3 Takeaways - Former Google CEO, Eric Schmidt: On AI, Tech, COVID-19, and Making the World A Better Place (#7)

Episode Date: September 15, 2020

Find out how former Google CEO Eric Schmidt sees the future of artificial intelligence (AI) and tech. As leader of NY Governor Cuomo's COVID-19 task force, he also provides insights on the world post-COVID, as well as the opportunities and talent that he is investing in through Schmidt Futures and his new podcast Reimagine.

Welcome to the Three Takeaways podcast, which features short, memorable conversations with the world's best thinkers, business leaders, writers, politicians, scientists, and other newsmakers. Each episode ends with the three key takeaways that person has learned over their lives and their careers. And now your host and board member of schools at Harvard, Princeton, and Columbia, Lynn Thoman. Hi, everybody. It's Lynn Thoman. Welcome to another episode. Today, I'm excited to be here with Eric Schmidt, former CEO of Google, founder of Schmidt Futures, and host of the new podcast, Reimagine. Eric's going to talk about how tech and artificial intelligence are going to transform our world, and also what he's doing now with Schmidt Futures and his new podcast,

Reimagine, to make the world better after COVID. Welcome, Eric, and thank you so much for our conversation today. Thanks, Lynn. I've known you for a long time, and I'm really impressed with what you've done. As I am with you. And in fact, before we start our conversation, I'd also like to say thank you, Eric, for all you've done building Google as CEO from a startup to be the search firm it is today. Without search, which Google pioneered, our computers would essentially be useless. Search forms the backbone of the internet. It's what makes it possible for us to work from home on the web, to stay connected to other people through LinkedIn, Facebook, and Zoom,

to entertain ourselves using streaming services like Netflix, and to find and purchase food and other items from Amazon, Walmart, and elsewhere. So thank you, Eric, for all you did building Google. Well, thank you. And all I can tell you is, can you imagine having a pandemic without Google, Amazon,

and all of the other tech companies and the services they provide? Imagine if we couldn't do Zoom. Imagine if we didn't have this kind of connectivity. How would we get through it? What must 1918 have been like? I can't imagine. I cannot imagine what 2020 would be like without all of those great web companies and services. And it was just over 10 years ago, in 2007, that the first iPhone came out. It's hard to imagine that cell phones, email, and texting have only come into widespread use in the last 10 or so years. Eric, you're very involved with artificial intelligence, with companies you've founded, companies you advise, and as chairman of the U.S. National Security Commission for Artificial Intelligence. What lies ahead with AI? How will AI and technology change our lives going forward? You know, most people think that AI is the killer robot, who's inevitably female, by the way,

taking it out on the male inventor who was evil in some complicated way, because that's what we see in the movies. It's not going to happen, certainly not in our lifetimes. What will happen is the systems that we use are going to get smarter and smarter and smarter. And smarter meaning making recommendations, giving you advice, trying to give you insights. Conversation, for example, will become routine. And the conversational developments that have happened in the last year or two are really extraordinary. The ability for computers to understand language enough to be able to answer basic questions, to be able to have a reasonably intelligent discussion about something

is the first part of the next set of breakthroughs. And what is the next part after that? Many people debate what happens after conversation. There's a large number of scientists who believe that the right thing will happen is that aggregate data knowledge will occur. In other words, that the building blocks of how we learn and so forth will occur. There are other people who are more skeptical and think that we need other breakthroughs in the algorithms of AI before we can even replicate what a child can do by age two or three. So there is not agreement on it. What there is agreement on is that AI is

transformative for pretty much every field. I'll give you an example. You're a physicist, and you wake up in the morning and you have some ideas and you go to sleep, you know, having dinner. Well, what you could do as you go to sleep is you could say to the computer, I want you to read everything in my area of physics and tell me something I don't know that's consistent with this insight. And then the next morning, you wake up and you go, oh, my God, the computer saw a pattern that humans didn't see. What does that mean for my theory? So the real activity of AI is not going to be the AI sort of interfering, but rather

the collaboration between people trying to solve a problem, whether it's a social problem or a research problem or a fun problem, and they're going to actually solve it in a new way. So AI involves computers analyzing hundreds of millions of pieces of data in fractions of a second and making recommendations or giving advice or making decisions based on algorithms that humans won't be able to understand. How do we oversee AI to make sure it's only used for good? Well, there's a whole field now around AI ethics. And the rough presentation of their view is that you want to establish some basic principles about AI ethics. And the first is that the AI does what it's supposed to do. In other words, it does what the inventors and developers did,

