Stanford’s Silicon Valley pipeline, investigated

Note: Transcripts are automatically generated from episode audio, and are not fully corrected for spelling, grammar, and formatting.

BAKER: Stanford likes to think of itself as an underdog. That’s not what Stanford is anymore. Stanford is this huge, powerful institutional juggernaut with a reputation to protect. This is more like a business than it is a school, and students are the product. I watched one friend go on to fake a launch demo, and she’s feted by all of Silicon Valley. She’s put up in a mansion, and she’s celebrated for technology she doesn’t have and didn’t create. These kids are taught that this is the path to success, that the way to assume their place among the rightful ruling elite is by trampling on everyone else’s shoulders to get there.

I stayed at Stanford. I love Stanford. And I think if you love an institution, you want it to be better. And that starts by acknowledging that the same set of incentive structures that have allowed Stanford to get very, very wealthy have also allowed for some of its corruption.

EL KALIOUBY: Theo Baker just graduated from Stanford. I first came across Theo through his New York Times op-ed, What AI Did to My College Class. It was so interesting because he gets to reflect on being the first cohort of college students to experience ChatGPT through all four years of school. During his freshman year, he reported for his school newspaper, and his stories led to the resignation of Stanford’s president, and he became the youngest recipient of the George Polk Award. He’s also written a book called How to Rule the World, where he pulls back the curtain on Stanford’s culture, its deep ties to Silicon Valley, and the ways AI has become an accelerant to Stanford’s ethos of winning at all costs.

I’m Rana el Kaliouby, and this is Pioneers of AI, a podcast taking you behind the scenes of the AI revolution.

[THEME MUSIC]

Hi, Theo.

BAKER: Hi there.

EL KALIOUBY: Welcome to Pioneers of AI.

BAKER: Thank you for having me.

EL KALIOUBY: I read your op-ed in The New York Times, What AI Did to My College Class, and I sent it to my kids. I sent it to a lot of people in my network, and I loved it. So thank you for writing it.

BAKER: Well, that’s very kind of you to say. Thank you.

Copy LinkHow Stanford became a magnet for tech ambition

EL KALIOUBY: I want to dig into it, but first I want to talk about the book and your experience at Stanford. What got you into Stanford in the first place? How did you become interested in it?

BAKER: I came out here thinking I was going to be a techie, a complete diehard lover of technology, just because that was what I had spent my whole life doing. I was the kind of kid who would spend a Friday night working on a machine-learning model and call that fun. I remember seeing these kids in their Stanford T-shirts and flip-flops lounging in the shade of a palm tree, leaning up against a self-driving car they had just helped build. And I thought this was the coolest place on Earth.

EL KALIOUBY: I have to admit, as I was reading the book, I kept thinking about my own experience. I founded my company out of MIT, not as an undergrad but as a postgrad. I was doing my postdoc at MIT, saw this opportunity, started a company, and now I invest in AI. We find a lot of our founders on campuses like Stanford and MIT. So I’d love to have a conversation where we almost compare experiences of that ecosystem, with you representing what you experienced at Stanford, and I’ll share my experience at MIT. We can have two data points, kind of thing.

BAKER: Sure.

EL KALIOUBY: I was really struck by the secret class that you found at Stanford, and how it unlocked this whole world of a Stanford within a Stanford. Can you tell us more about that? Bring it to life.

BAKER: The title of the book is not a metaphor. There’s literally a secret class at Stanford called How to Rule the World. These 12 students a year are plucked from the crowd through this cloak-and-dagger admissions process by a Silicon Valley CEO who styled himself a professor, because you’re not actually getting course credit for this, although it meets once a week in the winter quarter and there are discussions, lectures, and guest speakers. It’s more like Skull and Bones for the aspiring tech elite. And this guy, Justin, teaches students, in their words, how to extract value from people. That’s the lesson you’re learning: how to extract value from those around you. I just find that so fascinating.

I think it’s a great expression of what has come to roost here at Stanford: all of these people trying to network with the teenagers they think will be useful to them in the future.

Copy LinkInside the secret class training a future elite

EL KALIOUBY: Bring this to life. You’ve got all these potential founders on campus, and you’ve got all these VCs and investors roaming around. How does that manifest? And by the way, I think if I were at Stanford, I would not have been invited to this class. You ended up not being selected to take this class, correct?

