on my writings, and I went and I spoke to some experts, and I asked them what they thought of this idea. And they said the AI is not good enough yet, but even if you want to use AI, you'll do better by taking the full model and asking it what would Tyler Cowen say, rather than training it on Tyler Cowen. You train it on only Tyler Cowen, you're restricting its knowledge rather than having it draw the knowledge of the whole world. So like, big mistake right off the bat. But I also think AI right now is not good enough to write books. It is very, very good at fact-checking things I write, which I do all the time. It is very good at answering questions. It's very good at helping you find examples. So I'm writing this book on mentoring. I asked, uh, a free model recently, what are some examples of famous mentees who fell out with their mentors. And it gave me the example of Jung and Freud, which I had not really known of, maybe in the back of my mind, but I wasn't aware of it. So I went out, I've been reading their correspondence, which is fascinating. That will be in my book. I wouldn't have found out about that if not for the AI. That's great. But the book, the words will all be me. And at least circa 2025, I think if you do it some other way, It's quite a mistake. As AIs develop, we'll see. We need to keep an open mind. Uh, yeah, I think Jordan did it the wrong way.

We have time for two more questions. Um, I'm wondering now, in the age of AI, how, as a young person who was interested in intellectual things and intellectual ideas, you know, the, uh, value of specializing in, uh, intellectual endeavors. So for example, and Luke, you kind of specialized in René Girard and Mises, and that's, uh, led you to a career— a part of your career has been that. Um, is there a value in taking one thinker or one topic, maybe like classical liberal economics, and going for it? Uh, will AI just be smarter than us and we won't be of any value anymore, or What do you think?

AI is smarter than us now in many ways. So if I took like a 50-question exam on economics and the best AI model took the same test, it would beat me. That's not the only kind of smarts, but it's significant kind of smarts. That said, AI on its own cannot right now do anything. It needs direction, it needs evaluation, it needs context. It needs management. It needs coordination with a mix of other humans and other AIs. So I'm trying to restructure what it is that I do. I can tell you my intuition for the future, and that is people who are either very good specialists or very good generalists, and they know how to work with AI, they will do very, very well. And if you're not one of those things, you might have a lot of trouble. And it will take us a while to get that through our heads. Simply being pretty good at a bunch of things and not being great with the AI, that's what I worry about the most. Now, if you're doing physical labor, of course, you're, you're quite protected in your job. But if you're an ideas person, as most or all of us here are, I think extreme specialization or extreme generalization and knowing a lot about AI is the way to go.

Um, my question has to do with, um, the imagination and its correlation with ambition. Um, I have a number of friends who are in education, and they talk about, um, the challenge of a well-developed imagination in students. And something I hear maybe in this, uh, conversation is the use of AI as a tool for question asking and maybe to help develop the imagination and how that might correlate with ambition in the way of talent?

AI, the internet also could expose you to entirely new worlds, show you how interesting they are and how exciting they are. I think you still need human guidance in that process and you need role models, you need mentors. It could be parents, siblings. So it's still a story fundamentally of complementarity. But any new thing you want to learn about, and I do mean any, AI can do it. It's quite remarkable. And when I before called it magic, it does to me feel like I'm living in a world where magic is real. That does partly make me nervous, but it's here and we need to go ahead and, you know, tame the magic and turn it toward being better, deeper, or caring, whatever, you know, human beings. I don't see why it can't do that. It can be useful as a therapist. It can be very good for self-critique. You have to ask it so it's not just a suck-up. Those are things we need to learn over time. Like, how long did it take us to learn how to make TV pretty decent rather than terrible? You would say decades, I think. We need with AI to do it much more quickly than that. Like, again, it's really a tall order. I'm actually optimistic AI will help us, in fact, but we've still got to do it. Like, the clock is ticking now. Let's do it.

Thank you.

Now I will invoke my right as the host to ask the last question because your questions have brought something up for me, especially yours. About the hyper-specialization. You're right, I wrote a book about René Girard. I don't think I have any particular talent in talking about René Girard, but I'm writing a new book. And I listened to Tyler's podcast with David Perel on writing and AI, which affected me. And I was in the middle of writing my book when I read that. My question is, is there danger in trying to AI-proof yourself?

Or your writing.

And what I mean by that is there— it's almost like the danger in trying to be anti-mimetic is a little ridiculous because you just end up imitating something that you don't even realize. I've tried to AI-proof my book by talking about incredibly, uh, emotional things like my relationship with my father. It was Alzheimer's. Like, well, AI could try to replicate that, but it has no experience of ever doing that. Where do you stand on this idea of— is it, is it wise to try to AI-proof ourselves, or is it inevitable that we will end up producing something that is derivative anyway?

I can tell you what I'm doing for myself. I don't think it's the same answer for everyone. So I'm fully aware of the fact that if I ask AI to write a 900-word newspaper column, which I do all the time, it's often very good. You can debate if it's better than me. I think a lot of readers don't care if it's better. They want to hear what Tyler says, in kind of a silly way maybe. But like, I can keep on going with that in terms of my profession, I think, until my profession is over. But nonetheless, I don't want to be a secondhand, second-tier version of something that's better. So the more I see of AI, the more I avowed to go out and meet people face to face and to travel and have person-to-person encounters and to talk to crowds and work more with mentees, mentors. On average, they're younger, but they don't have to be younger. It's great when they're older than my 63 years of age, but of course most of them will be younger. That's how I've been AI-proofing my life. Now, I think that's a good thing. And I've been doing it now for almost 2 years. I've been very happy doing it. It's more rewarding for me. I think it's probably better for the world, and I still have time to write. I'd say I write a bit less because I know the AI can copy that in some different ways. I don't think AI-proofing is necessarily a good thing if you're doing it just to run away from the AI. Is probably a mistake. But it's also going to depend on your age. So I mentioned that I'm 63. I feel quite protected from AI. So I don't know where AI will be 20 years from now, but I know where I'll be 20 years from now. Like, one of two places. You know, in my home, not doing maybe that much, or not on this earth at all. And that's like, it's an easy thing to plan for. But if you're 23 and you're having to calculate where will AI be 35 years from now, and maybe the AI will invent new biomedicines that will help you live to 97, so you've got to even think further ahead than you might have thought. It's a blessing, but still it makes life more complicated. I don't think you can just neglect the notion of AI-proofing at all. You have to realize it's competition. Just like, say, manufacturing employment in this country has fallen from about 35% to 7% from the 1970s to today. It's had a big effect on the Midwest, probably Grand Rapids. Uh, you can't do nothing about that. So we're seeing a new version of that. Grand Rapids is doing great with AI. You know, people in general can do great, but Grand Rapids did great by realizing the world is changing. You have all these new sectors, new manufacturing, healthcare, other things I don't even know about. It seems to me your culture entertainment sector is quite vibrant. Like, you did it. Congratulations. But you were aware of what was happening. And maybe it was hard at first. I wasn't here then. So I would say take this very seriously. But if your only plan is to run away from the AI, it's like having your only plan in life be to run away from the bear. And you can say to the guy next to you, I don't have to be faster than the bear. I only have to be faster than you. Like we all know that joke. But if that joke is all you've got, you're spending your whole life running from the bear. And that to me is not a life worth living. So start with the question like, what is my ambition in life? So it's much more than just running away from the bear, and take it from that.

Thank you. On that note, thank you so much, Tyler Cowen, for being with us tonight. Thank you so much.

Thank you.