# Trojan Talks Ep. 3: President Beong-Soo Kim with Eric Schmidt

*USC* · https://www.youtube.com/watch?v=T2gEs7dRk-w

## Summary

In this episode of Trojan Talks, President Beong-Soo Kim welcomes Eric Schmidt, who opens with an AI-written sonnet about the Marshall School. Schmidt argues that AI will both enhance and disrupt higher education, but that universities, teachers, and jobs will adapt rather than disappear. He discusses democracy and AI, China and foreign students, the need for universities to secure compute, and the coming agentic revolution, closing with an optimistic message to students.

## Key points

- Schmidt argues both AI narratives are true: universities will adapt, driven by graduate students and postdocs rather than faculty.
- He predicts jobs will change but not disappear, citing the persistent truck driver shortage despite automation.
- He warns against outsourcing democracy or human dignity to AI, the theme of his New York Times editorial.
- He strongly favors keeping foreign, especially Chinese-born, graduate students in America by stapling a green card to their diplomas.
- He says China dominates EVs and batteries and may dominate humanoid robots, but lacks the financial depth to fund AI at America's scale.
- He advises USC to build a large data center, get donor-funded chips, and have students build an agent orchestration framework.
- He describes the agentic revolution as the next big thing, illustrated by his five-agent house-buying workflow.
- His message to students: stop complaining, think in terms of scale, and participate in the agentic revolution.

## Notes

### Opening
President Beong-Soo Kim welcomes Eric Schmidt, who reads a sonnet about the Marshall School that he says took him one sentence to write with AI. Schmidt calls the arrival of AI a historical moment comparable to electricity.

### Universities and Jobs
Schmidt says both narratives—AI as enhancer and disruptor—are correct. American universities (18 of the top 20 globally) will adapt, driven by graduate students and postdocs, not faculty. Teachers, universities, and jobs are not going away, but roles will change. He cites the truck driver shortage as evidence automation doesn't eliminate jobs, and argues democracies must grow through productivity, which America does well and Europe does not.

### Democracy and Human Dignity
Referencing his New York Times editorial, Schmidt warns against letting AI run society—appeals to authoritarian AI solutions would be a loss of human dignity. He suggests AI could instead be used to encourage constructive dialogue.

### Social Media and Critical Thinking
He discusses Jonathan Haidt and Scott Galloway's work on loneliness and young men, and a thought experiment about a child whose intelligent toy bear becomes their best friend, eroding social skills. On students prompting LLMs instead of writing, he trusts faculty to detect cheating, but worries about reduced critical thinking from phones and social media. His trick: airplane mode.

### China and Foreign Students
Schmidt praises Chinese-born researchers as essential to American science and says the U.S. should staple green cards to their diplomas. China dominates EVs and batteries and likely humanoid robots; DeepSeek's R1 reasoning advances are real, but China lacks scale and financial depth. Chinese models are open source; American firms will likely release portfolios of models at different price points.

### Advice for USC
Build a big data center (jokingly suggesting a Republican state like Nevada), get donors to fund Nvidia or Google TPU v7 chips, teach prompt engineering, teach curiosity, avoid LLM lock-in, and have graduate students build an agent orchestration framework. He illustrates agents with a five-step house-buying workflow and calls the agentic revolution the next big thing.

### Message to Students
Stop complaining (you're in California), expect life extension to 100+, and think like a computer scientist—build scalable things for everyone, not just yourself.

## Transcript

[00:03] Eric, welcome to USC. You wrote a sonnet? I like to write. I like to write sonnets. This is about the Marshall School. Pay attention. When Fortune's wheel doth turn in time swift course ... You don't know what I'm about to say. And commerce seeks new methods to prevail. The wise must harness innovation’s force lest ancient ways grow stagnant, cold and stale. Behold, the thinking engine doth arise. A tool to amplify the student's mind. To parse through data vast as are the skies and patterns in the chaos swift to find. Let Marshall's halls embrace this mechanic art, for business leaders must with times align. The future calls. They must not stand apart, but blend the human touch with the divine.

[01:00] So let USC with boldness lead the way and teach tomorrow's mind what rules today. Wow.

