A robot gets lessons on how to recognize and manipulate objects. As artifical intelligence improves, a new kind of arms race is likely, involving killer robots, says the new book 21 Lessons for the 21st Century.
Will Smart Robots Fight Our Wars and Take Our Jobs? Maybe.
A rise in global nationalism combined with advances in technology signal trouble ahead, says the author of a new book. But there's hope.
Yuval Noah Harari’s last book, Sapiens, was a global bestseller. In his new book 21 Lessons for the 21st Century, one of the world’s most exciting young thinkers turns from the past to the present—and the future. What will life be like in 50 or 100 years? Will the liberal democratic order that has underpinned western societies survive? Will artificial intelligence (AI) and biotech render us economically redundant? Will digital dictatorships control our lives and thoughts?
When National Geographic caught up with Harari by phone in Tel Aviv, he explained why nationalism cannot solve the global issues facing us, why AI could put millions out of work, and how meditation gave him insights into his own mind that science could not.
The big idea you explore at the beginning of your book is that “the merger of InfoTech and biotech might soon push billions of humans out of their jobs and undermine both liberty and equality.” Can you tease out the details for us?
First, I try to emphasize that what we are facing is not just a revolution in information technology, artificial intelligence, and machine learning, which get most of the attention these days. We are also deciphering the secrets of the human body and brain, how we think and how we behave and why we do it. And we’re getting more and more computing power and better and better artificial intelligence. The combination of the two changes everything because if humans are, in principle, hackable and decipherable animals, and we have a better and better understanding of how this animal functions, and we have the necessary computing power to make sense of all the data we are gathering about ourselves, then what you get is algorithms that can potentially understand you better than you understand yourself. When you take all this mixture together, you get cooperation that can understand our desires better, governmental politicians that can manipulate our choices better, and more and more jobs that computers can perform better than human beings.
Another scary prediction is that “Big Data algorithms might create digital dictatorships in which all power is concentrated in the hands of a tiny elite.” Again, can you break that idea down for us?
When you look at the 20th century and the struggle between democracies and dictatorships, we tend to think about the struggle in terms of different ethical ideas. But it was also a struggle between different systems for processing information and making decisions. Democracy is a system that distributes information and power between many individuals and institutions. Dictatorship works by concentrating all the information, decisions, and power in one place. Given the technological realities of the 20th century, dictatorships were simply less efficient. One of the main reasons why, for example, the U.S. defeated the Soviet Union in the Cold War is that the distributed system of the U.S. just took better economic and political decisions than the concentrated system of the Soviet Union.
But some people have this overly optimistic view that there is some law of nature that says that under all conditions and circumstances distributed systems work better than centralized systems; therefore democracy will always be more efficient than dictatorship and will always defeat it. But this is unwarranted. When you look at history, the development of technology and economics and politics, what you see is a pendulum that swings back and forth. Sometimes distributed systems have an advantage and sometimes centralized systems have an advantage.
In the 21st century we may be entering an era in which centralized systems, again, work better than distributed systems because of the power of machine learning, big data algorithms, and artificial intelligence. What the Soviets couldn’t do in 1960 will become a possibility in 2030, when you have good enough algorithms and enough data. It’s not a certainty. But we should beware of the rise of digital dictatorships.
Among the most provocative statements you make is, “When a thousand people believe some made-up story for a month—that’s fake news. When a billion people believe it for a thousand years—that’s a religion.” A lot of people will be offended by that statement. Is there no place for religion in today’s world?
[Laughs] No, there is certainly a place! It’s very difficult to get large numbers of people to cooperate effectively without mythology. Whether this mythology is about some God, a nation, or some ideological system, without fictional stories it’s almost impossible to arrange large-scale cooperation. So, I’m not denying the effectiveness or even benevolence of religion, I am just denying the veracity. It’s a very different issue.
Almost all religious people will accept the claim that religion is based on fictional stories, with one simple caveat: all religions except mine are fake news. If you ask a Jew, then the Jew will tell you, “Yes Judaism is the truth, but Christianity? Jesus was not the Son of God and did not rise from the dead! This is fake news.” Then you ask a Christian and he or she will say, “No, no, no, that’s the truth. But all these stories that Mohammed had a revelation from the Angel Gabriel and the Koran is the Word of God, this is completely fake news!” Then you ask a Muslim …
You call nationalism, or “the worship of nation,” a dangerous myth. Explain your thinking here and why you believe we need “a new global identity.”
