Podcast: The case for global optimism with Steven Pinker

TRANSCRIPT: The case for global optimism with Steven Pinker

Steve Pinker:

As long as bad things haven't vanished from the face of the earth, which they never will, you can get the impression that things are unchanged or even are worse than ever even when they're improving. But if you compare the number of wars and the number of people killed in wars in the '60s and the '70s and even the '80s, we're actually much better off today.

Ian Bremmer:

Hello. And welcome to the GZERO World Podcast. This is where you'll find extended versions of my interviews on public television. I'm Ian Bremmer. And today we are talking about laundry. And by laundry I mean human progress. In 1920, the average American spent 11.5 hours a week doing laundry, and that average American was almost always a woman. Dudes just kept wearing their dirty clothes. By 2014, the number had dropped to 1.5 hours a week thanks to the invention of the washing machine. By freeing people of laundry, and I mean, doing laundry because freeing people of laundry is something we don't approve of here.

This new technology allowed parents to devote more time to educating their clean children, and it allowed women to cultivate a life beyond the washboard. Laundry is just one of many metrics, thankfully that my guest today, Harvard psychologist Steven Pinker uses to measure human progress. But are the various metrics of success that Pinker employs enough to offset war in Europe, famine in Africa, global pandemics, fake news, AI or Mageddon, and that's just your average Tuesday. To talk about all that and more, I'm joined by the laundry lover himself, Steve Pinker. Let's get to it.

Announcer:

The GZERO World Podcast is brought to you by our lead sponsor, Prologis. Prologis helps businesses across the globe scale their supply chains with an expansive portfolio of logistics real estate, and the only end-to-end solutions platform addressing the critical initiatives of global logistics today. Learn more at prologis.com.

Ian Bremmer:

Steve Pinker, thanks so much for joining us on GZERO World.

Steve Pinker:

Thanks for having me.

Ian Bremmer:

So you are portrayed as this relentlessly optimistic macro thinker, but you would say, "Hey, I'm just looking at where the data is going."

Steve Pinker:

Indeed, and the view of the world you get from data is different from the one you get from journalism because journalism is a non-random sample of the most dramatic, usually the worst things that happen on earth on a given day. A lot of the positive developments are either things that don't happen, like there's a country where a war did not break out. A city that has not been attacked by terrorists or things that unfold gradually, a few percentage points a year which can compound, but there's never a Thursday in October in which it's a headline. And so if you don't look at data, if you look at headlines, since as long as bad things haven't vanished from the face of the earth, which they never will, you can get the impression that things are unchanged or even are worse than ever even when they're improving. It's only when you count the number of wars, number of deaths in war, longevity, child mortality, extreme poverty, number of leisure hours that you see that there actually has been improvement, a fact that you could miss if you just follow the headlines.

Ian Bremmer:

Life expectancy, infant mortality, education rates, I mean, almost all of those things, if you look over the last say 50 years, unprecedented improvement in the human condition.

Steve Pinker:

Indeed, and it's not magic. It doesn't happen everywhere all the time. Some things do get worse sometimes because the world does not contain any force that just lifts humanity up. Quite the contrary, the laws of the universe grind us down. It's only when people apply their ingenuity to try to improve the human condition that every once in a while we succeed. If we try not to repeat our mistakes, then progress is possible on average, but not everywhere always.

Ian Bremmer:

Now, by that I generally agree with this. So let's have a few minutes on celebrating that, though I'm then going to focus on some places where maybe it doesn't apply as well and particularly right now. But if we wanted to look in the last couple generations, so since you and I were kids, what are a couple of individual data points, things that have happened that feel the most like magic compared to who you were at that point?

Steve Pinker:

Yeah. Globally, a decline in extreme poverty would have to be one of them. The global poverty now is estimated to be maybe eight or 9% of the world's population. 200 years ago it was 90% of the world's population, and even 30 years ago, there were a billion more extremely poor people than there are today. So that would be way up there. Illiteracy probably today among young people, maybe 10% are illiterate, it used to be more than half. Even the number of wars, and not withstanding the horrific war in Ukraine, the war in Sudan, they recently concluded war in Ethiopia which has led to an uptick in the wrong direction. But if you compare the number of wars and the number of people killed in wars in the '60s and the '70s and even the '80s, we're actually much better off today.

