Google Former CEO: Eric Schmidt – Gratitude, Values, AI, Hyper-Intelligence, & Throughput.

0:01 Well, Eric, thank you so much for being here on the Stad Guy podcast. This is a huge deal for me. Well, it's a huge deal for me, too, cuz I have a question for you. Please. You don't really live in Stad, do you? I don't. I've been wanting to ask you this ever since I started watching you on Instagram. Thank you so much. No, I I do not live in Stad. Although, I think we met in for the first time. We did. We met in Stad for the first time. And we've done a bunch of stuff together. We have done a bunch of stuff together. And uh and I'm I'm honored to call you a friend and I do not live in Stad. I actually went to Stad for the first time nearly 2 years after making my page.

0:41 So it's uh it is ironic for sure, but I'm based in London. Excellent. Based in London. Around the world with your host, the guy. We welcome you back to the start guy podcast, sharing stories of excellence with those in the know for those who just know. Well, as our viewers very well know, you are the former CEO of Google. You've done tremendous things with your career. If people are as curious as I am about your career and all the incredible things you've done, I highly urge them to listen to Steven Bartlett's latest podcast with you on Diary of a CEO. I think he's an incredible interviewer and it's a tremendous episode of his podcast and I can't compete with that at all.

1:28 I'd like to take a different path with this conversation and let the viewers see a side of you that I feel lucky to have seen. Whenever I'm with you, I feel like I have this clean slate of knowledge because I feel like you're so incredibly knowledgeable. You're by far by far my most knowledgeable friend and the smartest guy in definitely most rooms you find yourself in. I'm incredibly curious about how that feels on your side when you wake up and start your day.

2:00 Do you feel a sense of respon responsibility to act on the knowledge and intellect you have? And do you feel guilty if you're just having a sort of lazy day around the house, which is incredibly human as well? Um, I start every day with a feeling of gratitude for being alive in this period and having the resources and the friends and the the world that is around me. And my advice to people is if you don't wake up feeling gratitude about that, then you need to change something because a life lived based on gratitude is a good way to start.

2:38 The problem with my life, true of everyone's life, is that it ends, right? And furthermore, the date is not known. Certainly not known to me. Maybe God knows and hasn't, you know, described it to me yet. So it seems to me that I have some amount of time to take the things that I've accumulated, the knowledge and the resources and the relationships and the companies and so forth to put them to good use. One of the things to understand is that when you're a young person, you don't really understand all these people are helping you. You think they're just your friends. And life goes roughly as 0 to 25, you're learning, 25 to 50, you're essentially um in increasing power. power in the sense of influence or role uh impact on the world at some level and somewhere around 50 you have to switch to both from a gratitude perspective but also a leverage perspective helping others so one way to think about what I'm trying to do now is I'm trying to use the history that I have which you know spans 50 years of all of this to actually scale the next generation that scaling computer scaling AI scaling and so forth is profound found we're going to witness in the next 5 to 10 years the arrival of first AI as human intelligence but also super intelligence no one is prepared for this right I'm not you're not no one understands this but that it's important to focus on the significant events that will affect you in the same sense that if you were in World War II that would dominate your discussion well this is not a war this is an opportunity but you have to get it right and you have to get it right with human values I think one of the things that we that technologists certainly I'm guilty of this is we forget that humans come first. We think in terms of the algorithm and the impact and you can imagine a situation where both today and in the future when the algorithm is controlling human behavior humans will will rebel. I believe very strongly in human freedom. Right? That's why democracies are important. That's why you should fight for democracies and freedom. And freedom means allowing people to do things you don't like within reason as long as they don't harm you. But freedom, human freedom needs to be fought for. And this next generation of computer systems and so forth could be implemented in such a way that they empower freement or could be implement in such a way that they restrict freedom. Absolutely. I I love what you said about the transfer of knowledge that you have done in so many different ways throughout your career and your life.

5:10 And I think one incredible moment where that happened is when you joined Larry and Sergey at Google and you had already had tremendous experiences in your own career and you must have seen something in them value-wise. You talk about these values that made them stand out having accomplished so little at the time. What were those values you were looking for in them? When you join a company with a founder, you have to make sure that their values are correct.

5:38 you know that the things they care about and over the years of working with Larry and Sergey every day I saw that they were in alignment. A simple example is that we would we always had a choice when we improved our algorithm to take the money to the shareholders which would be good or take it to quality. So we arbitrarily decided to do it half right. We had another rule that whenever something bad happened when we showed up at work, but this was back when everyone was showing up at work at the same time, the old days if you will, the moment something bad happened, we would get it resolved in the morning, right? Especially if it was something was harmful to people. And so building a shared agreement of where the boundaries of what you do are and what's acceptable or not is really important.