Nationalism is not an eternal part of human psychology or society. Lots of people think that it’s somehow imprinted in our genes, but humans have been around for more than two million years and for about 99 percent of that time, they did not live in nations. They lived in very small communities, clans, families, and tribes, whose chief characteristic was that everybody knew everybody else. The miracle of nationalism, which appeared only within the last few thousand years, is the ability to make millions of strangers who don’t know each other nevertheless cooperate effectively and care about one another. As such, it has been a very good development.
But in the 21st century, we are bumping up against the limits of nationalism, because the main problems we face now, as opposed to 1,000 or 100 years ago, are global problems. Our three biggest problems are nuclear war, climate change, and technological disruption, and you cannot solve any of these problems on a national basis. If you’re afraid, let’s say, of genetic engineering of humans, and the implications of that, you can ban genetic engineering in the U.S., but if the Chinese are engineering super-humans, very soon you will be tempted to break your own ban because you wouldn’t like to stay behind. If we think that technology, like in genetic engineering, or AI needs to be regulated to prevent the worst outcomes, we must have a global agreement. At present, largely because of the rise of nationalism, we are going in the opposite direction.
Over the last two to three years there has begun a very serious and accelerating arms race in AI. And if we continue with this, all the world’s nightmares about AI are going to be realized. Every country will think, “We don’t want to develop, say, killer robots, we are good people. But the others are unscrupulous and evil people and we can’t allow them to have an advantage on us, so we must do it first.” This is the logic of the arms race.
At the end of the book, you speak personally about the effect that meditation has had on your life. You even claim it can offer insights that neuroscience cannot. Can you talk about that?
When I was doing my PhD at Oxford in history, a good friend recommended I try meditation. I was very dissatisfied with my life and I had no idea where to turn. So, after about a year, I said, “Why not?” At the time I was 24 and thought I was a very intelligent, well-educated person and in control of my life. Ten days of meditation revealed to me that I had almost no understanding of my mind or control over it. Thoughts and desires just pop up without any command from me and I have a very limited understanding of why and how.
Science gave me a lot of very powerful methods to deconstruct the dominant fictions that shape the world and tell you how to behave, whether it’s religion, economic, or political systems. But it didn’t offer any real solutions to the problem of misery and suffering in life. What meditation gave me is, first, the understanding that much of the suffering in life is generated by the mind. You can change the whole world around you, but if you don’t change your mind you can be in the best situation in the world with enough food and loving people around you, and still be miserable. Realizing this was a very important step both to understanding myself and also to living a much happier life.
We have a lot of powerful tools to understand the brain, but the only mind we can access directly is our own mind. I can put you in an MRI scanner and directly observe what is happening in your brain and see things you don’t know about, like which neurons are firing or which biochemical processes are happening in which part of the brain. But I cannot see your subjective experiences or have direct access to your pain or your love. If I want to explore the subjective experience of love the only subjective experience I have access to is my own. The idea behind the meditation I practice, Vipassana, is to have a methodical system for observing the subjective experiences that are happening in the mind.
Having read your book, many people might assume that you are a pessimist about the future of mankind. Yet you actually believe we can rise to the challenges we face. How can we do that, Yuval? Are you a pessimist?
I would summarize my position in three simple statements: I think the world of things is today better than ever before in history and this disqualifies me from being a pessimist. Second, things are quite bad. Yes, it’s better than ever before but there are a lot of problems still around. Thirdly, things can get much, much, much worse.
That doesn’t mean they will inevitably get worse. Things today are better not because of some divine miracle, but because of wise human action. We have managed to a large extent to reduce the problem of famines, epidemics, and plagues, and even of violence in the world. We are living in the most peaceful era in history. This is because of things we did. This implies that we have the power to prevent the worst outcomes and to make things better but it’s on us! If we make wrong decisions, nobody will come to save us.
A good place to start is a combination of realizing the immense power that humankind now has and being humble about our understanding and knowledge. There is a mismatch between the immense power of manipulation that humans now have over ecological systems and even our own bodies, and the limited understanding we have of these systems. In the ecological system, the result is something like climate change. We know how to cut down forests, how to extract oil and burn it, but we have very little understanding of the impact this has on the climate and ecological systems.
The same thing could happen with our ability to change things on the level of the body and the brain. To prevent the worst outcomes, we should combine and deepen understanding of our power, which would make us much more responsible, with a humble appreciation of the limits of our understanding, which should make us far more careful about the kind of manipulations we carry out in the world.
This interview was edited for length and clarity.
Simon Worrall curates Book Talk. Follow him on Twitter or at simonworrallauthor.com.