Ian Bremmer:

And are there any particular turning points that you would point to? I mean, is it the semiconductor, is it the internet? Is it people just having mobility they didn't have before? What are the things that facilitated that?

Steve Pinker:

Yeah. The globalization probably had something to do with it. All of these things are statistical probabilistic. You can always find exceptions, but generally, if the world is more knitted together in commerce, if you get rich by buying things instead of stealing them, if you have customers that you don't want to kill, if you belong to a club of trading nations, it doesn't guarantee that you won't go to war, but it lowers the odds. It changes the incentives. The fact that wealth no longer comes from land as it did a couple of hundred years ago, if you want to get rich, you don't need more and more and more farmland, but better tech, more knowledge. Rise of democracy, not a perfect guarantor of peace by any means, but seems to be a-

Ian Bremmer:

And not moving in the right direction now.

Steve Pinker:

... and not moving in the right direction now, but statistically is a contributor to peace. Also, it's hard to put your finger on this vague thing called norms, expectations, values, but there's probably less of an emphasis in most of the world or much of the world on national glory, preeminence and more on human wellbeing. That is nations are somewhat more interested in keeping their citizens educated and rich and well-fed compared to having more square inches on a map.

Ian Bremmer:

Now, you sound very Canadian when you say that.

Steve Pinker:

Yes.

Ian Bremmer:

Which you are.

Steve Pinker:

Which I am.

Ian Bremmer:

Now, why are the Canadians seemingly a little more well-run and happier than the Americans these days?

Steve Pinker:

A better question might be why is the United States punched below its wealth in so many measures of wellbeing?

Ian Bremmer:

Given how much capability economically-

Steve Pinker:

Given how much capability-

Ian Bremmer:

... the Americans have, no question.

Steve Pinker:

... so the United States is an outlier among affluent western democracies. Canada is not so different from the other commonwealth countries like Australia, New Zealand, Britain itself, Western and Northern Europe. But the United States is the anomaly because it is rich. It's more or less democratic, but it has lower life expectancy. It has poor scores on Math tests. It has more obesity, more drug addiction, more violence.

Ian Bremmer:

And you think we'd be fat and happy, but it turns out we're fat and unhappy.

Steve Pinker:

Yes. Literally, we are far

Ian Bremmer:

So why do you think-

Steve Pinker:

... and less happy, by the way.

Ian Bremmer:

right

Steve Pinker:

The United States is less happy than it ought to be given its affluence and freedom.

Ian Bremmer:

So what are your views? I mean, you're a cognitive psychologist, you live around Americans. What's your view? Why is that?

Steve Pinker:

There's probably an optimum amount of government support for Social Services, for old age, for the poor. The United States probably falls below it. Now, that can be too much and that can be too little, and I think the United States is probably on the wrong side of that curve. So people are more anxious about losing their job if they get sick, about becoming a destitute. There's less government attention to poverty and segregation and other problems that the government tends to take on in other countries. Also, the United States has a tradition, especially parts of the United States, the south and the southwest, of hostility to any centralized government and planning. It's much more of vigilante justice. Every man defends his own home, defense is honor, and must secure the credibility of his implicit threat by responding to insults. And so Americans are more likely to get into stupid barroom brawls and lethal disputes over a parking spot and road rage. Europeans and Canadians are more likely to call police.

Ian Bremmer:

To not have parking spots.

Steve Pinker:

Yeah, exactly.

Ian Bremmer:

Why?

Steve Pinker:

That might help too.

Ian Bremmer:

No, fair enough. So I think that's a good way of putting it, and there's no question that if you take the objective data of US growth and wealth as well, over the last 50 years, there's been an extraordinary uptick. It hasn't come with a sense of wellbeing of the average American.

Steve Pinker:

Not as much as you'd expect. I mean, it says it's still a pretty happy country compared to the 193 countries in the world, but less happy than it should be given how rich it is.

Ian Bremmer:

How much do you blame the structure of the media as opposed to just the job that journalists have always had? Because I mean, it's not new. I mean, people that are writing or writing things that people want to read on the day, but has that changed over the last decades, even before we talk about social media and algorithms, but has it changed just the media itself?