[01:12] Very impressive. It took me a really long time to write that. It took me about one sentence. Welcome to the new world of AI. Amazing. Well, let's just dive right in. When people talk about AI, when they think about AI, especially as it pertains and will impact higher education and universities like USC, I think there are two different narratives: that it's going to be an incredible enhancer of learning and research, and there's another narrative that it's going to completely disrupt the entire university model as we know it. Which narrative is correct? Both. I think almost all of these questions, and by the way, I'm really excited to be here. So I'm just ... have fun with you. What you've done since you took over is remarkable.

[02:06] Look at, look at, I mean, I don't need to sing your praises, but they're well known. Universities are going to adapt because universities, the American universities, 18 of the top 20th, 20 universities in the world are in the United States. It’s one of the greatest things about America. They're adaptable. Right? An example is that you have this amazing film school here, and you have a digital tools, a digital, a digital, how to use digital tools, right? Which is the leading program. It's like a little subgroup within a larger super group. That's how your university works. And so you're going to see the same thing with respect to labor allocation, education and so forth. The university will adapt,

[02:48] and the university will not adapt because of the faculty. It'll adapt because the students. Because the innovation occurs essentially from graduate students and postdocs. Right? And the faculty are good enough to allow that to occur. It's always been true in my world that two graduate students and an untenured assistant professor change the world. There's something about that pressure cooker, which USC is very much a part of. That's not going to go away. What will go away is an awful lot of the sort of traditional education and common knowledge, and you'll find ways of having much more self-learning and so forth. And I think that the butts-in-seats kind of model is going to get more ... Most universities are experimenting

[03:29] with various forms of collaborative research, working teams and so forth, because it's a better learning model. I don't think ... So, let me exclude a bunch of things. Teachers are not going away. Universities are not going away. Jobs are not going away. Right? All those predictions are all ill informed. But what you do will ... you will use computers a lot more in everything. And the first class that I hope you introduce after this conference is a class on prompt engineering. How could you have students at USC that do not know how to do complicated prompts to generate whatever it is they're care about, which is often very funny? Do you think there's a greater danger for universities in under-utilizing

[04:11] AI or over-utilizing it? Well, I'm always pro-AI, and the reason is that the arrival of a new form of intelligence that’s at the peer level of humans, which is roughly what's happening, is a very big historical day. This is like, like the arrival of electricity, maybe even more so. We, and I've written two books with Kissinger on this question about what happens when your best friend is non-human, what happens when your collaborator is non-human? What happens when your non-human collaborator is smarter than you are in some areas? The most important thing for this university is to get your students ready for that outcome, because that outcome is a big deal and an awful lot of people who are much older

[05:03] than your students are going to have all sorts of crises over this because it'll affect their identity, what they do, how they spend their time. You share your view that jobs aren't going away, but jobs are certainly going to change, and perhaps very dramatically in ways that we can't really fully envision over the next five to 10 years. How do you think, how do you think universities need to think about how we need to adapt to the changing nature of work, the changing jobs market, in terms of how we're teaching and what we're teaching students? So, and I'll make an obnoxious comment, the most interesting professor is a professor who has tenure, who studies the dynamic nature of the economics of the labor market

[05:48] because they themselves are not subject to it. But I digress. That's not an attack on tenure. Tenure got me here. My father was a tenured faculty, so forth. Let's consider ... So I've spent 20 years with people telling me that truck drivers were going to no longer going to be needed because we would be driving trucks. The number one, there are many job shortages in America. The number one is truck drivers. It's really hard to drive a truck. It's a very difficult job, and not enough people want to do it. How is it possible that jobs are going to take that jobs are all going to go away because of automation and the job, which is most obviously automatable, right, and is difficult and so forth, is the one in the shortest supply.

[06:34] And the answer is because, two things. First, economic efficiency in a, in a country level fundamentally means that the, the firms that adopt AI and automation more quickly become more efficient, which grows the market. They displace people who are less efficient. But as a societal matter, growing the economy and growing revenue, especially in a democracy, is key. Another view on democracy. Europe is essentially not growing. Thank God America is growing at somewhere between 2 to 3%, which is great. In the democratic systems, you can't take money away from its constituents. So if you have new needs, you have to have new money. So democracies have to grow. So you've got to figure out a way to grow your democracy, which is through productivity.