as opposed to something else, and puts some bounds on its ability to learn and solve things and so forth. In other words, that the intent is what the creator intended. And the second one has to do with algorithmic bias. So one of the observations about people is that we accept bias that humans have. We criticize it, but it's there. So if you train a computer to mimic what humans do, guess what? You get a system that has bias. And there are famous examples where computers came up with a bias against a woman or a minority or something, which was clearly

inappropriate. It didn't do it on its own. It did it because the data had the bias inside of it. So researchers are looking to how to filter, if you will, the training data and the algorithms to make sure that there is not inherent bias in the decisions. Synthetic biology, the intersection of computers and life sciences, is another area of interest from you. And you're a founder of at least one company that I know of in synthetic biology. Where is synthetic biology going? What do you see next? When you hear about synthetic biology, again, you think of, oh my God, we're creating new organisms. God knows what they'll do. But in practice,

synthetic biology is quite different. Synthetic biology broadly is the manipulation of living systems to improve them in some complicated way. And there's tremendous interest now in synthetic biology being used to improve industrial processes, material processes, agricultural chemicals, those sorts of things, to get rid of the bad stuff, the bad part of these life forms, if you will, and only get the good part, make them more effective, have them have less collateral damage, etc. There's evidence that synthetic biology will apply to, as I mentioned, chemicals that are broadly used in the industrial processes, in the food processes,

in everything that we do. This synthetic biology revolution is a chemistry revolution, and it's also a business model revolution because it's a new set of companies which have very high valuations because people realize that when this new substance is inserted, just like with technology that I'm used to, it has a huge downstream effect on the economics of the business, almost all positive. What's interesting about synthetic biology is it started off with a focus on essentially solving some of the problems of renewable energy. And so the original thesis many people were very excited about was that synthetic biology would be used to build essentially replacements for oil, less

carbon loading, less pollution, those sorts of things. And it didn't really work. The technology worked, but the business model didn't work because those were very mature technologies. And people have now figured out that synthetic biology is early enough that you're better off running it in areas which have relatively high gross margins. In other words, they have profit that can be earned and learned and so forth. And so the promise of synthetic biology is that it initially helps you build new businesses and new chemicals and new solutions. And that eventually, as the costs improve and manufacturing improve, it can be even broader and more scalable back to its original intent.

So making it a little more concrete, how will it change drug discovery, for example? The combination of synthetic biology and AI is extremely powerful. So one way to think about drug discovery, again, is the chemists wake up in the morning and they try five compounds, none of which work, and they go have dinner and go to sleep. And then the next morning they try five others. Well, you can automate that using AI to look for targets. And some of the greatest breakthroughs in the last few years have been combinations of biologists, chemists, and artificial intelligence researchers who've been able to come up with algorithms which sift through all those choices very quickly. The same principle applies to synthetic biology because you're dealing with life forms.

And so you have many different choices of how to put something together. The combinatorics become very complicated. And so to the degree that we can allow the development cycle to be more precise, or at least eliminate false choices and narrow the search space, it makes the discovery process far more efficient. So looking at some of the hot topics today, one of them is information and privacy. What information is out there and available on each one of us? Well, let's think about yourself. You have a mobile phone. The mobile phone knows where you are and tells that to the telco. That's accepted for security and safety purposes.

The mobile phone also remembers your various activities and stores a log of all of its activities. The mobile phone manufacturers have various rules about how much of that data is stored on the phone and sent to servers. But in theory, that information could be sent to the servers and used. In China, all of that information is sent to the servers and is available to the government. In the West, there are restrictions on that, and both Apple and Google have announced that they don't keep that data for various reasons. If you go to the application level, each of the apps has a record of the things that you've done. So for example, Google keeps a record of your search history, but under agreements with various governments, that information is kept for

specific periods of time. So the compromise around privacy seems to be that the information is kept for specific periods of time. So the compromise around privacy seems to be that the information is retained in the digital system for a while in case you turn out to be a criminal, which of course we're not. But in theory, if you were, the government could use a search warrant, at least in the West, and get that information. I think that's the stable outcome here, because I don't think people want no record at all, because then how would you discover a terrorist act? But people are very concerned about long-term implications of the storage, and they want this information eliminated. One more point. There's a great deal of computer science research that indicates that you cannot successfully de-identify data.

