Yuval Noah Harari and the power of storytelling.

Yuval Noah Harari can be liked or disliked. As I like him as a historian and philosopher of history, not necessarily as the "pop philosopher" he sometimes becomes. But it's hard not to respect him as one of the sharpest thinkers of modern times.

When I saw Harari on Lex Fridman's podcast, of which I am a faithful viewer, it made me jump. Much more so than when Lex interviewed Musk or Zuckerberg.

But what does a historian and philisopher of history have to do with technology? Well, history is made in a huge way by the development of technology. Not exclusively, because like Harari I'm not a Marxist, but to a large extent certainly. Seemingly obvious, and somehow it doesn't break through so much in our awareness.

I catch myself constantly writing about AI in my short columns here. So it must be about AI too, so here it is. Because Fridman Harari's talk is also about it.?

But let's start with what Yuval Noah Harari thinks what intelligence is versus what consciousness is. We hear about these two concepts in the context of AI practically all the time lately. Because we are naturally afraid of the unknown, and the unknown is large language models. And if this unknown is intelligent and, moreover, as one former Google engineer stated, sentient, then isn't it able to dominate us and, in the extreme version, completely subjugate us.

In Harari's terms, intelligence is the ability to achieve goals independently. In this sense, ChatGPT is intelligent to at least some degree. Consciousness, on the other hand, is the ability to have feelings. It is difficult to imagine, in this view, that a computer would ever become conscious. Although Harari does not rule it out-I do.

So, if computers will (probably) never become conscious, is there anything to fear from generative algorithms?

According to Harari, definitely yes. What has made us, as humans, the dominant species on planet earth is our ability to create and tell stories. Our closest cousin the chimpanzee is also intelligent and in many areas surpasses us. We wouldn't want to go to a boxing sparring match with a chimpanzee and, being on a desert island, it's more likely our cousin will climb a tree faster and has a much better chance of survival. However, chimpanzees can't create history and thus abstract concepts that unite them into communities as efficiently as human communities.

For what are concepts like money, nation and society if not stories? We believe in the digital record of money, even though it does not give us food, because there is a magic behind it, that we can always buy food with this record. This is even more evident in the example of cryptocurrencies. For aren't these purely the effects of human-created story?

Generative AI, whether we consider it intelligent or not, regardless also of the quandary of whether it is capable of acquiring consciousness or not, is already capable of generating stories on its own. And this, according to Harari, poses the greatest threat to humanity. I fully agree with this, unfortunately.

Marxists believe that history is created by the forces of economics. But how then to explain the phenomenon of Adolf Hitler? This one of the greatest criminals in history came to power because he was an excellent storyteller. He told his citizens what they wanted to hear: that their rapid impoverishment was the result of betrayal and they were great, but everyone had betrayed them. The gigantic genocide that is the Shoah is also based on a myth invented largely by Hitler himself. Also, Mao, a monstrous genocidaire, was an excellent storyteller. When he placated Western journalists he posed as a charming intellectual, when he spoke to a crowd he sharpened his peasant accent.

Stalin, at the time of the Battle of Moscow, where he was on the verge of losing, gave a famous speech that was completely stripped of communist mythology. Instead, it referred strongly to much older myths like the nation, the binding of the people to the land, the idea of mother Russia.

History shows that a psychopathic and murderous individual can create it if he is an excellent storyteller.

Many military experts are far more concerned about fake news created by artificial intelligence than autonomous weapons. The substitution by Ukrainian forces of fake news imitating Putin, where he announces the seeming imposition of martial law in border areas, enabling mass mobilization has caused hysteria among civilian Russians.

At critical moments, war is won with fakes. This is evidenced by the examples of fake Allied landings during WW2, in particular the fake message found by a dead soldier on the beach about the landing in Sicily and the construction of fake tanks around Dover posing as D-Day near Calais.

According to the experts I'm talking to (I'll be publishing an interesting video segment on the subject soon), we're still a long way from developing algorithms to block fake stories.

As Harari rightly says, humans are designed in such a way that we want somebody to listen to us : "your spouse, your mother, your politicians". Is it not the case that being listened to by AI is enough for him? AI, unlike even our loved ones, focuses its attention solely on us. And here's the quote that stuck in my memory the most: "we used to create machines catching our attention, now we create machines to catch our intimacy."

How humanity reacts to technology? According to Harari we are not good at it. The learning curve of industrial revolution was quite scary. On the way we invented colonialism, nazism, communism. It all costs lives.

In the case of the AI revolution, we can't let that happen.