Eric Brynjolfsson
Appearances
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
Well, that's a good question. Fei-Fei Li started it along with John Ichimendi, and they recruited me out to Stanford. You were MIT? Yeah, I was at a major city in Massachusetts. You're picking bones with me because that is the other city that thinks it's a major city. Well, I'm sorry. It's too bad we keep beating New York in all the sports, but that's just the way it goes.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
Good, yeah, you can cope. So the idea is that AI is doing these amazing things, but we want to do it in service of humans and make sure that we keep humans at the center of all of that.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
A lot of technologists are very focused on the technology, but I'm an economist, as you mentioned, and there are political scientists, sociologists, artists, and we're all working to use AI to help lots of the other parts of the world and of academia.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
Yes.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
having ai or having technologies that let humans be human in their most essential way or what totally too many people think of machines as just sort of trying to imitate humans there's this iconic test of artificial intelligence called the turing test that many people are familiar with which is how much can you make a machine mimic a human to the point where you can't tell which is which and that was i think a visionary idea when alan turing proposed it in 1950
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
But in a way, it's a very constraining idea because machines can help us do new things we never could have done before. And that's a much higher ceiling. And so we want to look for ways that machines can complement humans, not simply imitate or replace them. And so I wrote a paper recently called The Turing Trap, trying to steer people away from this idea of just imitating humans.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
I think it's still part of the conversation. I don't know. Maybe in the press, there's things that go up and down in cycles to some extent. AI is becoming much more powerful. There continues to be rapid progress. And the good news is we can have tremendously higher productivity and wealth and have medical solutions addressing poverty in the environment. But it also raises a number of risks.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
misinformation or people using it in a way that creates weird interpersonal dynamics, AI boyfriends and girlfriends, you know, maybe millions of those that people have as their primary relationship, pathogens and even catastrophic risks.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
Well, the great thing is AI can help us discover new drugs as well as, you know, as medicines, new materials. There's just a study from a grad student at MIT describing how researchers using AI were able to discover 44% more materials than a randomly assigned group that didn't have access to the technology. So, big difference.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
But some of those new materials could, you know, lots of them could do good things. You could also create dangerous ones, you can flip the bit on a drug that's meant to make you healthier, and it could make you much less healthy to the point of killing you.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
I'd say I'm a mindful optimist. What I mean by that is that they're sort of blind optimists. I run into a lot of those in Silicon Valley who are just like, hey, don't worry, just chill. It always works out in the past. You know, it's going to be great. Just sit back and we're going to have a great time. And there's a lot of pessimists who basically say the opposite.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
And they both make the same mistake, I think, which is they take the agency away from us. It's like this technology is going to do stuff to us. And, you know, whatever it is, is what it is. I think that we have a lot of choices. One of the reasons I came to Stanford, the Center for Human-Centered AI, is that I think we can help steer the technology in ways.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
And if we do it right, we could have the best decade we've ever seen, but it's not inevitable.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
That would make sense. That's what people, when I go to Iceland, that's how people know me. It's true.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
That's exactly right. Simon Kuznets and his team in the 1930s basically developed what we now use as our national accounts, GDP, productivity. It's all based on this system of accounts. Paul Samuels got one of the greatest inventions of the 20th century. I agree, but it's also been massively abused and misused. Nowadays, if you see a headline, economy grew by 3.2%, they mean GDP increased by 3.2%.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
And why is that an imprecise or not useful metric? GDP measures... basically everything that's bought and sold in the economy. What that means, with few exceptions, if something doesn't have a price, it's not counted in GDP. So we're missing a lot of important stuff, you know, clean air.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
A classic problem is if I cook a meal for myself, that's not part of GDP, but if I hire somebody to cook it, or if somebody pays me to cook it, then it is part of GDP. So you have a lot of little weirdnesses like that, a lot of household productions out there. And one of the biggest ones, I'm the director of the Stanford Digital Economy Lab,
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
All these digital goods that are often free, Wikipedia, search, Facebook, texting, email, if they have zero price, other than the electricity and a few other things, they're basically not counted in GDP. Yet people get a lot of value from them. Right now, the average American spends a little over eight hours per day looking at a screen of some sort, TV, computer, whatever.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
That means they're spending slightly more than half their waking hours interacting with bits, not with all the other things. And that means a big part of our lived experience is these things that are not being well measured by traditional GDP.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
Well, you're absolutely right. We have a lot more wealth than we had before. But we also did before as well. I mean, there's, you know, television and penicillin things were also not counted very well. But wealth is not the same as happiness, as we know. And so it doesn't automatically translate one for one.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
But it does mean we have an imperfect measure of the value that our economy is creating for us.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
Easterlin. It's called the Easterlin paradox that what you just said, that as people got richer, they didn't seem to get happier. More recent research found that actually it just sort of is diminishing returns. Like you're saying, it's not that it actually stops.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
So the B stands for the benefit. As I said earlier, GDP is basically a measure of production, what it costs to produce things. The GDPB is trying to capture the consumer surplus. It's the value between what the most you would have paid, what you actually have to pay. So with Wikipedia, if you would have been willing to pay $15 a month and you pay zero, you're getting $15 of consumer surplus.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
Yes, we are. How do you do that? So if I were to do a survey and say, okay, I'll give you $500 to stop using the internet or stop using email, more realistically, for 30 days. Some of you would raise your hands and some of you wouldn't. If I said, okay, what if I gave you $50? What if I gave you $5? You get fewer and fewer people being willing to give it up.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