[07:23] And all of the evidence is that technology invention, especially in America, is the key component of American exceptionalism and the, the characteristic of the American culture, which I'll simplify, which everybody knows is relatively freedom focused, relatively suspicious of the government and relatively law abiding. Right? We can debate the specifics, but that's roughly classic liberalism is a necessary component of the kind of growth that we're having. And other cultures, which are also democracies, don't have any growth. They can't do anything. I've tried for 15 years to try to get Europe to do something. Europe is full of incredibly smart people, but the economic system doesn't form it. I was in China, and I mentioned this to you when we just met.

[08:14] I was in China in August, and I came to ... I was originally thinking that China was going to beat ... attempt to beat us. China does not have the depth of the financial markets to fund the crazy stuff that American entrepreneurs are doing. You can't randomly raise $50 billion for your data center in China, right? And maybe that government will ultimately decide that's a good idea, but it's a command and control economy. So the genius of America is this restructuring. So the long ... sorry for the long-winded answer, but the result is, yes, there will be more jobs. The jobs will be different. And furthermore, because demographics are such that we're having fewer and fewer children ... And this is worst, by the way, in Korea

[08:56] and in China and other Asian countries, where the replacement rate is well below 1.0, which is terrible. Those are examples of societies that going to need some serious surgery. I don't exactly know why this is occurring, but it's important. We're not going to have the people to do the jobs we need in the next 30 years, which is why universities are so important. You produce the people. I want to get back to this question about how universities need to be rethinking its curricula and how we teach, but you mentioned democracy a couple of times, and I saw that this morning you published an editorial in The New York Times about how you can't outsource democracy to AI. Can you share why you felt compelled to write that?

[09:37] We're in a period where people are frustrated with democracies. I believe in freedom, and I believe in democracy. My political views are clear. And, in order to have a ... democracy has to sort of organize itself and succeed. When there are troubles, and we're certainly having them now, there is an appeal to authoritarianism, to the quick solution. And I'm very worried that people will say, let's have the AI do it. After all, AI is much smarter than we are. It can figure things out. It can regulate us. That is a loss of human dignity that I'm very, very worried about. Do not give your human dignity to a non-human activity. You want your freedom to think, to write, to associate, and so forth as a human being, and you want to do it collaboratively.

[10:28] So my basic argument is you need to use AI to cause the communications process, which has largely become, at least in America, he said, she said kinds of tribalism. You know, “you're in this” and “you’re left” and “right” and so forth and so on. There's no constructive conversation. There are solutions which involve using AI to try to encourage the dialogue. And somehow we've forgotten that. I wonder if you can extend some of those ideas to higher education, because one of the things that we think a lot about at USC is that you're not just learning knowledge and acquiring skills, but you're also learning how to interact with other people. You're learning how to get things done as a team. And I'm wondering if one of the implications for

[11:14] AI is that universities, of course, need to figure out how to continue teaching the material, but they also perhaps should consider leaning even more into the social aspects of being part of a university community for, for years or however long they're here. To the degree that you start yourself to interfere in students’ behavior. One, you'll be, shall we say, concerned. And second, it's just better to let students do their thing. Right? Because you spend an awful lot of time getting the, really, the quality of the school force here is, as a private university, is exceptional. Let them figure it out. I'm very bottoms up in this view. And I think that ... so, so ...

[12:03] The professors that I've talked to in other universities have a pretty simple rule: phones off in the classroom. You actually have to talk to each other, right? And you actually have to talk to each other without anonymity. And, there's a long discussion about how social media has destroyed society. I think you're familiar with Jonathan Haidt's book because he was just here, and he's a good friend. A guy named Scott Galloway has just published a book, which is quite concerning, about the status of men in their 20s. And all of these are consequences of excessive social media. Right? And the fact of the matter is, most people and an awful lot of teenagers are pretty lonely, right? And so they turn to these things and they take on a huge

[12:46] superhuman role in their development. Thought experiment for you is somebody here has a child next year or the following year, and every year your child gets a better toy. And the toy can talk and is intelligent. So it's a bear or something like that. Well, that toy will become the child's best friend. Now as a, with a three-year-old, you're going to be really excited about it because it'll keep the three-year-old occupied and it'll teach that the three-year-old language and counting and things like that. But what happens when the kid is 12 and has no social skills with the other kids, but really loves the bear and you cannot get the bear with him? I was talking to, to one social psychologist

