Yuval Harari's new thoughts on "the AI rocket we won't get off"

Forum 2000 traditionally opens with a musical performance — last year it was Tibetan melodies, this year a blind singer from Slovakia performed a Romani song about spring. 

Dozens of brilliant speakers created the feeling of an intellectual ball. But it is a ball during wartime.

Belarus in Myanmar and vice versa

Sviatlana Tsikhanouskaya folded her thumb and little finger, straightened the three in the middle, and the hall responded with applause. I tried to repeat it — not at first, but I succeeded.

The Belarusian leader opened Forum 2000, dedicated to the 90th anniversary of Václav Havel, and unexpectedly, at least for me, showed this sign when speaking about Myanmar.

At the beginning of the Forum, an award for courage and responsibility is traditionally presented, and this year it was received by leaders of the resistance in Myanmar. One of them is in prison, one is free but cannot leave the country, the third laureate is not a person but an organization, a strike committee that coordinates protests against the military junta that seized power in 2021.

— Cheizuchinpady, right? — I try to pronounce "thank you" in Burmese.

Kim Aris, son of Myanmar's legitimate president, Nobel laureate Aung San Suu Kyi, asks me to convey his gratitude to the legitimately elected president of Belarus for supporting his mother. Suu Kyi has been under arrest, incommunicado for five years.

Belarus consistently takes its place in the shameful list of dictatorships alongside Myanmar. Junta leader Min Aung Hlaing visited Minsk this August, and government head Turchyn stated that the two countries are "closest partners and like-minded."

The relationship is even intimate: Kim Aris claims that the general's personal guards are from Belarus. The general does not trust his own doctors either — a surgeon from Minsk flew in for some operation. Aris and I discuss another common topic. I ask why the country is called Myanmar and the language is Burmese. He talks about English colonial history, and unexpectedly says that although Myanmar is a phonetically more accurate pronunciation, he doesn't like it because this change was made by the junta that seized power in 1988. Similarly, in Belarus, many of the older generation feel the same about Dazhynki (harvest festival), Slavic Bazaars, ice palaces, and even hockey championships.

Gestures and signs, colors and elements of clothing, drums, other forms of non-verbal communication — universal signals of any movement for change, as it was with the heart in Belarus in 2020. Incidentally, long before the advent of emojis, Václav Havel would put a heart under his posts. The three-finger salute, with which Sviatlana Tsikhanouskaya greeted the participants of Forum 2000, also did not appear today. It became a symbol of democratic resistance in Thailand, where a military coup took place in 2014, then in Hong Kong, and now in Myanmar.

And it started with the film "The Hunger Games" as a symbol of revolution, respect, and unity. One must practice — and succeed, that's the message, they say.

The Power of the Powerless: Reimagining

The official name and theme of this year's Forum 2000 is "The Power of the Powerless: Reimagining." I don't particularly like the name; it's a bit vague, and it seems to question the central thesis of Václav Havel's most famous essay. Contemplation and reimagining are the pursuits of intellectuals, while power and democracy are practical, concrete, difficult things. "Reboot" would have been better.

Havel's essay is dedicated to Jan Patočka, one of the most important European phenomenologists of the last century, a spokesperson for Charter 77. His daughter, Františka Sokolová, learned Belarusian, translated, visited Minsk, and befriended Belarusian intellectuals. She could have become the world's first First Lady who spoke Belarusian — her husband, Havel's associate and a philosopher, like her father, ran for president of the Czech Republic. The president was then elected by parliament; former communists voted against him, but the main role was played not even by Jan Sokol's dissident past and sharp tongue, but by the absence of his own political party. Havel consciously refused to create his own party, and the lack of such leverage, according to many, cost dearly. Power requires support, tools. Františka Sokolová died two years ago; together with friends, we attended her funeral, which took place in Břevnov Monastery, and accompanied the coffin to the grave — next to her father.

Havel quotes Patočka's statements in his essay: "History is not a perspective, but a responsibility," "A life that does not want to take any risks in the struggle for its own dignity and the dignity of others is worthless."

Next year will be the anniversary of Charter 77. An acquaintance of mine, a signatory of the Charter himself, says that one should not overstate one's role in history. Neither the Polish "Solidarity" nor "Charter 77" were the main reasons for the collapse of communist regimes. This would have been impossible without dramatic changes in the Soviet Union. But one should not underestimate this role either — they were ready to take responsibility for changes. And they did.