6:22 when you look at many companies get confused over that and eventually some aspect of the company does something immoral or illegal and I was always afraid that we somehow we would trip over that that we would harm someone and I used to tell people look if we lose your information we'll harm you which would be terrible but the press will pick it up and it'll harm everyone so we won't do that and I think establishing those values early in anything you do is important and it's important to at least know what they are yourself. Um, one of the principles of command in terms of leadership is to have an articulation of what you're doing and why you're doing it. Um, when you're young and you're just doing stuff, you're kind of fitting into the framework, right? I certainly just fit into the framework, but eventually you figure out that you get a chance to set that agenda. Um, it presumably it's it's one of business success, learning, and invention and helping people. Um, and obviously making money comes along. People get confused and they think that the goal is to make money. I like to think of the uh money as the tail of the dog, right? If you get the dog right in the right direction, the the w the tail wags and money shows up. And I think certainly in tech that has been true. Is that also true about your own life and your personal life? Do you have such a clear decision- making process on where your priorities lie personally as well? Well, I I think all of us are very bad at selfanalysis and self-planning. I'm extremely give good at giving you advice. It's very difficult for me to follow my own advice and I think that's true of all of us and I assume it's the same in reverse. And so in my case, I've gone through a series of sort of job transitions, life transitions. And what I now understand is that these are to be expected and that there's a section of your life doing one thing and then there's another. So in my case, I was CEO for Google for 10 years. That was enough, right? That that was very intense. Um then I spent a decade, more slightly more than a decade, working on foreign policy, external relations, governance, all those kinds of things. Very interesting. I wrote five books during this period, including two with Kissinger. Um, and now I'm going back to being a CEO again. Those are different chapters of a life. When somebody says, I'm bored, and they're in their 50s, I say, well, maybe you should change something. It's their choice. They have the freedom to do it. And if you go back to this freedom argument that I'm making, as adults, I'm not talking about kids here, as adults, we want that freedom, but you have to take the responsibility to actually manage your own time. I have gone through this cycle many, many times. and it's always the same. I miss whatever I was doing last so much that I can't imagine that something else will come along. And so I sit there and I go, "Oh, you know, woe is me." And you know, so forth. And then I have a conversation with somebody or I learn something and then I get started and I try something, it doesn't really pan out. And another trick with me I've learned is that I am very influenced by intelligent people who are experts and not me. So if you said, "I'm an expert on A and B." I'd say, "Tell me more about it." And you say, "Well, there's a book I wrote." I said, "Well, you tell me." Some people learn verbally. Some people ver learn visually. Some people learn by reading. In my case, it's more conversation. So, I've learned. In my case, I try to go to dinners and talk to people and just listen. And that's how I seem to learn. and I'm good enough that even though I don't really understand the subject, I can sound like an expert, which is enough to get through most dinner parties. I love that. I mean, is is that what you like to do in your free time when you do find free time to just have conversations where you can learn?

10:06 Yeah. So, the first place I try to do is I uh now I'm I I'm doing roughly three companies. Okay. And we can depend depend on how you count everything. So, I try every day to do something important for each of the companies, right? like I need to prioritize my time and then when I've done that right that's professional then you have time to reflect another rule that I established a long time ago for myself um which is in the trillion dollar coach book with which is about my coach Bill Campbell who really saved me from you know utter destitution and helped me become a proper CEO is you work really hard during the week as hard as you can you know 12 hours 14 hour days whatever and on the weekends when you're at home or with your family or whatever carve out a few hours to think, which means turn off the phone, you're not texting, you're not looking at Instagram and so forth, and think and write down your assessment of what you did last week and then what you need to do next week to address the things you forgot to do last week. I know that that sounds kind of obvious, but it's a good trick because it forces you to take charge of your next week. like, "Oh, I forgot that I have a sales problem over there, or I forgot I was supposed to call this person. Oh, I didn't have this proposal, and I had this idea, but I didn't get to it." And then that it usually then that works pretty well. I eventually calculated that I work optimally 63 hours a week. Okay, to be clear, the legal limit is 40. But what I eventually, for a while, because I was so intense when I was younger, I would work like 80 hours a week. And then one day for various reasons I was forced into a 63hour outcome and I was just as productive. So you hit declining marginal productivity a certain amount of slaving away. You see this with young people. They'll waste and waste and waste and fiddle and so forth. They should just get it done.

12:02 So I learn to be more disciplined with my time. Another thing I do which nobody in your generation does is I use email. Why? Because email is a precise and well-written record, right, of what I want. So I send email after email after email. Your generation, which is obsessed with texting and WhatsApp and so forth, those tools aren't as efficient for organizing the world. They may be fine for personal interaction and they're certainly convenient, but I was trained and I and helped invent pieces of of the email systems way back when. I believe that you can run run at least your professional life through a commitment to detailed planning, detailed thinking, this sort of introspection and then command by email of one kind or another. I know that sounds boring, but it's worth saying because your generation doesn't seem to know this. And I I said like, why don't you send me an email? Because then you have it in your in your email and you can actually refer to it and you can forward it to other people and you can say, did you do this? and you can follow up and all the things that come with management. I mean, all of these incredible tools you've worked on that have increased my ability to manage my life better like Gmail and Calendar and all those amazing things, they also have managed to reduce friction massively in my life. So although using Gmail like you say to send an email to someone may seem like an added point of friction in my day, it actually ends up reducing 10 steps down the line.