Steve Pinker:

It has that there have been quantifiable trends for the media to become more negative, more pessimistic, even as the objective indicators of the world have improved. But the news is getting more and more morose, and you can quantify that by just counting the number of negative words and number of positive words. So that misinforms people about the state of the world, it means that they become cynical about the causes of global improvement, that they fail to appreciate what has been working. Often they just check out and become either fatalistic like, "The world's going to end by 2050, so we should just enjoy ourselves while we can," or radical, "The system is so dysfunctional that we should just burn it all down because anything that replaces it is bound to be better than what we have now." By not focusing on the successes, many of which are only visible through data and not through events, since they're often non-events, I think the media can give a misleading portrait of the world.

Ian Bremmer:

So we've talked a fair amount about globalization and all the things that have been positive of the comfort. I want to focus on some of the pushback. First, we're coming out of a pandemic, and for the last three years, the human development indicators that have come out of the United Nations, which had been getting better and better and better are over the world for 50 years, have shown five years of moving back, lack of progress over the last three for the world, including on indicators like extreme poverty. How much does that concern you?

Steve Pinker:

Oh, it does. And again, it would be a shock if it hadn't. A pandemic is a real event that makes hundreds of millions of people worse off, it makes the economy worse off, it almost had to. The question is, will it reverse the curve? Unlikely, we don't know. But just taking it as a point of comparison, the Spanish Flu epidemic of 1819, if you look at the curves for life expectancy and prosperity, you can see a definite notch, but then as the pandemic receded, the progress resumed. Likewise for HIV/AIDS in Africa, life expectancy really went down, but then with the spread of antiretroviral drugs, it resumed its upward course.

Ian Bremmer:

Now, if it was just a pandemic set against trends of evermore globalization, I would immediately say that sounds right and be optimistic. I'm not so sure in the sense that as you intimated before, we don't have the same level of support among populations for open borders, free trade. Indeed, we see a lot of retrenchment. We see a lot of industrial policy. We see a lot of near shoring, French shoring, whatever term you want to use. It's definitely not, "Let's make sure that things are as cheap as possible for everybody. Let's make sure that we take care of global poverty." It's, "My backyard that really matters," that's us versus China, it's Russia versus the advanced industrial economies. It's pretty much every populous movement around the world. Do you think that globalization, not necessarily reversing, but at least plateauing, do you see that happening structurally leaving aside the pandemic?

Steve Pinker:

Yes. Maybe, but maybe not is the way I'll put it. Because of course, you're right. The populist movements are pushing back against globalization, nationalism, both from the right and from the left, but there's at least a force pushing toward greater globalization, namely the basic laws of economics that say that's a way to get rich. Going back to David Ricardo and Adam Smith, so that's always going to be true, and people are always going to, together with all the other things they want, they are going to want more cheap stuff. They're going to want to get richer, and so that's at least a push in the direction opposing populism, which one's going to be stronger and for how long? Impossible to say.

Ian Bremmer:

I mean, there's been this huge pushback against the highest level inflation in the United States that we have seen in over a generation, and yet all you would've needed to do is reciprocal removal of the tariffs, US China tariffs from the Trump administration. You take a hundred basis points off of inflation. There'd never been such a political force to do that, and yet that was considered a third rail. You couldn't possibly do that politically because you need to beat up on China.

Steve Pinker:

Yes.

Ian Bremmer:

I mean, if that's structural, that's a serious rejoinder to the idea that we're just heading back to globalization because people want cheap stuff.

Steve Pinker:

It may be specific to China which, of course, is a rising power threatening in many ways, but still, people might want to buy cheap stuff from Bangladesh and Vietnam, and Philippines and Indonesia. So China might be an exception simply because it does appear to be such a threat, and it is a third rail.

Ian Bremmer:

If China is an exception, but they're soon to be the largest economy in the world, and they also have much higher levels of trade with lots of countries that they can politically align with them, does that potentially lead to a broader decoupling, which would truly undermine all the progress you're talking about the last 50 years?