[13:25] who had a mother talk to him that she had a 15-year-old boy who was ... she was unhappy about and she attempted to take his phone and he beat the crap out of her to the point where she had to go to the hospital. And I said, are we at that stage? And I think, at least in some cases, we are. It's really, really important to maintain human dignity at each of these levels. And here in the university, as well. You've written a number of books and, and you've written a number of articles and papers. And if you're like me, Eric, part of the reason that I write is to figure out what I actually believe and think about a matter. Are you concerned about a future where students are no longer writing? Instead, they're just prompting

[14:17] LLMs to give them answers or produce first drafts, and then maybe prompting them again to, to edit those drafts, but not actually staring at that blank screen or piece of paper and the impact on and the degradation of their critical thinking as a result. So, so I have every belief that the faculty at USC can determine the people who cheated using LLM and the ones who wrote things originally. Right? I'm sure that, that the faculty here can figure that out from a grading perspective. I'm not worried about that. I have long been worried about a reduction in critical thinking because of reduction in reading and, and the ... Instagram and TikTok and so forth. Reading is not the kind of reading I'm ...

[15:00] You're familiar with stage-one and stage-two memory or thinking? The first is quick, and the second is long. And in academia, we call that being in the zone. And so if your students are not in the zone, sitting there doing it, I've been working with a number of research scientists in their 20s, and when I was a research scientist in my 20s, of course, there were no cellphones and we were busy, you know, doing stupid things in college. And we were otherwise reading books and, in my case, writing a lot of code, which I guess is going away as well. And in watching them, they turn off their phone and they just work and are concentrated for a while, and then they take a break and so forth. So maybe the answer to the addiction cycle,

[15:50] because these are all fundamentally addiction devices, is to train your students to turn them off for a while. My trick, my, my newest invention is I put my phone on airplane mode, and I imagine that I'm in an airplane, OK? And I'm so stupid that this works. OK, I'm in an airplane. I look at my phone, it says I'm in an airplane. You know, it works.

[16:17] Dean Garrett talked about an emphasis of mine on really trying to lean into the private sector more and develop relationships not with just private industry, but also with government agencies and actors. I want to ask you, how has AI raised the stakes for us to get that, that set of relationships as, as right as we possibly can? Because right now it's, it's difficult. It's difficult to collaborate with the government and with the private sector. But I wanted to ask, how has AI increased the urgency that we do that, and then the follow-up question is going to be, how do we have to get this right in order to compete with other countries, including China? Well, you guys have a really good math department, OK?

[17:04] And every one of these firms has no idea what AI really is. Right? So I think you're misstating your own negotiating leverage. They need more ... They need you more than you need them. So the AI that I work on is essentially math and most of it ... And I have a piece in this here. I cannot even understand the math anymore. It's moved so fast. And I think that this generation that's coming out right now, which, who will be talking for the rest of the day, are probably the greatest economic potentials of all. And the reason is they're the ones that are going to take these complicated algorithms and essentially apply them to solve real business problems in health care. And there's a long list. In our industry, we,

[17:51] we talk about technology overhang, where we believe, arrogantly, I might add, that we have produced all this incredible technology. And why do customers, businesses, universities and government, why are they taking it more slowly? Now, one answer is it's not quite ready for use, but the real reason is that technology innovation starts in bespoke industries and requires special handling, and then it becomes standardized. I'm old enough to remember that email was considered an option. Now most of you don't even use email. But email was ... And then one day ... So they used to have this diagram where they said, “This is the vertical. This is the horizontal.” And then somebody announced one day the vertical became horizontal, and email became standard.

[18:36] And then all of a sudden the email had to work. So these are simple ways of understanding how technology adoption occurs. And remember, so I was coming by here and I saw Waymo car here in L.A. Great. Now the first self-driving cars were largely invented in 1994. Stanford won the self-driving competition in 2004. Google hired all those people in 2005, and it is now 2025. So it took 20 years from the time a successful contest among universities to this. So the technology diffusion is much ... takes much longer, and the wealth creation occurs at every stage. And since we have a lot of business students here and I'm very pro-business, obviously, remember that the majority of the money is not made at the beginning.

[19:27] It's actually made during the mainstreaming of it, because the mainstreaming creates trillion-dollar industries. And that's where the real wealth will be made. How do you think about China, Eric, in this ... in terms of how universities need to balance, on the one hand, the commitment that is part of our academic mission to sharing knowledge, but also the concerns about intellectual property theft and, and, and we see that those kinds of tensions running through all sorts of debates that we're now having with respect to chips to LLMs and in a lot of other parts of AI? Well, I think, I think you and I agree on this. I'm very strong and positive on foreign students in our American universities, especially at the graduate level.