And that gives you a downward sloping demand curve. A lot of people think it's worth at least $500. Not so many think it's only worth $5. And the area under that curve, we call that consumer surplus. And if you do that for lots of different goods, you start getting a sense of how much value all these different goods are creating.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
No, it's something I came up with on my own where we decided that we should measure consumer surplus and not just cost. And so we started doing some small scale surveys and we got some money from different groups, the National Science Foundation, the Sloan Foundation. So I guess part of it is a government project. And we'd love to get more support to do it at a larger scale.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
We're doing about 250 goods right now, including digital goods, non-digital goods. We ultimately want to get a representative basket of several thousand goods that we can track periodically alongside traditional GDP and see how they compare.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
Yeah, I mean, I think already there is a standard among many antitrust experts that came out of Chicago, actually, that, you know, we should look at consumer welfare as the ultimate metric. But that was a few decades ago. I feel like we've moved past that now. Well, that's the concept, but in terms of measuring. So, Lena Kahn has a different measure, and there's an ongoing debate.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
But I like the consumer welfare standard, which is, you know, is this concentration, this merger, the spinoff, whatever, is it making consumers better off or worse off? And that's hard to measure, but our tool, GDPB, gives us a set of measures for that. And you're right. In many cases, there's a tremendous amount of value from Google search or from email.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
One of my students was over in the European – commission, and they were upset about all the money that the big companies based here in San Francisco and around the United States were making on European consumers and saying, oh, this is a very unbalanced thing where there's a lot of money going in this direction.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
And he pointed out, well, actually, if you measure the value that people are getting, there's far more value that French citizens are getting from these services than what they're paying. So the net gain is in the other direction.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
Not as much as them having their own tech firms. And so when they have companies like Mistral, I met with the finance minister in France, and he was coming around to the view that, well, actually, maybe tech companies can create some value now that we have one of our own. Yeah. You didn't quite say it that way, in fairness.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
I agree. I mean, look, I moved out here four years ago from Boston, another pretty innovative place, but I really underestimated how amazing the culture is out here. I am constantly meeting people doing startup things. They have these grand... visions and dreams. And I think they're mostly pretty sincere about wanting to change the world for the better. I moved here in summer of 2020.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
I remember it was COVID. We had this garden party and I sort of half-jokingly nervously said, hey, are you guys all going to be moving to like Austin or Miami or something? Did I kind of miss the party? And they laughed and said, oh, don't worry. It kind of seems like, you know, San Francisco and the Bay Area is going downhill. They said, don't worry, there'll be something.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
And we didn't know about ChatGPT at that point. Maybe a few of the people in the room might have been working on it for all I know. But then, of course, there's this explosion. And if anything, I think the tech innovation scene is even more concentrated in the Bay Area now than it was when I came in 2020. And there's just a whole wave of innovations. People talk about it, but I'm an economist.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
I don't really fully appreciate it until I'm here. There's a cultural element to it, an attitude. It partly attracts people from around the world who have this mindset of wanting to change the world. People help each other to do it. And like you said, it's been a tremendous engine of creativity and wealth creation.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
Yeah, it's a tragedy. Part of it, to be fair, I think it's overrated. Like London Reed was saying, I heard, you know, apparently it has the lowest murder rate in 60 years. And, you know, I come up to San Francisco a fair number of times. I don't think it's like what is described by Elon Musk or others on Twitter. Well, there's a... Good timing. But part of it is real.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
I just walked over from Union Square and there were definitely some homeless people on the street. In the green room, I gave London Breed a little bit of a hard time, you know, about why do we allow this? I think in some ways, I'm not a sociologist, but maybe all the wealth and success allows a lot of sort of slack and allows them to get away with a lot of mismanagement.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
I'm not singling out her or anybody in particular, but... I do think a lot of that government is not managed as well as it could be. They can get away with it because there's just so much innovation and wealth being created that you can have a lot of slack. I'm hoping that will get tightened up a bit because it doesn't reflect well on California or on San Francisco.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
And with all the money being poured in to try to support San Francisco and the Bay Area, we should have the cleanest streets, the best police force, the safest neighborhoods, and we don't.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
I think I'd be a terrible mayor. Economists do think differently about problem solving, right? Well, a lot of them are probably just common sense, but a few of them, you know, where economists maybe differ. Like I'm a huge fan of congestion pricing. Almost everyone who I know who is not an economist doesn't think that's a good idea, but. Well, I guess some of these guys may be economists.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
What are the first rules of taxation? You tend to get less of what you tax. So if you're taxing work and investment, you're going to get less of that. Why not tax pollution? Why not tax congestion? I also sometimes advise Singapore, and they've put in place all these rules that economists have.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
No, no.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
I've been there and met with senior officials, just like I meet with senior officials in lots of different places. I mean, not like an official advisor or anything. Do you have a badge of some kind? A badge? Do you have a cape? No, no. I am just a professor. But what I like about Singapore is they listen to professors and they don't listen to professors in the U.S. Congress.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
And that's one of the reasons that they're successful over there. I think, you know, I'm biased. And one of the reasons so many people are leaving California is because it's so expensive here. It doesn't need to be. And the reason it's so expensive is because a lot of people would love to live here. But the housing prices are insanely high because, you know, it's supply and demand.
Freakonomics Radio
Is San Francisco a Failed State? (And Other Questions You Shouldn’t Ask the Mayor)
It's elementary. It's economics 101. These are things that are actually really easy to fix. And, you know, we could do a lot better.