According to the Israeli historian, Ph.D. Yuval Noah Harari, the neural network ChatGPT is to some extent more dangerous than a nuclear bomb. This statement was made by the author of the book “Sapiens: A Brief History of Mankind” in an article for The Economist.

A recognized expert in the field of military history noted that the neural network will not create more powerful atomic weapons on its own, but will produce smarter artificial intelligence.

At the same time as benefiting humanity, for example in the field of finding a cure and solving the environmental crisis, artificial intelligence poses a huge threat to civilization. It is about his ability to independently create unique texts, music and images.

The professor argues that almost the entire culture of mankind is built on language and is a collection of man-made artifacts. Thus, a neural network that has learned to generate human-like texts is actually “hacking the operating system of our civilization,” Harari states.

Recall that the confrontation in the development of neural networks in China and the United States can already be compared with an arms race.