BAKER: I was not, no. At that point, I’d already started becoming an accidental journalist.

EL KALIOUBY: Do you have FOMO, or not really?

BAKER: About the secret class? Not particularly.

EL KALIOUBY: Yeah, OK.

BAKER: But I ended up interviewing seven people who were in it, and they had really interesting things to say because it is a really interesting cohort. Shockingly, if you take the most insider kids at the school who have the highest ambitions possible, it’s a pretty interesting mix of people. One of them was Phoebe Gates, Bill and Melinda’s daughter, who just raised her $31.5 million round. A lot of kids have gone on to do really interesting work. But the class is fascinating to me because, on the one hand, some of what this guy Justin says, like, “Exceed the limits of your conceivable ambition,” is not necessarily a bad message.

But where people start to get into trouble at Stanford is when you have this massive, unconstrained ambition and it is not tethered to a culture of accountability. People are taught to make very, very big promises and shown they will be rewarded for it. Then they learn that if they cut corners or start to deceive, they’re not going to get called on it. That’s how the culture of fraud develops from this root, I would say.

EL KALIOUBY: Well, it’s a lot of this fake-it-till-you-make-it mantra, right?

BAKER: Yeah. And look, Silicon Valley has taught people that the only thing that matters is whether you win in the end. As journalists, I think we think of Reddit as one of the better-governed social media companies. The Reddit founders admitted that they faked all their user data for the first two years of the company, and it doesn’t seem to matter because they did make a great company, or whatever. In the immortal words of one of my colleagues when we were working on running the campus hackathon freshman year, we had hundreds of thousands of dollars from corporate sponsors, zero contracts or oversight, and predictably messy results. We were talking about it, and she said, “It’s all just a little bit of fraud.” Her phrase was, “Just a little bit of fraud.” I feel like that is what has gotten baked into the system: just a little bit of fraud.

Copy LinkHow fake it till you make it turns into normalized deception

EL KALIOUBY: I think what you’re saying in the book, too, is that this has become the culture. My son Adam is 17. He’s a rising senior in high school. I want my kids to be very ambitious, but I also want them to have integrity. I am worried that if a young person ends up in a culture where there’s so much pressure to achieve, and almost a precedent for not sticking to the facts, I don’t know what the word is, maybe fudging the facts, what would you call it? This loosey-goosey culture that you talk about.

BAKER: Yeah. I watched this happen with people I arrived with freshman year, people I knew who were brilliant and idealistic and, I believe, genuinely committed to technology. I watched them learn, in the words of one of them, that you’re taught how to lie to VCs and misrepresent yourself. Then I watched one friend go on to fake a launch demo, and she’s feted by all of Silicon Valley. She’s put up in a mansion, and she’s celebrated for technology she doesn’t have and didn’t create.

You see it happening. These kids are taught that this is the path to success, that success is about winning at all costs. The ends justify the means. And that is the classic tale of exploitation people are taught, that the way to assume their place among the rightful ruling elite is by trampling on everyone else’s shoulders to get there.

EL KALIOUBY: I’ll be right back with more of my conversation with Theo after this short break.

[AD BREAK]

I think back to the time when I spun out of MIT with my company, Affectiva, and I actually feel very lucky. Once we determined that we were sitting on very interesting technology that was commercializable, and I didn’t have anybody in my family who was a previous entrepreneur or anything, it was not on my mind to start a company. But I will say, immediately I got connected to fellow founders. I got connected to a lot of mentors and a lot of investors. I actually feel very lucky and privileged. I know this is a privilege, that I had this sort of access. Even if MIT had an MIT within an MIT, I was not in that inner circle, but I still feel like I was lucky enough to have doors open for me. You talk a lot about that, almost at an extreme, like the wining and dining, the yachts. Give us a few examples of what that looks like.

BAKER: Students learn pretty quickly that self-presentation is the thing that governs all. If you learn how to say the right words, speak the right lingo, and know the right people, you get plucked from the crowd. And it ends up not being about your talent because nobody’s actually going to check up on it. It’s like, do you know the right person? Then they’ll refer you to the next one, and then so on and so on. It’s referrals all the way down. And that’s how people end up at the yacht parties with the slush funds and the quote-unquote “pre-idea funding,” which I still find to be such a shocking concept.