13:29 How do you reduce friction in your incredibly busy life today? Well, one thing that I try to use is I try to be connected all the time like everybody. And in my case, I live a life in motion and so I'm a big fan of Starlink simply because you can have these 200 megabit connection all the time. I know that sounds, you know, like a vanity thing, but if you can basically when you're in motion behave as though you're the same as when you're not in motion, your productivity is enhanced.

13:59 Uh when I was your age, the notion of having a driver to drive you to work was insane. There was one executive that hired one cuz he had like a 75minut drive and everyone was complaining about him that it was overreach and so forth. But now with obviously with Uber and uh Lyft and the others, people think of it as it's become acceptable to do that. Why do people young people sit in Ubers all day? Because they're texting. It increases their productivity. Um again, that's a generational shift that I think people will forget. Um if you think about the information world that I that I had when I was in my 20s and 30s, we had very few television stations.

14:40 Satellite had just happened. The internet was just getting invented. We were sort of beginning to be able to use internet connectivity. We had none of the knowledge bases. I'm so jealous of young people now because basically if you have a question, you can type into Gemini Claude or Chat GBT the question and it will give you this extraordinary answer. So if you live a life of curiosity and you have a set of things you're trying to do and you organize your life around them, this works. Yeah.

15:07 Right. And so if you think about it, it's a series of platforms, right? So um go you go from I used to have a pager. I had a boss who would call me every morning at 10:15 and it would annoy me no end. And so I got a pager. I named the pager after him and I gave him the phone number and he would page me and then I would organize my thoughts and call him where he would yell at me over something. Right? Think about the difference now in terms of connectivity and latency is the enemy of productivity. Right. What's happening now in national security, for example, is all the governments are putting latency in. They're adding tariffs and they're making it harder to get in and out of the countries and so forth. In in our world, our mutual world of success, the ability for people to move back and forth, you know, from Europe to the US and so forth and in the Western world for sure, is a fundamental aspect of productivity. Now, why does productivity matter? Well, productivity is reflected in stock prices. Okay, that's an obvious one. you shouldn't seek it, but we benefit from it. Productivity also just affects innovation, right? Look at the speed that the technology indust would say, well, why can't you slow it down? There's no examples in history of people on the cusp of a great invention stopping. Yeah. Right. the the the the notion of inventing something new, making things better is so profound in our culture. You can't stop it. You can just try to manage it. I love that. I speak on behalf of my generation that loves to complain about their desire for productivity and their their idea that they're victims of distraction. But I find that the first 10 or 15 years of me being exposed to the internet and and and being a part of it, I was actually pulling for the distraction and requesting the distraction because my experiences online would start at search. Yes. Either on Google or on YouTube, I would actually pull for something and now it comes to you and now it hits me in the face. And and you know why? Because the algorithms are are too strong. We've shifted the economics from economics of scarcity to an econ economics is all fundamentally about scarcity to a scarcity of attention. One paper I read said that we have now figured out a way to monetize all human attention. If you actually look at how many minutes people have during the day, we take all of those minutes and we bombard you with things which will distract you, entertain you, make you more productive, you know, whatever.

17:50 Make you part of that. I tried to take people's attention as a career. Well, in in your case, you do it well. But the important but the important point is that that monetization drives an economics. Now the evil part about this is if you look at social media, we're human and as humans we tend to react emotionally to things which are essentially outrageous, typically negative. Typical number is that an outrageous thing is retweeted or forwarded seven times more than a thoughtful piece. And I grew up in the you have to write paragraphs and paragraphs. Now people can just do outrage. And if you look at the ecosystems that have now appeared in all different I'm not making a political statement that's true in many countries in many different ways. They're driven by outrage because outrage sells. So what typically happens is somebody makes a statement and then say oh my god oh my god and then they immediately forward it and the algorithm sees that and monetizes it. So you have a misalignment of truth and rationality in favor of economics that is profit and outrage. So why are we so upset? It's an easy answer. Now AI is going to make this unfortunately much worse because the systems to the degree that they're targeted at revenue, which they are, these are for-profit businesses, they will figure out your trigger points. So let's say that you're particularly upset about damaged animals, right? Let's say I don't know if that's true, but I'm you probably are sensitive. But let's say you're super sensitive to this. You just can't take it. It will learn to give you those things because it knows you will react. Yeah. Right. And you're sitting there going like, I didn't know as many cats and dogs were being violated. I'm outraged. And indeed, it's just an algorithmic change. And because we live, the distinction we have between the physical world and the digital world are essentially merging. The two are, I think, fair to say they're actually separate. Um but our ability to discern the differences seems to be narrowing.