In other words, if I give you the data and you take my name off, you can eventually figure out who it was because you can correlate that with other life patterns. So it looks to us, at least to me, that you should have a restriction on the amount of time the data is kept. Another hot topic now is oversight. And how should we think about overseeing big digital companies like Facebook, Amazon, and others? I mean, as you know, Facebook has more than double the number of users that live in China and almost 10 times as many in the US. So some of these companies have just gotten so large. So there's much discussion about how should we oversee them?

What do you think? So first of all, I would say that every company that is named in these situations is under many, many different laws already. So there are laws about privacy. There are laws about commerce. There are laws about what data can be released and where. And every company is struggling with the difference in laws. Usually the laws in the EU are the most restrictive, and the laws in Asia, and in particular China,

are the least restrictive if you're able to operate. Of course, China may not allow you to operate anyway. And so there are complicated negotiations between the governments and these companies. And that's, I think, where we are. The industry would like self-regulation for obvious reasons. And my guess is that the self-regulation will work until a big mistake happens. And so at that point, there'll be enormous political pressure to regulate whatever mistake occurred. And I think that's the likely cycle. I don't think that the regulation should be a priori. And the reason is it's extremely difficult to anticipate where the technology will go. And premature regulation, which is what people

love to talk about, could forestall a technological solution that was better. So I'm always in favor of, oh, okay, we had a problem. Who screwed up and did they admit it? And how are they going to fix it, et cetera? The most obvious one being the interference in the 2016 elections, which was widespread in the EU and in the United States, that needs to stop. How do you think about calls to break up some of the big companies? One of the questions is, why are these companies so big? And it's unfortunate for the people who are arguing for the breakups is they're big because they're useful when they're bigger. So it's fundamentally easier for a consumer to see a global view

and get access to that. And for many, many reasons, these are network effect companies where there tends to be one singular winner that's a global platform. So that explains Facebook, et cetera. There are some exceptions. So when you think about breaking up, you want to think about what is the objective you have. So I'll give you an example. Senator Warren, when she was a candidate, argued, among other things, to break up the Apple App Store from the iPhone. And on first blush, you say, well, Apple's a very big company. It's the most valuable company in the world.

Maybe that makes sense. So the App Store subsidizes the iPhone. So the result of breaking them out would be that the iPhone price would go up. And would the App Store price go down by the same amount? Probably not. That's probably a price increase, I'm guessing now. Then the App Store in Apple model, and I'm using Apple because it's easy for me to talk about a company I'm not so close to.

Then what happens is that the App Store is unbundled, but you're going to want the security and safety of an app store. Wouldn't you continue to use that one? So in other words, when you think through the breakup, you may not get the benefits that you think you are. And so my long experience dealing with EU regulators is to say that when you start, you need to say what you're really trying to do. And you need to come

up with a regulatory outcome that will in fact generate it. So here's an example. You don't like Apple? Block them. Okay, well, that will work. But now you're denying your citizens. Again, this is a rhetorical point, so don't get the wrong point here. So now you're denying citizens that extraordinary innovation of Apple. So I push back pretty hard on the breakup scenarios because I think the questions are ill-formed. In thinking about regulation and governments, China looms very large. How important a market is China and what are they doing now? So China is an extraordinarily impressive computer market. And there are areas in China where China is well ahead of the West.

They're clearly ahead in face recognition. No one disputes that. Their face recognition and gate following and so forth is far better than any other country because they've invested so much in it for the obvious reasons. They're also very far ahead in electronic commerce. So last time I was in China, one of my friends explained he had not used cash for two years, that that was how pervasive the use of, in this case, Alipay and I guess it's WePay are. And they have many apps inside of WeChat, which allow you to have some level of interaction between the phone and the vendor, which is pretty powerful in terms of what you can do. The system is such that China has unified security, identity, the phone, and the payment system, and your presence where you are into one