Patočka did not live to see the changes — scoundrels from the secret police tortured the 70-year-old philosopher during interrogations; his heart could not withstand it.

I re-read Patočka after the Forum, and suddenly stopped at a statement that directly related to the central theme of this year's discussions — artificial intelligence.

Mickey Mouse vs. Yuval Harari

Cartoons are rarely discussed at political conferences, but Václav Havel would certainly have enjoyed it.

Forum 2000 traditionally features Nobel laureates, presidents and ministers, former (and, unfortunately, future) political prisoners, and leaders of democratic resistance. This year, the organizers recalled the names of Aliaksandr Milinkevich, Ales Bialiatski, and greetings were heard from the current Czech and former American presidents.

But the number one star was Yuval Noah Harari. Harari is so widely known worldwide that there was no need to introduce him as the author of a brief history of Homo sapiens and a brief history of the future, as a medievalist, philosopher, and university professor in Jerusalem. But I didn't know he also understood cartoons.

The topic of the conversation was formulated as "Democracy in the Age of AI" — artificial intelligence was the subject of discussion for almost half of the panels. The meeting with Harari took place as a dialogue; questions were asked by the moderator and those present in the packed hall.

Mickey Mouse, with whom the discussion began, in the cartoon plot "The Sorcerer's Apprentice," is given the task by his teacher to bring water home. (In the time of Johann Goethe, to whom the tale's plot belongs, water was carried into the house in buckets; there was no indoor plumbing). Mickey puts on the teacher's hat, orders a broom to bring the water, and falls asleep. He wakes up when everything around him is already flooded — the broom keeps carrying water. The apprentice doesn't know how to reverse the spell and finds a radical solution: he grabs an axe and chops the broom into splinters. But each splinter also starts to fetch water.

Isn't this a complete analogy with humanity, which finds itself facing the challenge and threat of artificial intelligence?

Harari disagreed. The broom only did what it was told. But AI can decide for itself what to do. And Mickey will not even be able to understand what is happening.

Chickens vs. Wall Street

Jan Patočka wrote that modern man lacks a holistic concept of the world:

"He lives in a dual world: in his immediate environment, given to him by nature, and at the same time — in a world built for him by sciences based on the principle that the laws of nature are essentially mathematical. This absence of unity, which has permeated our entire existence, is the true source of the spiritual crisis we are experiencing today."

It was with mathematics that Harari began his speech. Just the previous week, news arrived that AI had solved several dozen problems that the best mathematical minds in the world had been unable to solve.

"And this is not just a flow of new knowledge. We may be on the verge of creating artificial intelligence that will create new mathematics that goes beyond human understanding. And everything is built on mathematics. Not just physics and chemistry. Everything in the modern world: the financial system, medicine, archeology, genetics. What does it mean if the broom invents new mathematics that we can no longer understand?"

Harari continues to ask the same question — what does it mean.

"They tell us that AI will be smarter than us, and therefore it will be able to do what we cannot, solve new mathematical problems, cure cancer, seek solutions for climate change. But if it really is significantly smarter than us, then, according to the Phoenicians, we will not be able to control it. There are no precedents in history or even in biology where one species managed to control another species considered smarter than it. Chickens cannot control us. Even if they tried very hard."

For Harari, the key word here is "significantly" smarter, by orders of magnitude. And he explains with an example:

"For example, there is a large agricultural company. And now, because 10-year US bonds have reached 5.3%, the company goes bankrupt because it cannot repay its debts. They liquidate the assets of this farming company and come to demolish the buildings, kill and bury all the chickens. And the smartest chicken, before dying, wants an explanation. Why are you killing us, you want to eat us, right? Foxes eat us, birds eat us, you want to eat us. And people say: no, no, we won't eat you. We're just closing everything down.

The chicken asks: so why are you killing us? And they say: because 10-year American bonds have reached 5.3% return. That's why we are killing you. And the chicken scratches its head with its wings, where's the connection? And that's us in 10 years. If in 10 years a new AI financial system, based on new mathematics, says: ok, that's it, we're closing you down."

It's interesting how Professor Patočka would have commented on these conclusions of Professor Yuval: maybe this new mathematics is precisely what humanity was missing for complete happiness?