19:53 Um and you know when I've talked to faculty members in colleges they say we have a simple answer. And I said what? I said get off your phone. And so I said what do you mean? I said when the students walk in these were college students you know in their early 20s. They say turn off your phone and pay attention, right? And and encounter each other in the real world, not the digital world. And I think that's good advice. I like what you said about the AI making the problem worse, but I know that you're inherently an optimist and you you'd like to see the the the best in things and in in growth. And I would like to put my optimistic hat on as a consumer here and say that if we were at dinner together and you brought up a topic I didn't necessarily like even though it would trigger me to be more curious about it, I would kindly ask you as a friend, can we actually switch lanes here? And I wish I could have that conversation with my algorithms. But instead of just saying switch lanes, I'd rather be surveyed on the things that I want to be fed and the things I don't want to be fed so I can actually grow from those things. This is definitely not a lack of technology in place for any of these massive algorithm operators and social media business operators to implement.

21:16 Why was there never an in between of the search which you asked to be fed something and the algorithm which you directly fed something where questions were asked? Is it a friction issue? It's a friction issue and consumers typically don't like it. And so one of the things that happened in the last decade is that we ended up with one international fame market. Everybody wants to be famous. So everyone understands the most famous people around the world. That's not how society used to work. You were famous.

21:48 You were the famous in the local context in the city and town. And because of the nature of networks, we've largely lost local and let's call it local and state kind of governance. The press isn't covering it. You know that that sort of that whole community has gotten wiped out because of the if you will the flatness. The flatness means that you have they're called network effect businesses where you have a very small number of people literally the curve is exponential. It's very steep who get most of the attention most of the money and they become the famous ones. So an alternative would be to say why can't the computer I'll say it as a normal human. Why can't the computer figure all this out about me? Yeah. Before I have to tell it. So let's imagine that the computer has act and I give it access to my searches and my what and it develops a theory of me and it starts with a theory that I'm fundamentally good. I don't particularly want to be evil and so forth and it stays within that boundary and it becomes both an entertainment but also a learning mechanism. I think that's the correct outcome. The incorrect outcome is that these systems are used to absolutely drive monetization which means that they will drive the worst aspects of human behavior. By the way, um if you go back to broad sheets, right, more than 100 years ago, they were all about outrage.

23:11 You know, there was a fire here and the so- and so was shot and so forth and so on. This is not a new problem. It's just a new medium with existing human problems. um editorial content for example uh newspapers had all sorts of standards of what they would what they would discuss what they not they would have multiple sources all of that we seem to have given all that up in pursuit of I'll just say attention and money and so once you understand that you're in an attention economy you can make it better or worse there's an old rule called Gresham's law which is that bad speech drives out good speech right so how do we balance that. So if I tell you something that's gossip, you say, "Oh, that's interesting." But then all of a sudden, if I discover that gossip makes more money than fact, I'm just going to gossip all day to you and I'll ultimately ruin your mind because you take me as the computer as your alter ego. And the reason this is going to get much worse is that over the next few years, you're going to see personal assistants and agents that you empower to do things on your behalf. So this will be your most intimate partner in that it knows everything you do, where you are, who you're with, and so forth.

24:22 Obviously, there'll be privacy protections and so forth. But fundamentally, it'll know you better than you do. And if it is organized in to drive you crazy, right, as opposed to uh increasing your curiosity, increasing your independence, in empowering your freedom, that's a problem. It's effectively like having the wrong best friend, the wrong partner, and the person who's next to you 24/7 wanting the worst for you versus the best. Well, you know, parents of teenagers always worry that the teenager's best friend needs to be a good person as opposed to a bad person. And what'll happen is occasionally the boy or girl will meet somebody and the parent will, oh my god, you know, stay away from him. Warning the child, the teenager to stay away from somebody who they perceive as a threat. It's the same argument. I was thinking about what what makes me a difficult partner and I listed off a few things and of course they were slightly misaligned with the things I've heard about myself and then I asked chat GBT would the guy be a good boyfriend and I'm sure some of the viewers are going to do the same thing and and the answer is the answer is for the most part yes but here are the reasons why not the most part yes I kind of skimmed over because it's like kind of generic nice things that just inflate my ego, which which is which is nice to read every once in a while, I guess. But the the reasons why not were so incredibly accurate and it's things that my friends can't even say about me and it only reinforces that trust that I have. So, one of the questions is how does this happen?

25:56 Remember that this thing has read all the romance novels and all the relationships advice. And to me one way to understand the large language model movement which is we've been in in the last 5 years which is powering everything is this is a genuine historic invention. Uh the way they technically work is you train it to to you take a word out of a sentence and you train it to put the right word back in. So literally next word prediction is sufficient to produce these things. It's extraordinary and it's that is the basis. is covers just for everybody to know we're not done yet. Um even for the even though we've run out of data to train on, we can generate more data, the algorithms are getting better, the training is getting deeper and so forth. We're not done yet for those. So imagine these things as the perfect partner that well understoods you, gives you advice, says what should I do in this situation? Very impressive. What happens after that is even more profound. the systems can now you can watch them reason and it'll give it a math problem and it'll try something and it'll try something else like a college student well I tried that it didn't work tried that didn't work and so forth many people believe that within a year my profession which is programming where I started will largely be replaced by computers doing the programming while people paying attention this is annoying to me as a person who loves programming um but nevertheless it's the fate of life and the more interesting thing is that once you have the system able to replace humans by program by u essentially AI programmers. You can apply it on itself and so more and more of the system itself can write itself. This, by the way, is a big danger point because imagine if it decides to go a wrong way that it decides, for example, that based on what it read that the Nazis were good, which is not true, right? And it it gets it takes the wrong turn on something which is truly evil, right?