Steve Pinker:

It would push back against the progress. Would it undermine all of it? I think that's probably too extreme, simply because it's going to be in China's interest to sell stuff. It's going to be in our interest to buy cheap stuff. It's not the only thing on people's minds. They might sacrifice a bit of cheap stuff at Target for rallying around the flag and feeling that they're not being bullied by China. So I think that there's always a set of forces pushing and pulling. That's why even though there could be pushback, there could be a slowing down of progress, I don't think it's reason to predict that it will be reversed, that we'll go back to the 1950s or 1960s.

Ian Bremmer:

Now, one place that it does seem like we're going back to the 1950s and 1960s is the nuclear balance. Lots of nuclear agreements that are now being ripped up that are no longer being focused on, the Chinese are going from 400 warheads to 1500 warheads over the next 10 years. No nuclear agreements between the Americans and the Chinese, for example. I mean, the last time we had this level of nuclear concern was the Cuban Missile Crisis in 1962. How existential should that be seen as a risk? How much of a spanner does that place in your general concerns that, "Well, the backdrop of the world is more peace, less war and people are happier?"

Steve Pinker:

It is a significant concern. I think it ought to get more attention in our politics. The number of mentions of a nuclear threat in the recent presidential debates, including during the Democratic primaries, was very close to zero.

Ian Bremmer:

That's right.

Steve Pinker:

Considering that it is, to put it mildly, a significant threat, I think it's a scandal that it doesn't get more attention. And it is simultaneously possible that there are fewer and less destructive conventional wars, but there is the looming threat of a low probability but catastrophic nuclear war, and so I do think we should worry about it.

Ian Bremmer:

The world seems to be getting better, but the tail risks of cataclysm from climate change, from artificial intelligence, from genetic engineering, from nuclear weaponry are actually going way up.

Steve Pinker:

Climate change, it seems extremely likely that there'll be a lot of terrible consequences, but probably not existential ones. I think we should worry about bad consequences even if they're not existential, and a lot of the existential ones are hypothetical anyway. But I think the nuclear threat is unique because that's what nuclear weapons are designed to do. They're designed to kill people and blow stuff up, unlike other things that are more hypothetical. This is why-

Ian Bremmer:

It's a proven technology.

Steve Pinker:

... it's a proven technology.

Ian Bremmer:

And all we're doing is building more and taking the guardrails that we used to have away.

Steve Pinker:

I think the sheer number is less important than the mechanisms of strategic stability, of better detection, of possible attacks, of channels of communication that would prevent response to a miscommunication, misinterpretation, an emotional leader. So I think fewer is better, but a regime of communication and transparency that makes it less likely that they'll ever be used is what we ought to concentrate on.

Ian Bremmer:

Now, let's move to the topic du jour, which is artificial intelligence. And before I ask you what you think about AI, I'm really interested in what AI so far has taught you about human beings. What have you learned that is different about how people behave, about how the brain functions on the basis of what we've seen from AI over the last year?

Steve Pinker:

In AI and in my field of cognitive psychology, so AI is building machines that are intelligent and cognitive psychology is trying to figure out what makes humans so intelligent. There's a bit of tension between two different approaches. One of them is, it comes more out of logic and traditional computer science. That is, you've got statements, propositions, almost like English sentences. You've got rules for manipulating symbols almost like logical deduction, and you go from premises to conclusions. Then there's more the so-called neural networks that are loosely inspired on the structure of the brain where there are many units that each record a probabilistic bit of information about the world, not an actual fact, and the probabilities are aggregated millions, billions, now trillions of them. What we've seen is that the neural net approach is capable of much more than I think anyone would have thought possible. And so in the debate that I am embroiled in cognitive psychology is what makes us smart, you've got to take more seriously the possibility that by absorbing lots of statistical patterns, a kind of intelligence can emerge. It doesn't have to be step-by-step reasoning.

Now, I actually think that the human brain has a hybrid, and I tend to think, by extension that our best AI is going to have to involve a hybrid sometimes called neuro symbolic. That is both the neural network approach and some symbol manipulation. And we see that in the fact that the large language models like the GPT family, even though it's astonishing what they're capable of, I think our intuitions break down when you process a petabyte of information and store trillion statistical patterns. I would've guessed, it couldn't do something like generate coherent prose and answer questions sensibly, and it can. On the other hand, we also know that these systems-

Ian Bremmer:

Make mistakes.