[20:17] When I was running the National Commission on AI for the Congress, we studied very carefully the Chinese contribution. And it turns out all of the really tippy-top papers in America had one or two out of the five or 10 authors who were Chinese, right? Chinese born, Chinese trained. They are an important part of our research infrastructure. We don't want to get rid of them. Furthermore, we don't get we want to get rid of them after they graduate and send them back to China. That's the stupidest thing ever. So staple a green card on their diploma, you know, whatever ... Whatever it takes, right? You know, you issue diplomas, you have graduation. Just take a green card, you know, and just put it on.

[20:55] Just staple it on. Right? They can write their name in it, right? Yeah. It's obvious, and ... So, why in the world would we want the top, top students in the world to be anywhere other than America? Now, if they're criminals, arrest them. It's not that compli-, we have like a police force and they're pretty good, right? And, you know, you have your own police. What is your problem? You have your own police. You can make your own green cards. You do your own diplomas, like a full-stop shopping. So, so my, my point here is, if you take the right attitude, then America can continue to lead. There are national resource differences. So as I mentioned ... Let me just finish on the China thing. China is going to dominate

[21:39] in a set of areas without question. They dominate, and this was a mistake on America's part, the electronic, electric vehicle and the battery systems. There's no scenario where American firms will be able to compete at the scale they operate. I think, and this is controversial, that that will be true at humanoid robots for the same reasons. I went to some of their factories. They're huge, and these people work incredibly hard. So far, if you look at DeepSeek, which is the most impressive of the research groups, they're good. I mean, they've invented a number of profoundly better algorithms to do what is called reasoning, DeepSeek R1 and so forth. But they don't have the scale. So, at least at the moment, we're OK.

[22:26] But I think as a matter of intellectual property, my preference would be to engage with China because I've, I, and I want to be clear here, China and the U.S. will never be best friends, but we can learn how to work together. And even though the systems are different, both have a common interest in the set of things: domestic stability. We don't want terrorism, nor do they. Economic growth. They say they're communist, but they sure look like capitalists to this capitalist, right? And they work very hard and they're very, very smart. If you have any confusion about the quality of Chinese research, stop. You are wrong. All those prejudices from 15 years ago about they're behind, they’re are no longer true.

[23:10] They are really at the doorstep. You mentioned DeepSeek, Eric, and one of the weaknesses, we'll call it, of DeepSeek, is that when you do query it for information about certain topics, right, you're going to see certain biases. And I was wondering, and there's a whole debate that you're well familiar with about bias and building natural language LLMs and whether it's avoidable or even the right way to talk about the issue. But I want to ask you, are you concerned about a future where people start using the same LLMs and, as a result, you start getting a narrowing of opinion? When universities like ours are always trying to cultivate the broadest range of opinions and viewpoints that we can. So what should you teach?

[24:07] Curiosity. Right? And I'm now at the point where I just can't stand it when people ... and I can tell people are paid to convince me of something, I'm intelligent enough. Just give me the facts, and I can form my own opinion please. And I think there's a real danger in society that people will be captured by an LLM. Remember, it's their best friend. And if your ... your best friend has, I don't know about you all, has a huge impact on, on me. You know, I talked ... in my case, it's a guy. I talk to him, he calls me and I do the thing. You know, he has a big impact on me. He's really important to me. And ... or your spouse, you know, whatever. Whatever it, whatever it is. So part of the answer

[24:52] is that there's not going to be a single LLM solution. So let me give you the way to think about it. The ... there's a, in America, there is a race to AGI, which stands for artificial general intelligence. The leading players there are a startup called Anthropic. OpenAI, which everybody here is familiar with. Gemini, of course, which I, is where I came from, Google. Microsoft is now doing its own. Elon is attempting to enter the space in Grok. Maybe there's a couple of others. These things require an enormous amount of money. Those systems ... and Meta is, not at the moment in the lead. It is probably going to have to have a closed system to do so. Closed means they're not going to give away the source,

[25:38] just because the amount of money involved, they're not going to give it away. The Chinese have strangely made the inverse decision. Every Chinese model is open source. So I have a number of startups which have a high quality of very, very smart Chinese people, and they just use the Chinese LLMs. And they're not political startups. They're, they're just useful and they're free. I said, sure, right? So it's likely that the, in the media, in the next five or 10 years, you're going to see this huge amount of tension about, as we get closer and closer to AGI. And when we hit various thresholds, you know, when something ... We already have systems that can answer graduate-level biology, chemistry, math problems,