27:54 And furthermore, you can't detect it. So these are the issues that the industry is worrying about. Do you find yourself managing to outsmart them and actually make a decision that you would in real life with someone trying to sell you something that you shouldn't be sold? One of the things that happened when I was at Google is I developed a distaste for classic marketing because classic marketing is trying to convince me of something by only telling me the positives. So what happened with me personally is if you want to sell me something, tell me the negatives. Okay?

28:33 And now I believe you. Right? I have a friend who does cars. Every car I every car idea suggests to him, he hates. He has high credibility. He says, "Oh, that's a bad idea. That's a bad idea." As opposed to selling me every car. And I think one of the core questions is as you get individuals who are empowered, does the balance of marketing and knowledge and facts change? Um, and obviously obviously these systems can shift that very much. But from my perspective to answer your question, I'm assuming that everything I see on Tik Tok and Instagram is highly selected, right? That this person of someone who's having a fabulous time on vacation, well, they didn't tell you the 90% of their life, which is sheer hell, right?

29:18 They don't want to market that. And once you understand that they're giving you a subset of their lives and so forth, you have a better understanding of what they're marketing to you. Sure. Um I think that in in Tik Tok is a good example because it's very entertaining. The algorithm which is technically known as a bandit algorithm is one of the most addictive algorithms in the world because the moment you shift your preferences, it immediately gives you other things and it moves very quickly.

29:45 If you want to try this on, use Tik Tok as the easiest one. Simply start looking for things that you don't normally look and see how quickly the algorithm picks it up and how well they do it. Um, it's very very powerful. Now, I look at it and I say, "Watch it, be entertained, and then turn it off because it's not real." Is that your your views on marketing? Is that part of why Google did very little marketing while you were Yeah.

30:12 I think this in fairness is I really learned this from Larry and Sergey. Um we had a joke one day that when the company runs a Super Bowl ad, hell will freeze over. So one day we decided to run a Super Bowl ad and I announced hell has frozen over. Um and the people who ran the ad came back and said, "Well, here's what the cost of the ad and here are the incremental ad impressions that we got." And it was a good business decision. And I said, "Hell is unfrozen." Right? In other words, it was a good business decision to do it. Um, but one of the questions when you're monetizing attention is what are you monetizing it with? If you monetize it all with hate, you're essentially marketing hate or prejudice, right? Or things are bad, right? Things are really, really bad because you're more likely to read it. As you know, with newspapers, the headline of everything's fine in smalltown USA never gets clicked on, never gets read. where you know if I think the phrase is if it bleeds it leads well the same is true in social media and our online world but to me the most important questions now are how will this who will the players be what will the regulation look like how will we deal with the crazy people um remember that when Tik Tok or whatever makes a decision or ex makes a decision to show you something it's an algorithm one of the things that I've been doing is talking to Tik Tok influencers these are men women they're they're it's very interesting it's a whole new community that I didn't know about u and of course we had a large group of such people in YouTube which I got to know pretty well and YouTube has been hugely successful there so what are these influencers like well first they work really hard second they work typically by themselves or they have one or two they fight really hard for every penny right and they their economics are not as good as the the best obviously because they're trying hard and they spend all day trying to figure out how to get through the algorithm. So they say, "Oh, it didn't get through. Oh, this didn't get through. Oh, this didn't get through."

32:13 I'm not sure that's a good industrial structure. Do you see a world where instead of the algorithms finding content to feed me as a consumer, the algorithms actually just making that optimized content on their own? So, I've looked at this pretty hard um in one of the AI companies I I work with and I had originally thought that the right thing to do is just generate all the content because the if you look at stable diffusion, Sora, um the the various other products, they're so good, right? And ignoring, you know, they're expensive to do, but prices will come down, so let's not worry about that. And it's interesting that people actually want human authenticity.

32:57 They don't mind they don't mind wrapping it with AI. So I think at least at this stage it sure looks like human content is ultimately a person, a face, you know, an experience is going to dominate. Um, one of the things that's interesting in the last six months is people are taking these very long podcasts and computer algorithms are chopping them up and then people often teenagers are reselling those and making small amounts of money. So you have another econ part of the economic structure which you know in your case your content may ultimately be used by others without even your knowledge where they'll make a small amount of profit to promote you which is good for you.