Steve Pinker:

... they make stuff up. No competent lawyer would present a judge with five cases that are made up out of the blue as happened recently in a lawsuit involving Avianca Airlines, but GPT does that all the time, and that's because it doesn't have the concept of a thing, an object, an event, a fact, a person, a place.

Ian Bremmer:

No. It's a probabilistic formulation of what comes next after the statement or question that you just raised.

Steve Pinker:

Yes. It's a pastiche, a mélange, a collage, but there's nothing in there that was programmed to have the concept of a paper, a person, an event. Now, it's astonishing how it could make up for that lack by crunching those trillions or quadrillions of bits of information. But there are, I suspect, built in limits simply because of the way it's designed. And I think it's significant that if you look at what do we convey when we utter a sentence, what is the meaning of a sentence? What most cognitive scientists would say is, "Well, you decompose it into things and people, and actions and times and places," that's, I would argue, what makes us intelligent. These models achieve intelligence by a different route, and I suspect some of their shortcomings might be remedied by taking a leaf from the book of the human brain.

Ian Bremmer:

So that may make us intelligent, and I worry that not so much that AI is going to suddenly do all of these things that human beings could never do, but rather that it will be seen as really convincing and compelling by lots of other human beings. We're already at the place where you or I, or others engaged in a conversation with AI, are prepared to believe things that are just not so.

Steve Pinker:

Well, in fact, this has been known by AI researchers for decades that people are too easily fooled. That it doesn't take much to fool a user or an observer into attributing a lot of intelligence to the system that they're dealing with, even if it's rather stupid. Back in the '70s, the first chatbot, the predecessor of all these large language models just had a list of maybe 20 or 25 canned responses to questions. It simulated a therapist. So if you say, "Last night I dreamed that X," it would respond, "Have you ever wished that X?" This is not very bright.

Ian Bremmer:

How does that make you feel?

Steve Pinker:

How does it... Tell me more. Tell me more about your mother. People poured their heart out to the system. They didn't realize that they were dealing with 25 canned responses, so that is a problem in how we interpret it. It's one of the reasons why the so-called Turing Test from the great mathematician, Alan Turing-

Ian Bremmer:

Whether or not you can determine if something is a human being or a bot after 15 minutes of conversation.

Steve Pinker:

... exactly.

Ian Bremmer:

Yeah.

Steve Pinker:

Some people called it Turing's worst idea because it all depends on, how easy is to fool human? And the answer is pretty easy.

Ian Bremmer:

And the answer is, I mean, essentially the Turing Test has already either been passed or will soon be passed, right?

Steve Pinker:

Yeah. And it is just not a very good test any more than if you ask a scientist, "How do plants grow?" And you say, "Well, as soon as we can come up with a silk flower that someone can't distinguish from a real flower, then we'll know the answer." Well, no, that's just a completely separate question. So it's a red herring, I think. The question is how does it really work?

Ian Bremmer:

But it's really important if what you're trying to understand isn't the AI but society.

Steve Pinker:

Well, yes. You're right. And in fact, one of the dangers of the large language models is that people will vest undue confidence in them.

Ian Bremmer:

Now, what does that mean for you in terms of limitations that need to be placed on AI? For example, I mean, children whose very malleable social functioning on the basis of what they're exposed to. Should there be rules that prevent children from having relationships with AI bots? I mean, should there be impositions for anything that is AI to have, whether it's watermarking or some other form of, "This is a bot," and should the companies that have platforms be legally responsible for ensuring that AI is indeed so labeled? Because of course, none of that exists right now.

Steve Pinker:

Yeah, I think so, within reasonable formulation of policies that don't go too far and stifle innovation. But yes, certainly when it comes to children, there are all kinds of restrictions that we feel comfortable imposing that we don't on adults, so absolutely for children. And some kind of labeling of providence is going to be increasingly important, but at the same time, since it's never going to be perfect because liars and sociopaths will figure out ways of removing the watermark, then it will involve more trust in institutions that guarantee providence. So with say, the proliferation of deepfakes of Joe Biden saying something that he convincingly, seemingly caught on video saying something that in fact he never said that was just digitally formulated. It may be that we are going to have to put more trust in CNN and the BBC and the New York Times and institutions that do the due diligence, that they are not just propagating deepfakes.