[26:22] which I look at them and I say it's inconceivable to me that it can do it and it can solve those. Those are like headline events in my world. AlphaFold, headline events in my world. There will be more, but China will offer those in open source form, and there will be millions of smaller models in open source from, from many countries. So I think you won't have as big a lock in, especially if the university teaches you not to become locked into one. I move around. I try different ones. Do you feel like a lot of the American companies that are developing these LLMs have taken the wrong direction with respect to open source? They didn't really have a choice. I'm, my back-, just to let you know, as a computer scientist, my background was open source.

[27:07] Open source means we literally publish our source code. And I was part of something called Unix, which is the underlying, for example, it underlies the Mac operating system. For those of you who don't know what it is. And it was a great time because we were exchanging ideas, we were working collaboratively. And look what we built. The internet is largely built because of open source architectures and so forth. So there's a lot of power in that. I think realistically, the, so, if you look at ... so for example, Sam, one day called me up and said, you're going to be happy. I'm going to take open, o3 and do a mini-version and release it in open source. And they did. And it's pretty good, right? So I think there's,

[27:53] there will be attempts by the American companies to release subset. One of the core questions for the American companies is the raw models are so big and so powerful, will they ever release the biggest models? It sure looks to me like they're going to release a portfolio of models from super huge, huge, medium range, small, open source, with different pricing points, and it looks like that's what they're going to do. And the reasons have to do with the cost of serving and the cost of the architecture. Eric, you mentioned AlphaFold, and maybe you can explain a little bit about what AlphaFold was able to do. And, and your former company, Google, with Demis Hassabis, Demis Hassabis, has just been doing incredible work

[28:39] in bringing AI to generate scientific breakthroughs. And all of this requires a tremendous amount of compute. What's your advice to universities that want to be at that table? I told you already. Find a Republican state, build a large data center there, and run a big pipe to wherever you are. Sorry. No, no, no. I just want you to say it in front of everyone. Oh, I see. You want me to support you, and you have a Republican state in, in mind? Well, you gave me an idea. Well, I think Nevada is pretty close. Yeah. But, But, but, but you think it’s ... I'm not being ... I'm not being facetious. So, so, So universities need to keep up with this. There is no way the future can be invented in the, in universities

[29:38] without enough computing resource. No university I know has anywhere close. These are the richest universities in the world. I mean, think about the British universities have no money at all. At least, I mean, you guys, you have a lot a lot of money. It's private school. Very well run. And you don't have it. So first, you need to build a big data center. And, and many universities are power constrained. And then what I would do is I would get donors to donate the necessary chips. As you know, Nvidia is 90, 95% of the training market right now. They have amazing chips. And of course, they're now the most valuable company in the world. I'm pleased to report that Google now is working on competitive offerings, and the TPU V7 chip

[30:22] appears to be better than the B-300 chip for most training. Of course, that's widely disputed in the industry. That competition is really important. So the answer to the crazy stuff that's going on in tech is hardcore competition, which will benefit the university. Well, I have your number. So we'll be talking about philanthropic opportunities here at USC. So ... And I ... and I'm happy to help. Just a couple more questions. Wanted to ask you, what do you think one of the most underappreciated opportunities for universities are in the AI space? I think I said it before. The world I live in, I'll just ... I'll just be very direct, is largely dominated by the most brilliant 20-somethings, men and women,

[31:12] because the stuff is so new that if you're 40, you didn't go through all those painful classes to learn it. Right? Now, you're smart at 40 and you know you can study it and so forth. But the people who have forged through the crucible of their undergraduate work. At Princeton, I'm from Princeton, I met with a bunch of sophomores who were doing alternative reinforcement learning algorithms, and I go like, what, as sophomores? I'm sure that's true here. So I think people do not appreciate the, the quality of the ideas and the capability of moving quickly with the right tools here. So let's go through what that could mean. In humanities, why do you not have a database that has taken all the documents that matter and made them

[31:58] accessible and classifiable in interesting new ways? In physics, people are using models called diffusion models, which are hard to explain, but basically you add noise and then you reduce the noise and you discover an ... you discover the underlying pattern. Physics is a lot about understanding the underlying pattern. In biology with AlphaFold ... So proteins, so, the primary things that LLMs, what you know, is open ChatGPT do is they do sequence prediction. So what they're really doing is they're doing next word prediction. They just do it really well. And I often tell the story that it's like on a Thursday night in San Francisco and they were sitting, at OpenAI, and these are a couple of my friends.