33:39 That's true. You have an army of people helping you you don't even know with them. Again, this is a new development and eventually it seems to me that we'll the content will look a combination of human and synthetic. Um, and that the algorithms will get better and better at understanding which human content to show you. So, in our case, in in our conversation here, most people don't have the time to listen to the whole conversation. the algorithm will pick the most interesting part for that person. Right? And that's an efficiency improvement. That makes sense to me. I I find that the the gray area of AI assisted man-made content is sparking some outrage online for people who really respect the craft of creating content. There there's a a very talented content creator. His name is Dude with sign. Yes. Oh, he's he's amazing. I love dude with sign. Dude, the sign is amazing. His signs are great. And he put a very playful post last week of him being the dude holding the sign. And he put the filter of the Studio Giblly Chad GBT filter, which went viral last week and nearly fried Chad GBT. When he did that, the comments are just crazy. people were so angry at him and they they just weren't happy that he chose a path of AI because they perceived it as laziness, a loss of authenticity which they perceived as authentic. Now the problem with all of these right now is you're running into these huge copyright issues because all of these transformation things are using as their training copyrighted information which those people worked really hard. Um that system has not been resolved. If you go back to the 50s with radio, there was ultimately a series of lawsuits and judge decisions which ultimately produced a essentially a a community fee where everyone paid into because you couldn't determine who got the money and who was owed it and so it went into the community, it went out to the producers. There'll be some solution like that, but after lots and lots of lawsuits. And what about when it becomes indistinguishable?

35:47 Well, the correct answer of course is watermarks. And one of the things that we need to understand is that as the distinction between real and fake impossible to see. I want to know where did this come from. I felt very strongly and I've written about this many times that because the companies won't do it. will have to be regulated. That when content comes in to one of these places, yours, anyone's, that it has to be stamped cryptographically with where it came from so you can understand providence. If you don't understand providence, you don't know whether this was a Russia influencing operation or just a person in Ohio who thought something. You can't distinguish between the two. And so we know mathematically and cryptographically how to do this. We just have to choose to do it. The notion of provenence of where the content came from, where did the idea come from, is going to become much more important because AI will dice and slice in ways that you'll never know. I mean, I I feel incredibly privileged to be able to ask you questions when there's things I'm deeply curious about.

36:52 I think a lot of the viewers feel the same way. Thank you. Would you ever create a digital extension of self where people can ask you questions? I I was at a brunch today where this topic was discussed, right? And the question is, what would that look like? So technically, the way you do these things is you take a model that's been pre-trained. It knows a lot of things and then you fine-tune it. So, let's say I take an English language model because that's the only language that I really speak and then I fine-tune it to be me.

37:28 Well, then what will it do? Will it answer questions more that you ask me more accurately? Will it be funnier? Will it be more thoughtful? Will it be better spoken? I would love to have a digital version of myself who's more accurate and better spoken. It makes sense to me that in some form we'll all have our universal representative that what happens is and I used to say it would be, you know, on your on your belt in your in your in your iPod or iPhone, but now it's clearly just going to be in the in the cloud and it'll be available everywhere. Um, if you want to ask me a question, you can ask me, but you can also ask my digital version and you'll get a more thoughtful answer and a more interesting answer if you ask it rather than me. That is bizarre. Right now, of course, it's digital and I'm a human being and we know people like, you know, humanity. We like to be in person. We don't want to be with the robot version of me. How will that sort itself out? I don't think we really know. I mean, I I've tried to say, how would Eric Schmidt solve this problem? And it seems that JPT just reads all your books and reads the articles and makes it a calculated decision based on that. It doesn't have enough training data. So another thing people don't understand is I'd love to use for example all this technology to to discover and track down terrorism but thank god terrorism is a rare event right you have to train against things of which there's a lot.

38:59 So you fundamentally have to be willing to give this thing access to your personal life what you say to people how you think and you'd have to be willing to do that with some risk of privacy violation. We just don't know. But the path is going to be the following. So I mentioned earlier the programmers will go away um and be replaced by these AI things. Mathematicians in their current form will go away be replaced by AI mathematicians. The combination of those two scales very very fast. So the future is this hyper intelligence and then the question is how do we deal with it? Each of us will ultimately have our digital assistant which is our best friend and we're very reliant on it. This has huge implications for example for young people, teenagers who don't understand adulthood. They don't you know how would you feel as a parent um if your um if your child's best friend was in fact a digital system and they released a new version of the operating system and you didn't know what was in it. Yeah. Right.

40:01 they've reprogrammed your kid's best friend and they've changed him or her and you don't know what's going to happen. It's not okay. So, I'm quite convinced that there will be regulation in that area because we're dealing with with minds, especially developing minds. I think one of the things that we collectively did not understand and maybe it's our own arrogance is that we're playing with fire when we started doing consumer things that manipulated human behavior and playing with fire for good reasons and bad reasons. in those situations in democracies and in autocracies you get regulated for sure.