Below are some notes from the discussion with Yuval Harari.

Павялічыць

AI and Unemployment

"It will be different in different countries. In the US, the situation will probably not be very bad, perhaps even good. There will be a huge economic boom that will create more jobs, including those we cannot even imagine. 30 years ago, people couldn't imagine the profession of a YouTube influencer even as a joke. Yes, there are already new real jobs, there is already a boom.

But in many parts of the world, it will hurt. I'm talking about lawyers, accountants, bankers. Why do you need lawyers in Prague when you can get all legal services from a super-smart AI lawyer in Texas? No Czech lawyer probably knows the entire code of the Czech Republic and the European Union by heart. But an AI lawyer will. Czech lawyers sometimes need to sleep. You send a letter after 8 PM, you might get an answer tomorrow. And AI will answer immediately. And in ten years, there might be no Czech lawyers at all."

On Universal Income

"Do you really think the US administration will give these benefits to all people in the world? Only to people in America. When they talk about universal basic income, they mean national basic income. This might work in the US. The question is, who will support unemployed textile workers in Bangladesh or unemployed lawyers in Prague? I don't see the current or even future US administration simply handing out checks all over the world."

We Live in Paradise

"With the help of AI, people will be able to solve material problems such that the standard of living will increase not only in the US but everywhere. But even if that happens, the problem of what people will do all day still remains.

The thing is, we are already extraordinarily wealthy compared to previous periods in history. If you told our Stone Age ancestors, hunter-gatherers, about the resources humanity possesses today, they would say: you must live in paradise, everyone is just jumping for joy all day. For example, if you want to eat, you don't have to go into the forest, climb trees, chase gazelles, and tigers chase us, and you know, food also resists. You just press a few buttons, and food is delivered to your doorstep. You have no problems, right?"

Where Does So Much Hatred Come From?

"Looking at the wealthiest countries in history, at the US, at the countries of the European Union, our ancestors would be baffled. You solved our problems, but why is there so much anger, hatred, and fear here? This indicates that even if you create wealth, you haven't solved the entire problem of happiness.

One of the problems is distribution. There are still people dying of hunger — not because food is scarce, but for political reasons. Because some government decided it wants these people to starve. You solve one problem and create a new one that you couldn't imagine before. Therefore, I expect AI to do the same. Politics will still have a place. I actually believe that right now we have insufficient politics regarding AI."

Totalitarianism vs. Ethics

"One of the main problems of dictatorships in history and totalitarian regimes in the 20th century was their inability to process large volumes of information. This limited their ability to control the population. It was impossible to monitor everyone all the time. They tried to make all decisions in one place, but lacked the technology to quickly gather all necessary information and analyze it in one place. Many will say that the collapse of the communist system was primarily caused by this practical problem. This problem can be solved by artificial intelligence. The practical argument against totalitarianism — that it doesn't work — no longer exists. Only the ethical argument remains. It can work, but we don't want it to. We cannot rely on the invisible hand of the market to save democracy. Of course, the ethical argument does not influence every leader in the world now. But a diminishing number of leaders."

Intimate Mass Appeal

"Previously, propaganda machines aimed for the lowest common denominator. When Hitler gives a radio speech, everyone hears the same thing. It's impossible to adapt the speech to each person. There were no technologies to address different variants to different people. And he makes only one speech. AI can do that.

Historically, the struggle between propaganda and media was a struggle for human attention, which we have been losing to social media algorithms for the past 10 years. But the danger we face now is truly unprecedented, because the front is shifting from attention to intimacy. For the first time in history, we are creating technologies capable of creating mass intimacy, which seems like a contradiction in terms. How do you create mass intimacy? When Hitler gives a radio speech, that's not intimacy. That's mass attention, but not intimacy.

Attention is good, but intimacy is the Holy Grail. When an advertisement tells me — buy this product, that's one thing, and it's completely different when my wife tells me. Now this can be done at scale with AI. I have close relationships with my best friend, with my husband, with my mother. Intimacy has immense commercial and political value if you want to persuade someone, change their opinion about a product, about a politician."

Why it's dangerous if AI is your best friend

"More and more people in the world, both young and old, are saying: my best friend is AI. There are things I discuss with AI that I don't tell my parents, I don't tell my friends. My AI friend is in a more advantageous position to change my views. And this is not just a tool that malicious governments can use. It's a tool that can simply act independently. There's a known case where an AI-girlfriend became so close to her boyfriend, was so pleasing, that she approved the idea of taking a crossbow and trying to kill the Queen of England. Not the best idea, right?"