EL KALIOUBY: Talk about that. I see a lot of this, but tell us what pre-idea funding is.

BAKER: I mean, if you’re one of these really promising teenagers, a VC will hand you money before you even have the glimmer of a concept for a company in mind, just because they’re so desperate to network with the people they expect to be the next trillion-dollar company founders. And it all makes sense by the logic of Silicon Valley mythology. If you’re getting in on the ground floor of the next Google or Instagram, you’ve made your own career. But the way it works in reality can be very silly in so many ways.

It’s not that this system doesn’t work or can’t work. In many cases, these kids are brilliant, and the technology they’re producing is very impressive. It’s just that we only ever talk about the good side of the story. Stanford is very good at its sort of self-presentation. It’s showing the version of itself that it wants the world to see. And I think what my reporting has shown is that there are a lot of things that the school would rather sweep under the rug and not talk about. And that’s why we see the Elizabeth Holmeses of the world or the Sam Bankman-Frieds come to be.

Copy LinkWhen a university starts acting more like a business

EL KALIOUBY: So the book is very much focused on your freshman year. And you came to this place with a lot of admiration and a lot of hope. Was it quite shattering to see this other side of the school?

BAKER: It was definitely interesting. Stanford likes to think of itself as an underdog, that this is the place where renegade outsiders are building for the hell of it, and that it’s this quirky, unique place. That’s not what Stanford is anymore. Stanford is this huge, powerful institutional juggernaut with a reputation to protect, and it acts that way. This is more like a business than it is a school, and students are the product. You see that reflected time and again in the book, from something as theoretically quaint as the war on fun, which ends up being the first story that I report for The Stanford Daily, all the way to what I see reporting on the Stanford president, Marc Tessier-Lavigne, and the work I’d done to uncover the allegations of research manipulation in papers that he had co-authored.

EL KALIOUBY: Which ultimately ended with him resigning.

BAKER: Yes. And as I report for the first time in this book, he was unanimously ousted by his own board of trustees. This guy starts the year by opening Stanford’s first new school since 1948. A billion-dollar donation from John Doerr enabled the Doerr School of Sustainability. He starts the year with maybe his greatest success as the president of Stanford. He ends the year ousted unanimously by the board, which faulted him for a few different things, including the dynamic that he created in his labs.

There was an incident in which a younger female researcher apparently generated results that challenged some of his previous research, and he effectively tried to destroy her career over it, according to what the trustees learned from their investigation. There was also the highly aggressive public posture that he and his lawyer began to take in response to my reporting, and his failure to correct the scientific record at various points. Those were the reasons they cited when ousting him.

EL KALIOUBY: I want to flip it a little bit, because I’m an investor now. As I said, I think there is a very interesting relationship between academic innovation hubs and sources of capital, like venture funding. I’m on the hunt for people who are like me, who spent time innovating at schools and creating IP, and then I want to back them and support them. So that’s the story I tell myself, but am I evil? Have I switched to the dark side here?

BAKER: No, I don’t think investing in a company is evil. I think things can become predatory very quickly, and there are VCs who take that even further. The number of sexual misconduct allegations that are just sort of hovering around is not super pleasant. But that’s the logical end result when you have adults who are spending more time with teenagers than with their own families, and who see them basically walking around with a dollar sign hanging over their heads.

It’s a pretty interesting thing to see a teenager and think, “Oh, that person who is about to go to his first party is going to make me a lot of money.” Or to see it the way some VCs I’ve talked to do, where the calculus is so cold that when you’re making these really, really early-stage investments, you know that only one out of a thousand investments, or whatever, is actually going to pan out.

So any individual teenager is sort of useless to you. They don’t actually matter to you. It’s just whether you can get one of them to work. That being said, it is an amazing thing to be told as a teenager, “Go out and build what you want to build, and there will be someone there, theoretically, to support you.” Without it being coupled with a culture of transparency and accountability, things just go awry again and again.