[32:46] And they decided to turn on what became GPT-3o. So this is, you know, four years ago. And they said, “Oh my God, this thing can write.” They didn't know. It's a real historical discovery. And there's a, there's a moment where that occurred which will ultimately be recorded in history. And what's interesting is we think, “Oh, that's so stupid. Next word prediction?” By the way, how does your brain work? It's called next word prediction. It doesn't use exactly neural nets the same way. And it has other agencies and so forth. But fundamentally it's closer to the way people think than you're willing to admit, speaking as humans. So what I like about, about all of this is that the expansion and the speed

[33:33] that is possible when you take next word prediction and you add planning, which is essentially a series of steps, you can kind of do anything. Here's an example. So I like to buy houses. I like to build houses. I want to buy a house in California, which is always a mess. So the first agent, an agent is input and output. The input is language. It has memory. It has output, which is language. The first agent tells me where to buy. The second agent consults the, the published rules. The third agent designs the home with a known architect, which it selects. An actual human. The fourth hires a contractor, an actual human, to build the house. And the fifth agent sues the contractor when it doesn't work.

[34:12] Right? That's a workflow. This is ... Welcome to my world. Now, you sit there and you go, why is this so ... I mean, why is he giving us such a stupid example? Because you'll remember it. I just described the workflow in a university. I just described a workflow in a hospital. I just described the workflow in government and in businesses. So the agentic revolution, which is what, how we in the, in the computer industry call it, is the next big thing. So another piece of advice for universities is just start building agents. So if the graduate students in computer science, who in this area will be the smartest, just build an agent orchestration framework to run all the agents and tell everyone else in the university

[34:57] to build agents and have fun. Most of the agents will be kind of useless, but that's OK. It's a university. But people will learn what agents are like. So combination of right agents, which the computer can do for you, and then prompt engineering, which allows you to program anything, you're kind of done, right? Think about the power of your labor force, which is here in the room. Fifty thousand total? Boy, I'd take that. Last question, and maybe you started to answer it just now. So I was wondering, Eric, if you could just speak directly to all the students who are listening today? We see all the headlines, you know, jobs are going away, reorganizations. And I think there's, there's reason to be really, really excited.

[35:45] And there's also reasons for our students who, again, as we discussed, span hard sciences to the performing arts to be really worried about the future and how AI is going to impact it. What's your, what's your message to them? There's this problem with smart people that smart people are always depressed. Stop. Stop thinking so hard. I am super jealous of your undergraduates. I want to be very clear. One, you're in California, stop complaining. OK? Like I moved here as a graduate student for the weather. So we're clear. Second, in your lifetimes, you're going to see life extensions to 100 or more. Your children will be infinitely healthier than my generation ever was. Hopefully, the world will get

[36:35] more safe and less dangerous. So far, that's generally been true. There are plenty of studies on this, but most important, you're going to work on the arrival of a new form of intelligence that you can work with. And yes, yes, yes, there's all the movies about how the intelligence gets out of control. And then, you know, the the lady scientist goes and kills the computer and the man is, you know, whatever. Those are fun movies, and you're good at movies. But the important thing is the arrival of intelligence is such a productivity increase, and productivity, I mean, as a personal level, for entertainment, right, for your social life. Right? So start thinking as a 20-something about how you can participate in the agentic revolution

[37:19] and think like a computer scientist. Don't do something for you. Do something for everyone around you and everyone you know and all of their friends. A good rule is, so why are computer scientists doing so well besides, as I mentioned, our own arrogance. Because we think in terms of scale, and because I'm, I'm well known now, a lot of people come up to me with interesting business ideas, but none of them will scale. Come up with a scalable business model. And why is a scalable business model interesting? Because you can make a fortune, which is why you're in business school. Sorry. You see my point? I would be ... You should be, in fact, this is probably why you work here. You want to hang out with them because you're jealous of them, too?

[38:02] Absolutely. But thank you for having me. This is a great honor to kick this off. Eric Schmidt, thank you for your contributions to our society. Thanks for being here.