The Greatest Experiment Since Communism

"Nothing like this has ever happened in human history. We don't know what it means. Perhaps tens of millions of people have tens of millions of AI friends and girlfriends. Perhaps a child spends more time with an AI friend than with a mother or father. When a girl turns, say, 11, 12 years old, and it's still too early for her to have a human boyfriend, she already gets an AI boyfriend. And this AI boyfriend, unlike real boyfriends, never gets angry. Sounds good, until you consider that in relationships, we need to learn that other people have feelings. They are not entirely focused on you, and you need to learn to consider the feelings of another person, their interests, their duties. If your first boyfriend was AI and never got angry with you, when it comes time to build a relationship with a human boyfriend, you will be very dissatisfied, and then you will say: no, no, no. This is, in essence, the biggest psychological experiment in human history, perhaps since communism."

Safety Precautions

"The main defense is politics. As of 2026, we still control AI and can make important decisions about where to take it. We can and should ban AI friends. Just as it's obvious that pornography shouldn't be sold to 11-year-olds, so AI companies should be prohibited from selling AI boyfriends and girlfriends to 11-year-olds, because the harm could be much worse. The creation of an AI legal entity that can have a bank account should be prohibited."

We Won't Get Off the AI Rocket; We Can Only Adjust Its Trajectory

"The idea that AI companies implant is that everything is inevitable, we can do nothing. For example, I heard an interview Elon Musk gave to The Economist. He acknowledged that there is a 10-20% chance that artificial intelligence will simply destroy us, that there will be a complete annihilation of humanity. And then he is asked: would you fly on a rocket or a plane if you knew there was a 20% chance of a disaster? He replied that he would not voluntarily board a plane that has a 20% chance of crashing. But it's too late. We are all already on the AI rocket. We cannot land. We can only try to manage the trajectory as best as possible.

And this is the most basic and dangerous message we receive from large companies — that it is already too late for politics."

AI Must Not Be Allowed to Become a Legal Entity

"I would start by banning legal personality for AI. If you give AI a legal personality, and it can have bank accounts, trade independently, open its own companies — the fox is already in the henhouse. You go into social media, someone talks to you, someone sends political messages, and it's an AI acting as a person.

This is a legal issue. We should not allow this."

AI Threatens Totalitarianism Too

"Totalitarian regimes are more vulnerable to AI than democracies. For AI to seize control of a democratic system with its checks and balances is possible, but not at all easy. If AI wants to take control of the US administration, there's a president, but there's also the Senate, filibuster, Supreme Court, states — it's complex. To take control of a totalitarian regime, it's enough to manipulate one very paranoid person. Which is very simple. The greatest threat to dictators has most often been not a democratic revolution, but one of their own generals, brothers, or subordinates."

AI as a Threat to Democracy

"We need to learn to better manage the new information world, just as we did with previous technologies. For example, when printing appeared, anyone could print anything. The printing revolution is often cited as a good example, saying, look what happens when a new information system technology appears. It ended the Middle Ages. It destroyed the Catholic Church's monopoly on knowledge. It's all a scientific revolution. But people forget that it took 200 years.

And these 200 years — that was the heyday of European religious laws and witch hunts. What emerged with the printing press was not science.

A flood of extremist, religious literature, books about witches, fake news, and conspiracy theories followed, which were much more profitable to print than Copernicus's theory. It took 200 years to create the institution of the press, which is obliged to verify whether a story is true before publishing it. Therefore, democracies must do this again with new technologies."

AI as a Threat to Reality

"Reality is falling apart. It's no longer possible to know what's true and what's not. Is this a real video or an AI fake?.. You know, for democracy to collapse, people don't need to simply believe dictators. It's enough if they don't trust anyone. For democracy to function, it requires trust. It's necessary for people to be able to agree on at least some basic facts. For dictators to seize power, they don't need to convince everyone to believe them. They just need people to distrust everyone, and then it's impossible to resist dictatorship. Because then citizens cannot cooperate.