solution. Such a solution would be completely unacceptable in the West because of the obvious privacy issues, which are very long. So I'm not advocating for that. But I'm saying with respect that China has managed in its own society to solve the problem in a way that's well ahead of what's been possible in the West. China is working very hard to lead in technologies like software, AI, microprocessors, energy, quantum physics, things like this. It remains to be seen how well they'll do. They've made a national priority of focusing on those areas, and the West needs to respond. And the best response is not for us to block China, right? It's for us to compete, for us to invest more in synthetic biology, more in software, more in AI, provide

the new solutions, provide the better solutions, be the innovator. That's the historic strong point for the West and in particular for America. How do you see our actions toward China? Many of the actions that we have taken as a country appear to me to be defensive or reactive, and they lack a context. And I am concerned that the decoupling of China is the one-way street that eventually leads to confusion and national security danger to the country. My personal view, which is just my personal view of China, is that it's better to treat China as a rivalry partner. They are clearly a rival, and they're clearly partners. They're rivals in the areas of competition that I'm describing. They're rivals over some national security issue,

but they are, in fact, partners over trade, certain areas where they're a very important commodity supplier for America. They're our largest trading partner, where they're a very important commodity supplier for America. They're our largest trading partner, and they're also our biggest strategic competitor. And they are to be taken seriously. They're not like the Soviet Union, who are economically weak and poorly run. China, for its objectives, is well run against the strategy that they've set out, at least for the moment. The debate of what happens to the future of China is all over the map. Many China scholars believe that China has endemic problems, which will lead to its failure in the next 10 to 15 years. But the last 10 to 15 years of hyper growth in China indicates to me that they've still got some oomph left, that they still have some

energy. In particular, in our AI commission, we've studied and released information that indicates that China is investing very, very rapidly in the key areas that I'm describing, and in particular in AI. What is our answer? We need more money. We need more people. We need to keep the really smart foreign graduate students that we're busy kicking out of our country. We need to keep them in America. We need to make them citizens. They need to keep the really smart foreign graduate students that were busy kicking out of our country. We need to keep them in America. We need to make them citizens. They need to found companies.

They need to pay taxes in the United States. The American model of innovation has got us this extraordinarily plan where the five top companies in the United States are tech companies, right? We want to continue that. Maybe different companies, but we want it to be America. And by the way, in China, the three most valuable companies are tech companies. That's the race right now. Let's talk about COVID. You're leading a task force for New York's Governor Cuomo on COVID. And you've also just started a podcast called Reimagine on COVID and reimagining the world post-COVID. Let's start with what do you

think the world will look like post-COVID? How will it be different? The pandemic has brought to light the structural failures and problems in all of our governments. China, where the virus first emerged, appears to have understated the extent of the outbreak in the first few months to an alarming degree. The American response is so poor in aggregate that it's hard to describe. Very few countries have escaped the impact of the pandemic in terms of the death rates, the increase in equality, and in particular, the impact on the less fortunate. We should not be proud of our global or national response to COVID. We should ask ourselves, why could we not do this better? Governor Cuomo, who's done generally a pretty good job in this area, has been data-focused, has been relatively transparent about the decisions, created a commission of which I'm the chairman,

which is examining what the changes on a go-forward basis will be. The committee has not yet released its report, but to give you a sense of the feeling of it, the most important conclusion that everyone seems to come to is how important connectivity is. And I, for one, was not aware in New York State of how many people do not have basic connectivity. By basic connectivity, I'm talking about a relatively slow broadband connection upon which their children can do schooling. So it's pretty simple. During the pandemic, if you have to work from the equivalent of home and schooling from

home and you have no connection, you really are an outcast. The governor has also prioritized telehealth for obvious reasons. Telehealth used to be slow and people were sort of uncomfortable with it. Now, the majority of people are using telehealth and it's working well. Adopting that will help everyone for obvious reasons. And the third that we've taken as a priority is the question of worker retraining and how we work. It's pretty clear that the pandemic will result in structural changes in the labor force. And the best thing to do for people who are affected, of which there are unfortunately many, will be to give them both retraining opportunities, but also just job

opportunities. What are the jobs that are available to me now that the current job is gone? I don't think we're done with the layoffs and the restructuring due to the pandemic, unless there's a significant new stimulus package, which again is temporary. The damage, especially to women and to people of color, people who serve us essentially in one way or the other, the people who live in more crowded homes, they don't have control over where they work, and then they lose their jobs. They live hand to mouth. Many, many of our fellow citizens are affected by this. We have to figure out a way to help them. How do you think that cities and businesses will change? What will they be like post-COVID? You know, in January of this year, the United States had the lowest unemployment it's