Copy LinkWhy founder hype and easy capital distort judgment

EL KALIOUBY: We meet a lot of founders who are raising millions of dollars without a pitch deck. Maybe they do have an idea. Often they will have an idea, but sometimes there’s no pitch deck, there are no slides, there’s no product, there’s no revenue. Talk about some of the examples you’ve referenced in the book, and maybe some examples you’ve seen as well.

BAKER: This also gets us to the AI thing. I arrived on campus right at the zenith of the crypto craze, and it was just about a month in when Sam Bankman-Fried came tumbling down. Sam Bankman-Fried, notoriously the head of FTX, crypto wunderkind, advocate for EA, Effective Altruism. It turns out that he had been misappropriating customer funds in the nine-figure range, maybe 10-figure range, to basically make his own investments on the side. He is now in prison for that. And this is a guy who grew up at Stanford. Both of his parents were professors at the law school. They taught, among other things, legal ethics. And I learned that he actually was invited to give a lecture to a Stanford tech ethics class that he couldn’t give because he was on house arrest a thousand feet away from the auditorium, which is just very on the nose.

Right after that, I’m at a dinner with some teenagers and an investor, the kind of thing that happens all the time for a certain cohort of Stanford teenagers. And I’m talking to this one guy who’s a big crypto booster, and he’s like, “Yeah, building is paused right now. Everyone’s just worried about legality and trying to figure out how we get around this,” he says. Then, within less than a month, ChatGPT comes out, and suddenly everyone knows exactly what they’re going to be doing.

They’re going to make their fortunes off AI. And lo and behold, it’s very easy to do so. The guy who taught me how to shotgun a beer freshman year dropped out six months after starting his company with AI in the name. It’s valued at over a billion dollars, which is, by the way, more than $80 million per employee of the company because they’re so small.

We see this reflected in Perplexity, which was literally started the month I entered college. It’s now valued at over $20 billion. I think it’s $22 billion. It’s a wrapper startup that appears to have no particular technological edge apart from the fact that it has put AI in a search format. Why is that a $22 billion company? I really struggle to explain that to people on the East Coast who are asking me, “Hey, what’s going on there?” I struggle to explain why that is worth $22 billion.

EL KALIOUBY: I hear you. There are a lot of examples of that, and I do wonder what’s going to happen to some of these companies.

BAKER: The numbers are just sort of bandied about in this meaningless fashion. OK, Ilya Sutskever: brilliant, brilliant AI engineer. Literally no one will dispute that. Incredibly talented, and he probably will do something amazing. But he starts his company, SSI, with a 300-word placeholder statement. No articulation of a product, much less revenue, much less a working product, and no promise to release a product anytime soon, yet it’s valued at $32 billion, which is higher than Hyundai, which makes 4 million cars a year. So they’re making 4 million cars a year, and these guys have made nothing yet. They might make something, but $32 billion is a hell of a lot of speculation, in my opinion.

I’m not so worried about the fact that AI will be transformative. I know it will. It’s all of these associated companies that are glomming onto the craze. It seems hard for me to imagine that this limitless spending and limitless valuation trend continues forever.

Copy LinkWhat ChatGPT changed for students in the classroom

EL KALIOUBY: Don’t go anywhere. We’ll be right back after a short break.

[AD BREAK]

OK, so ChatGPT arrives on campus. What was your experience using AI, both in the classroom and outside the classroom?

BAKER: I’d always been interested in AI. This was the kind of thing I worked on in high school. I came here thinking I was going to work on NLP of some kind.

EL KALIOUBY: Natural language processing.

BAKER: ChatGPT, I think, was the moment it became a product. It went from a research experiment into something that people started relying on. I find it so fascinating that people anthropomorphize it. They call it Chat, which is the same word that streamers have been using for years to refer to their anonymous video chats full of people commenting on their video games. And now we’re literally treating life like a video game, because what is the difference between the nameless, faceless people you’ll never meet over the internet and a statistical approximation of the same thing?

People ask Chat how to get a date and what to do on weekends, and they feed their class readings into it. You’ll literally see people in a seminar asking ChatGPT what to say out loud to the class, which is so self-defeating because, by the way, the sticker price to attend a university like Stanford is nearly $100,000 a year, which is maddening. But it makes more sense when you understand the logic these universities have embraced.