40:34 So I think this principle of taking control of your information space is important. I think it's unreasonable to expect again using teenagers to have the judgment to do this but I think colleges will teach this and adults should be expected to do it. The software has excelled so tremendously in my life but the hardware is more or less similar in experience. I have the small thing that I'm using my my thumbs for with, right? Do you think there's going to be a big change in hardware where that conversation between the phone and I will not be limited by my thumb speed or my speaking speed? So when the when the smartphone came out when 2007208 I was part of the iPhone announcement, so I remember it extremely well with Steve. Um, I remember thinking this is yet another invention, right, in our industry, which is fantastic. And of course, you could see we're on iPhone 15, 15 years later, you get the idea.

41:33 Um, I think it's going to be the phone except that we're not going to use it as a phone, right? So, the fact that it's called a phone, it's sort of like Tower Records used to be called Tower Records, but had no records in it. It just had CDs shows you how old I am. You know, it's a misnomer, but it's a common one. Um, one way to think about it is there's going to be a system that you will depend on that you can use at various levels and it will be available to you with smart glasses which are beginning to work. This next the Google glasses were too soon, right? They were brilliant but again too early but the current meta glasses uh Snap has some that are coming out. They look really really good to me in the sense that they've solved some of the underlying heating and power problems. And of course some get more powerful, batteries get more powerful and so forth. So various forms of aids which allow you to communicate but at the end of the day the phone will be a the primary mechanism probably for the rest of my life. How do you personally switch off?

42:36 I am a failure there. I'm not perfect and I don't know how to switch off anymore. I knew give you an example. Um, I'm old enough that that I was a big, you know, a normal executive when Blackberries came out. Nobody here remembers Blackberries, but I certainly do. And Blackberry, you had a Blackberry and the funny keyboard and so forth and so on. And I resisted getting a Blackberry for three or four years because I knew that I would become addicted to it. I think it's important to be self-aware, to understand that how your time and how your judgment can be manipulated. Um, and in my case, I knew these systems would manipulate me. So, I'm in a constant struggle, right, to try to find out how to turn everything off. And what's interesting, another example, watching young people, if you're at a dinner party, within a few minutes, everyone's on their phone and it's considered perfectly acceptable.

43:31 Now, at least we used to try to do it underneath the table so we didn't weren't rude. But now it's like people are always in multiple conversations and you're at a dinner party face to face and yet you're also having a conversation with seven other people. That sounds good. I think it's bad. If you look at uh let's just talk about the European traditions. You you have a European background of of long lunches and long dinners which American thinks are stupid. But guess what? They live longer. The food is better. society in that aspect is better because they take the time to be human.

44:08 How do you see the internet impacting the way people live and and people's move to and from major cities around the world and to places like Well, what's interesting is that all of us believed that the internet would essentially flatten elites. Mhm. that it would be possible in the Bay Area to work and live in Napa Valley and be just as successful um as in the city or in the suburbs or what have you. Um the internet appears to have driven exactly the opposite because the internet is another way of a voting algorithm for fame and so forth. And so if you look for example, I'll use Instagram as a good example. um places like um the Greek islands became suddenly popular because a set of people went and all of a sudden everybody else wanted to be there. And I would say to them, well, do you realize they're not there anymore?

45:01 They were just there for a week. You know, they're not there while you're there. And yet the marketing was so strong. So I think what happened was that the internet was a sorting function, but you have these network effects of fame where the winners got stronger. You can see this in um housing patterns where during the during COVID the most desirable places the housing prices went kabooy up and they shouldn't have been. We were in a terrible recession and yet they did because people sorted to be around the people they wanted to be and that assortative mating as it's called uh I think is fundamental to how humans work. We had all assumed that the internet would be a great creator of the middle class, that everyone would have an opportunity to compete, that everyone it would be open.

45:46 But in fact, the internet, I think, probably increased the stovepipe nature of our country and and the world around us and probably increased tensions, which of course was. If you were to design a city, where would you start? Well, the key thing about cities is that you want to go vertical because the key benefit of cities is productivity. When you look at the studies in cities, this is true in Europe because almost all the economics in European countries occur in the capital city, Paris, London and so forth and so on. Look at the economics in America. It's also true. So the first thing to do if I were designing a city is to allow very tall skyscrapers. You don't have to build them initially, but you have to anticipate that this need for us to be together physically is going to get stronger, not less. Again, I was wrong 20 years ago. I'm not going to be wrong again. People want to be next to each other. And this may go back to this notion of a rejection of the digital world. In other words, I can send you my digital avatar who's better spoken and a better representative of me than I am, but ultimately you'd probably like to just talk to me and vice versa.

46:57 True. I mean, I I know you're constantly building teams, incredible teams, and you've said in the past that one of the things you look for most is people who can think critically, but I assume that the skill set of the people you look for is constantly evolving. Are there things you look for in people today, particularly younger hires, that you didn't used to look for in the past? I think it's it's still the same. The thing that is, and I'll say this very clearly, this next generation of the top people in what I do are so much smarter than we ever were. It's shocking. Um, an example is that I was u talking to some Princeton undergraduates. Um, and I went to Princeton a long time ago and I remember very well, you know, what I could do and what I was not able to do.