had in 50 years. And our cities were thriving. People were moving from rural to middle American cities, middle American cities to the big cities, property values going way up, all of that. That was only six or seven months ago. We've almost forgotten that that was our going assumption. The underlying assumptions of cities is that density matters, and in particular, productivity is higher. My guess is that cities will return, because they always have in the past, but with some different assumptions. The economics will be different. Prices will be different. Certainly, real estate prices will be lower, especially for commercial businesses because of lack of demand. But the vibrancy

of cities has always been true. It's been true for hundreds of years. And there's every reason to think that it will return. One of the quickest ways in which it could return would be the broad development of antigen tests, which are the quick COVID tests, the ones that take five minutes, 10 minutes, the swabs and so forth. There are many of these that are in approval cycles with the FDA and are generally becoming available in the next month or two. So you want to imagine a scenario where you have a tall building that's empty. The people would like to go there. The employer would like to bring them there. They're more productive probably at work in many cases, but they don't want to for safety right now.

But imagine instead that you had a weekly antigen test, which took five or 10 minutes, and that you did that weekly, that the employer would pay and the employees would be happy with this because they would then know. Ubiquitous and broad testing really does allow you to have as normal a life as possible during what will be a lengthy pandemic. Universities that have implemented once or twice a week testing seem to be able to keep the outbreaks under control.

Now, universities are special cases because they're like ships on the ocean. But from the standpoint of the data, it looks like a weekly test regime will work well enough that you'll be able to spot an outbreak, in particular, spot an employee who does not know that they've been infected. It's not foolproof. That may be the eventual solution. Those tests are becoming available in the next few months. Talking about your new podcast, Reimagine, which is terrific. I really enjoyed your episodes with former FDA Commissioner Scott Godley, and also with Chile's former finance minister, Andres Velasquez. Can you tell us about Reimagine and why you started it? It seems to me that many people are stuck and that we're stuck now in this narrative in between the election

and the other problems that everyone is having, that we need a forum to talk about how these problems are going to get solved in a structured way. We'll see how well my podcast goes, but I'm very committed to getting the ideas of how we're collectively going to solve this problem. Since, bluntly, we've had a governmental leadership failure. We've had a failure where the leaders who are supposed to keep us protected have ended up in a situation where there are a thousand deaths a day in America, which is unbelievably harsh. People are not coming back. They're dead. It's a horrific answer. We need to do everything that we can to move the dialogue forward.

And it's not just the United States. In Europe, many of the countries are now struggling with new outbreaks. Most countries are still grappling with the effect of the virus. South America is shut down and really suffering. And the next big set of problems will be in India and in that part of the world. So the COVID is not over. And if Free Imagine can give us a way of thinking about how the world will look, I'm looking forward to its success. Do you see any opportunities going forward? Are there areas that we could focus on to make the world better?

What has been lost in the political discussion is that the most important thing just to get the infection rate, which is known as R0, down. If the infection rate gets down below one, the disease, whether you like it or not, naturally dies out and becomes more sort of manageable. And at that point, contact tracing approaches and so forth work. As long as the reproduction rate is greater than one, the disease is spreading. That's why there's so much focus on test positivity, because that's a possible indication that you have an outbreak that you don't fully understand. The Northeast of America, which went through, again, a horrific March and April, now has very low test positivity rates. So it shows you that if you work hard,

you can keep it low, and hopefully it will stay that way. Without getting the infection rate below one, it's going to be very difficult to get our country back to work because at the end of the day, people are getting sick. And this illness is a really serious one. People are having trouble with the illness today. They were sick in March or April. We don't fully understand their illness. Everyone is focused on a vaccine, which of course would be wonderful, but do the math. Let's imagine a vaccine were available to you in the first quarter and you took it. Would you fundamentally change your behavior and go back to all of the pre-COVID behaviors, or would you wait and see? Would you see how effective the vaccine was? Would you see what happened to your colleagues