Ian Bremmer:

At a time that the trust in those-

Steve Pinker:

[inaudible 00:26:36].

Ian Bremmer:

... organizations is at its historic lowest.

Steve Pinker:

Indeed.

Ian Bremmer:

So that's not credible. So given that that's not going to happen, what needs to be the response to ensure that democracies don't erode much faster than they have?

Steve Pinker:

Yeah. And I don't think we know the answer. I mean, it's tempting to say, "Well, let's just have more regulations," but regulations can often have side effects or be ineffectual. It may be a combination of things. It may be that this will actually increase trust in the media because ultimately, even though people can be fooled, they don't want to be lied to. They don't want to believe things that are false, so there is at least a force in the direction of people wanting some kind of veracity, and it may push them toward, if not legacy organizations, but some form of certification or credibility that films and texts are accurate. Maybe it'll be blockchain that will-

Ian Bremmer:

People don't want to believe things that are false. You're a psychologist. I want to pressure you on this because it seems to me that people are very happy believing things that are aligned with what they like irrespective of whether it's false or true.

Steve Pinker:

... so maybe a better way of putting it is that not everyone wants to believe in the same falsehoods, so you can't fool all the people all the time. And so even though there'll be people who will... In fact, all of us are all too receptive to claims that are consistent with our own ideology, our own politics, our own tribal loyalties. We're still skeptical of other people's sacred beliefs. And so in a democracy, in an ecosystem of debate, criticism, airing of ideas, even if some people believe falsehoods, but again, since not everyone has the same commitments to the same falsehoods, you do want the truth about at least the things that aren't essential to your core identity. Those are very hard to chip away at. Fortunately, we don't all have the same ones. So even though no person will necessarily gravitate to the truth, the ideal would be to set up rules of engagement of debate so that the entire nation, the entire community, while indulging people with all kinds of crazy beliefs, but that the greatest majority view, consensus view will be pushed in the direction of accuracy.

And so I think it's too easy... Seeing the negativity bias in the media and the fact that it has led to systematic misunderstanding, it just opens me more to things that happened in the past that went against the most pessimistic expectations that might still be operating. The fact that I think there's... And certain pessimistic habits that I think lead bias people in a negative direction, such as playing out in your imagination, the worst consequences of some new innovation, but not factoring into your mental scenario all of the countermeasures and pushbacks, those are easy to forget. And so I try to remind people of the countermeasures and pushbacks with no expectation that they'll prevail even that they'll succeed, just that they, in doing our planning for the future, we shouldn't neglect them. Something bad happens. What we tend to forget is people will react and we'll try to mitigate that damage.

Ian Bremmer:

Do we find talking about just how human beings interact with each other, general engagement society, what's changed as globalization has occurred? What's changed as new technologies have emerged? What's changed as people have been more impacted by algorithms as opposed to nature and nurture? How do we see human beings as human beings becoming different?

Steve Pinker:

In the various decision-making elites in universities, in companies, I think there's less of a focus on the local national culture. There used to be, just in my field for example, people used to read things like the Canadian Journal of Psychology and the British Journal of Psychology. Those are now marginal because why would you care about what's simply happening in your own country? You used to when you'd have to take the train to a conference that was only a couple of hours away, now you can step on a plane or read about it on the internet. The national borders are irrelevant, at least for that kind of information elite, maybe not for the population as a whole. But in terms of decision making, you look for the best ideas wherever you can find them.

Ian Bremmer:

Steve Pinker, thanks so much for joining us today.

Steve Pinker:

Thanks for having me.

Ian Bremmer:

That's it for today's edition of the GZERO World Podcast. Do you like what you heard? Of course, you did. Why don't you check us out at gzeromedia.com and take a moment to sign up for our newsletter? It's called GZERO Daily.

Announcer:

The GZERO World Podcast is brought to you by our lead sponsor, Prologis. Prologis helps businesses across the globe scale their supply chains with an expansive portfolio of logistics real estate, and the only end-to-end solutions platform addressing the critical initiatives of global logistics today. Learn more at prologis.com.

Subscribe to the GZERO World Podcast on Apple Podcasts, Spotify, Stitcher, or your preferred podcast platform, to receive new episodes as soon as they're published.