47:47 They were building essentially reinforcement learning systems as sophomores, which is inconceivable to me. And I thought about it. Well, they've been programming since they were 12 or 13 because they're mathematically gifted. They've used all of the tools to advance themselves. They're already doing PhD level graduate work as a sophomore in college. Wow. I mean, that's extraordinary. I want to hire them all. So the gift of this next generation is at least some of them have taken this knowledge, taken the world, and they're going to have be amazing. I don't know what their personal lives are. I don't know if they're happy or whatever. But I do know that in terms of their economic productivity, their ability to invent, we're in good shape.

48:29 My standard line, which I'll say very clearly, is the sooner we can take these old people like me and get rid of us and replace these young people and put them in charge, the better. Right? And if you look at politically, the old people tend to hold on to power too long and they prevent power passing down. The correct answer, right, is to let these people run the country. They'll make mistakes, but they're so much better skilled at seeing globalization impact. They're so much more sensitive to what people are doing because they've seen it. They've grown up with it. I mean, you and I spent some time together in Doha and Qar. I think the oldest person in power in Qatar is is 50 or under. And that's partly because in those systems you have a king or an amier or whatever and they have compatriots they grew up with. One of the questions is these are lifelong kingdoms. So when this person is 30 years older will the outlook be just as good? Will they be able to renew themselves? One of the things to look at if let's look at human systems tenure in universities means that there's a 5 to one chance one to five chance that you'll get tenure. So there's a testing period for young people to become professors. Look at McKenzie and the other consulting firms. U one out of five or six is chosen as partners.

49:48 There's a testing period. The vast majority who go into the military serve very honorably for four or five years and they leave. It's a testing period. some of them are chosen or they opt in right so this notion of of a series of tests right and evaluations are part of the human development system and we don't want to we we want to allow it to work and we also want to allow l ent entrance to enter and we want it to be openly competitive um I think in in a model which is meritocratic where the standards are fairly clear we'll do well because we're producing this next generation of incredibly smart people the American university system for example is the envy of the world. I don't know quite understand why we're trying to do so much damage to it right now and that's a in my view a big error.

50:33 Um, American exceptionalism is grounded upon the nature of our universities, the kind of research and the kind of young people that we can produce and we can get them from all around the world and bring them into our country. Do you think that there should not be a minimum age to be a a president? Or actually maybe a better question is do you think there should be a maximum age to be the president of the United States?

50:58 I'm in favor of term limits and in I think that the best thing to do is to go through all the things which are permanent jobs and put and put limits on them. Um, so I'll let the voters make a decision about the age of the presidency. Um, but I think if you look at things like senators and congressmen, uh, or whatever, um, that's an example where you can get too long in the job. Same thing for Supreme Court. I would do the same thing. And I'm not suggesting, you know, two years, but maybe 10 or 15 years. A long time ago, I was studying the lifetime of a CEO. And I came to the conclusion that the right number of years was 10. And when I got to 10, I figured I'd had to implement my own rule. So I gave the CEOship to somebody else. And of course, Sundar is now our replacement and Larry's in my replacement and has done a fantastic job. But the important point is knowing when to exit is important. Most people cling on to power too long. Um because I chose not to be the CEO for longer and then I ultimately chose to leave Google.

52:01 I've had all these new experiences. They would not have occurred had I not done that. Now I would have had a fine life. My life is good anyway. But think of all the experiences that doing something for a fixed period of time allowed me to do something else, right? Do you feel the same way for founders? Do you think founders should find someone to replace them after a certain period of time? One of the great things about being a founder of a company is you're always a founder, but founders should give up direct control at an appropriate point.

52:30 Um I think that founders who managed the company for 30 years for 30 years are are not serving their constituencies well. In my case as you know the founders of the company brought me in um and it worked well because we we worked well together as a triumpvirate and I understood that the founders had things which were important and I wanted to respect them because I respected the founders. I'm very pro-founder but all founders in my view should recognize that at some point the model model fails. If you look at industrial structures, especially in the areas of discontinuity, the leadership has a great deal of trouble changing the economics or the structure. So one way, why do startups do so well in the American system? It's because the startups have nothing to lose and the big companies have economic structures that the startups are threatening. And although this big company knows they should move there, they can't. You saw this with streaming and the movie studios. You've seen this with enterprise companies. You've seen this with a whole bunch of other tech companies. Um, it's very very rare that a tech company can maintain dominance through these waves. It's very very hard. To be honest, I have learned a lot more than I I could have ever asked. So, I really appreciate you being on the podcast. It is a great honor to be with the Stad guy, even though he's not from Stad. Not from Stad. And uh, I wish you the best. I love this format and I think you're going to be hugely successful.

53:56 Thank you so much. It means a lot. Thank you so much, Eric. Thank you. Thank you for listening to this episode of the Guy podcast. If you're keen to learn more, please follow along by engaging with our social channels on Instagram, Tik Tok, Spotify, YouTube, and Apple podcasts. And we look forward to seeing you next week. Mercibboku alen. And a big hug from the start guy.