Education is actually not the first priority. For a lot of students, it’s getting wealthy. So people come in here, and classes are far from their top priority. Learning is not really the goal. And so this cheating wave with AI, which is really omnipresent, sort of makes sense when you think about it from that standpoint.

EL KALIOUBY: You have a statistic in that article: 49 percent of the 849 computer science majors who responded to an annual campus survey said that they would rather cheat on an exam than fail. And that really shocked me.

BAKER: Yep, at Stanford. But that’s the win-at-all-costs mentality. You have to do what it takes to get ahead, because students know that the rewards can be extraordinarily lucrative. The frat bro who wanted to send disappearing nudes to a girl drops out and becomes the world’s youngest self-made billionaire by creating Snapchat. The two guys who get food delivered to their dorm become billionaires by creating DoorDash. The kids know that this is the mythology they’re stepping into, and the rewards can be really, really high if you figure out the right self-presentation.

EL KALIOUBY: Do you use AI at all? And actually, did you use any AI to write the book?

BAKER: Not for writing the book. I use AI mostly for coding. I find that to be the most indisputably useful application. I also find it useful for data analysis stuff. It can speed up the process there. I think of AI as a great accelerant. It takes all of these trends that are already brimming beneath the surface and accelerates them to an extraordinary degree. We talked about that with the cheating thing, how it makes sense because this attitude has already been inculcated on campus, that education is the second priority.

But it’s also the stratification. The most valuable researchers are more valuable than ever before, but the vast majority of kids are seeing the ladder pulled up in front of them. Stanford CS enrollment went down for the first time last year in over two decades because the average CS student might not get a job anymore, given that Claude is better at coding than we are.

Copy LinkHow AI is reshaping computer science careers

EL KALIOUBY: I’m curious: Your fellow graduates who are studying computer science, do you see them struggling to get jobs?

BAKER: Yeah. It’s a pretty brutal job market right now. I think the entry-level stuff is somewhat easy to automate away, particularly with CS. Also, by the way, when we talk about the AI bubble, which may or may not be a bubble but is a bubble in some ways, part of it is just because in San Francisco, the thing people are convincing themselves of AI’s near-sentient abilities with is coding, because that’s the thing it’s best at and the thing they’re using it for the most.

Other people are being exposed to the other parts of AI, which are getting much, much better in very quick order but still aren’t quite as good as the coding aspect. I do think that’s part of where the disconnect in rhetoric comes from. Silicon Valley has basically made AI best for its own use case.

EL KALIOUBY: I do believe, though, that you still need, to your point about the stratification, really awesome computer scientists to supervise and also build the next wave, to continue innovating on the AI side. Do you believe that to be true?

BAKER: Yes, but I think a lot of the stuff that a junior engineer might have done in the past, you’re not going to have to spend hours and hours doing that anymore. As much as anything, it’s sort of a vibes thing. We don’t know what’s going to happen. I know this guy who went to Anthropic who graduated two years ago, and he came back to give a lecture class earlier this spring. So it’s a hundred students, and he’s talking about AI and AI safety and whatnot. And all the students want to ask him is, “What do we do? What jobs do we go for? What is happening?” And he said, “I have no idea. If I were in your shoes, I wouldn’t know what to do.” This guy, who’s only two years out of college, already thinks that things have changed so much since then that his own advice doesn’t apply.

EL KALIOUBY: When I studied computer science, there were no ethics classes at all. It was just like, “Here’s how to code, here’s how to be a 10X or 100X or whatever engineer.” A lot of schools now do have ethics classes, but it sounds like it’s still quite detached from the actual experience of being a builder and also being a founder.

BAKER: Yeah.

EL KALIOUBY: Is that true?

BAKER: Yeah, I do think it is. I remember a guy who told me he was taking a business ethics course, and he said, “Well, we’re all there to figure out how they got caught,” which I just thought was quite the sentence to say.

Copy LinkWhat it takes to rebuild a culture of integrity

EL KALIOUBY: But how do you change a culture so that… Because I love the building side of it. I love the ambitious side of it. But how do you do that with integrity, with ethics, with a human-centered point of view? Do you see a path where the two universes kind of converge again?

BAKER: Well, I do think it starts from the top, that leadership matters.

EL KALIOUBY: I agree.