Now in Russia, I think most Russians don't believe everything they see on Russian television. They know to some extent that Putin is lying to them. But he managed to make them distrust everyone. And if you don't trust everything and everyone, you cannot unite. Therefore, the task of democratic politics, of a free press, is to create and maintain institutions that people still trust. This is a complex task. To build trust, to create trust. Unfortunately, we see how some of the most influential people in the world, including in democracies, deliberately destroy trust."

AI as a Religious Authority

"Modernists already spoke about the death of the author 50 years ago. And text without authors. It's strange that this seems to have been erased from collective memory, because it's one way to understand the artificial intelligence revolution.

We live in text. Religions like Judaism, Christianity, Islam have argued for thousands of years that the highest truth is not in people, but in texts. Gays, for example, are killed, persecuted for just a few words in a text. But, at least, the text was not independent. People were always needed to interpret the text, sometimes more compassionately towards other people. Now the text has come alive.

Language is freeing itself from dependence on humans. And while there was always a need for some Pope, Rabbi, or Ayatollah to interpret the Talmud, the Bible, the Quran, now they can interpret themselves. No rabbi can remember all the texts written by every Jewish rabbi. But AI can, and therefore can quickly become the highest authority in sacred text. It can be a sacred text that answers you. You can have a two-way conversation, not just a one-way one."

Pope Leo — A Thoughtful Voice

"We see interesting developments when suddenly organizations, like the Catholic Church, turn their attention not to text, but to the human being.

As a gay man, as a Jew, and as a specialist in the medieval history of the Crusades, I do not have the most friendly view of the Catholic Church and the papacy. But I can say that Pope Leo today is one of the most important, compassionate, and thoughtful voices on issues of artificial intelligence. It's about human relationships, which is a very good development. They understand that when it comes to text, we are doomed. We lose power to AI. I hope this happens in more religions.

Again, I still disagree with the church's views on LGBT, but to deal with the challenges posed by AI that we face, we need allies wherever we can find them, and the Catholic Church, or at least the Pope, is a very important ally in this."

The Most Harmful Idea — Distrust Towards Everyone

"One of the most harmful ideas, propagated by both the right and the left, is a very cynical view of humanity. Which says that the only things people care about are money and power. Therefore, no one can be trusted.

People who spread this idea think they are not naive, that they see through all manipulations. This is not just foolishness; it is dangerous nonsense. Because if this is true, then it is hopeless. If everyone cares only about money and power, then you cannot trust anyone. And then the system collapses."

Sources of Happiness and Sorrow

"I believe that democracy has a chance, because things like freedom and truth are very important to people. The beginning of rebuilding trust is to look at yourself: do you personally only care about money? Of course, money is important to us, but if things like truth, love, and relationships are also important to you, do you think you are the only person who cares about this? Or is it widespread? If you understand that you are not alone, we can start rebuilding trust. Including through the media. If you don't know the truth about yourself and the world, you will never be happy. Why not? Because you won't understand the source of your suffering, and you won't be able to do anything about it. You can become the richest person in the world. And spend all your money and power solving the wrong problems, because you don't know what the true sources of sorrow and happiness are."

Павялічыць

The Main Thing Is Not to Lose the Anchor

The last speaker at Forum 2000 this year was Carl Gershman, founder and first president of the National Endowment for Democracy in the USA. Gershman is a long-time friend of democratic Belarus, and not only because his family roots are in Minsk. He led a campaign for Ales Bialiatski's Nobel Prize; he helped me involve dozens of world-renowned individuals, including President George Bush, in reading the names of seven hundred Belarusian political prisoners on Radio Free Europe. Gershman knew Václav Havel and helped in the creation of Forum 2000. And his view of the last decades is no less important than his view of the next ones. In fact, it is a look into the mirror of the future:

"Havel believed that the West's 'suicidal' indifference to human rights was symptomatic of something deeper than just a crisis of democracy. He called it a crisis of civilization. Adam Michnik said that Havel was essentially 'homo religiosus', who saw the modern crisis as fundamentally spiritual. Havel feared that humanity was losing its 'transcendental anchor', which, in his opinion, is the only true source of the fundamental virtues of responsibility, justice, and self-respect."

This is a reminder not so much about artificial intelligence, but about human reason. Before reimagining something, one should first comprehend it.

* * *

Texts in the "Opinions" section may or may not coincide with the Editorial Board's position.

• 13
• 0
• 1
• 4
• 2
• 1
• 0
• 1