and friends? Would you feel comfortable doing things like going to ball games and indoor events and so forth that involve an awful lot of super spreader possibilities? You would certainly not. We don't have the disease under control. We seem to have lost the formula that the most important thing to do is simply get the rate of infection down. And the easiest way to do that is universal mask wearing, literally making sure people don't get exposed. Can you tell us about what you're doing at Schmidt Futures to make the world a better place? Because I was so fortunate at Google, we founded a group called Schmidt Futures, which is trying to focus on both talent and science. I observed in my whole career, the most important thing was exceptional talent that was global, that was diverse, inclusive, and full of interesting people, different backgrounds, different ideas,

very, very creative people. So we've put in place a set of programs. We're doing funding around research, and particularly around COVID and AI and biology, which is a key component of trying to sort of move the knowledge base forward. And we're also focusing on developing talent. The most important recent announcement is the program called RISE, which is an attempt to identify the top and most high potential 16-year-olds in the world, and in particular, find a way to get them out of their countries and get them to the United States to get them acculturated and get them to their full potential. We'll see if it works. That's a partnership with Rhodes. That's an exciting program. Eric, you're one of the most thoughtful, insightful, and crisp thinkers that I know. I don't want to constrain you by asking you a narrow or specific question. So let me ask you

a very broad one. What are three ideas that you have that would surprise people? I'll give you a negative one and two positive ones. I think a negative one is that it's going to be very difficult to control the inequality that naturally arises from the kind of technologies and the kind of approaches that we are pursuing. Because ultimately, these are network effect technologies. There's a small number of winners, and those winners tend to get most of the profits. This is the world I've lived in for a very long time. That's going to continue to be a problem. It's a very hard problem. It's a hard problem politically. It's a hard problem for the losers, et cetera. But the benefits to humanity are high for these winning technologies. I think that the acceleration that is on the more

positive side, the acceleration of the union of AI and biology is going to be extremely powerful. What I didn't understand about biology, but I do now, is they don't really understand how biology works. They just understand it at a certain level, and below it, they don't really. So what they have to do is they have to watch it and then use computers to say, oh, we did this and that, and then the following thing occurred. And combinatorially, they can begin to understand the language of biology, which is still today. It's a masterful thing developed by God in evolution. What an extraordinary thing that is the basis of all living things. The progress in biology in the next 10 years for good, for health, for safety, and so forth will be profound.

And my personal view on AI is that there will be future breakthroughs. They'll take longer. And those breakthroughs will be the ones that really give us a partner, a partner that can really learn faster than we can in certain areas. And we'll keep it under control. It won't bite back at us. Is there anything else you'd like to discuss that you haven't already touched upon? No, I think this is very good. Thank you. Last question.

What are the three key takeaways or insights you'd like to leave the audience with today? The first insight is a simple statement. We can handle the pandemic better than we have. There is no excuse for the rate of infection, the rate of illness, and the rate of death that we see today in any of these mature countries. It's not okay in America. It's not okay in Europe. These are governments that have a duty to protect. They can do a better job, and they need to have an honest conversation with their people. We can do better. If we acted quickly, we can get ahead of the pandemic before the winter, where there's a great deal of concern about increased transmission, and get through it and

have this be a two-hump and not a three-hump pandemic, and then move on. The second insight is that the ability to use AI to make the world more interesting, right? To make knowledge more available, to improve learning, to build systems, for example, that follow the way you learn and entertain you and educate you at the same time is beginning. And the third point is that we're just at the beginning of the next wave of entrepreneurs. We highlighted, for example, the biology ones, but there are plenty. And there's a new generation that's almost two generations younger than me now that are just coming out of school that have extraordinary ideas of how to combine these things. And they move even quicker than their predecessor generations.

And that is the best hope for us and for our future collectively. Eric, thank you for your work to make the world a better place through your work with governments, with businesses, through your new podcast, Reimagine, and also through your work at Schmidt Futures. And thank you as well for a really interesting conversation. Okay, well, thank you, Lynn. As usual, you're the most thorough and the most thoughtful person that I know. So thank you very much. Well, I would say that you are one of the most interesting and insightful people that I know. So thank you, Eric. If you enjoyed today's episode and would like to receive the show notes or get new fresh weekly episodes, be sure to sign up for our newsletter at 3takeaways.com or follow us on Instagram, Twitter, and Facebook. Note that 3takeaways.com is with the number 3. 3 is not spelled out. See you soon at 3takeaways.com.