One of the creators of modern artificial intelligence warned of a possible "Chernobyl" due to neural networks.
As early as 2026 or 2027, AI could cause large-scale disruption in the energy or financial system. This scenario was proposed by Geoffrey Hinton, who is known as one of the "godfathers" of artificial intelligence, in an interview with The Guardian.
The scientist compared the possible incident with Chernobyl, however, we are not talking about a nuclear disaster, but, for example, a massive power outage or stopping the financial system for a whole week. According to Hinton, such an incident could force world leaders to think more seriously about the threats of technology.
At the same time, the researcher does not rule out that artificial intelligence will give humanity a wonderful future, but there is another scenario: mass unemployment, growing inequality and serious social upheaval. How successfully people can deal with risks, he estimates, will become clear in the next decade or less.
Hinton also criticized the tech companies' race to create increasingly powerful models and said he doesn't trust OpenAI head Sam Altman at all.
It turns out that one of the people who stood at the origins of modern artificial intelligence now believes that it would be better for humanity not to create it at all.