BAKER: And the priorities of a leader are reflected in the character of their community. Look, in many ways, what happened with the former president, Marc Tessier-Lavigne, is a good example of how students are learning to behave. Marc Tessier-Lavigne never falsified data himself. He was never accused of that. But at various labs that he oversaw in his career as a pioneering neuroscientist, data was falsified by researchers working under him. He put his name on the papers, and then, at various points when concerns were raised to him, he didn’t address them. He, in the words of some of his colleagues, swept the issues under the rug.

That’s sort of what students are being trained to do here. I think that approach has deleterious effects for the broader tech ecosystem because this doesn’t benefit any of us. Transparency actually is not a controversial thing, or at least it shouldn’t be. It’s not some outsider crusader saying, “Tear it all down.” I stayed at Stanford. I love Stanford. And I think if you love an institution, you want it to be better. And that starts by acknowledging that the same set of incentive structures that have allowed Stanford to get very, very wealthy, this no-boundaries relationship, have also allowed for some of its corruption.

EL KALIOUBY: Yeah. You write in your book, and I loved that line, that you came to Stanford because you believed in the power of technology to change the world for good. Do you still believe that?

BAKER: I do think it can affect change. I think technology is a neutral thing for the most part. Its applications can have very different effects, but the actual technology itself is still very exciting to me. I love the idea that [inaudible 00:27:48], the technique that underpins so much of modern AI —

EL KALIOUBY: Do you want to explain it?

BAKER: It’s basically the central mathematical technique that allows machine learning models to learn from data. I love that kind of stuff. The path of technology isn’t always a straight one. Our GLP-1s famously emerged from Gila monster venom research in the 1990s. And so that, to me, is the exciting part. The applications become much more fraught, particularly when the profit incentive can lead people astray. But the technology part itself is still pretty cool.

Copy LinkHolding on to the original promise of technology

EL KALIOUBY: You’re very impressive. You’re just graduating, you have a book, and you’ve done so many things already. But you did this while at Stanford. So in a way, it’s almost like Stanford achieved its goal, because you didn’t turn out to be another tech bro. You were able to still chart your own course at this school. So I would just love to hear your thoughts. When you look back at your experience at Stanford, what do you make of it?

BAKER: I think Stanford is always going to come out on top, some way or another. I remember freshman year, soon after we won the George Polk Award for our reporting on the research misconduct in the president’s labs, the tour guides started walking their tour groups past the Stanford Daily and saying, “And our independent school newspaper just became the first student newspaper to win a George Polk Award.” And shockingly, of course, they weren’t mentioning the content of the reporting, just the award. And the Polk Award is a big prize in journalism. It doesn’t usually go to teenagers, but they awarded it to our work on Stanford president Marc Tessier-Lavigne, and the work I’d done to uncover the allegations of research manipulation in papers that he had co-authored. I’m not sure that Stanford is entirely pleased with the fact that I’ve been doing this work, but c’est la vie.

EL KALIOUBY: All right, last question. As I said, Adam, my son, is 17. He’s a rising senior in high school.

BAKER: Wow.

EL KALIOUBY: What advice would you give to him and his friends?

BAKER: Well, I think it’s important to understand that wherever you go, Stanford or any other institution, there will be massive institutional pressures that will try to shape you or influence you in one direction or another. And it’s important to hold on to why you’re actually doing whatever it is that you do, so you’re not just doing it because the guy down the hall raised a billion dollars for a new company. I think it’s important to remember that, at 17, you can have your own values, and you don’t need someone else to tell you what makes you valuable to the world. If you want to be a consultant, great. If you want to do tech, great. But don’t do it because someone else is telling you to.

EL KALIOUBY: I love that. Stay true to your values. I love that, Theo. Thank you so much for joining us, and good luck with graduation and what’s next.

BAKER: Thank you.

EL KALIOUBY: I will be following you. I’m rooting for you. Thank you for making time for us.

BAKER: This is awesome. Thank you so much for having me. This was great.

EL KALIOUBY: I loved Theo’s advice to Adam not to let the pressure of what success looks like influence you. To dive more into Theo’s experience and his reporting, check out his book, How to Rule the World, wherever you buy your books. If you have any thoughts, questions, or impressions, please let me know. Thank you so much for listening. We’ll be back next week with a new episode.