Yes, artificial intelligence could destroy humanity. And in a rather short time too. This is the alarm raised by Geoffrey HintonNobel Prize winner for physics in 2024 and one of the fathers of AI. The Anglo-Canadian scientist explained that there is “a 10-20 percent chance that within the next 30 years artificial intelligence will cause the extinction of humanity”. A prophecy, the one on the microphones of the program “Today” on BBC Radio 4different from the previous one, in which he spoke of a percentage of around 10 percent.
The hypothesis of an apocalypse caused by AI exists and should not be underestimated: Hinton’s warning is peremptory. Also because the pace of evolution of artificial intelligence “It’s going much faster” of expectations. “You see, we’ve never had to deal with things smarter than us” the expert’s emphasis: “And how many examples do you know of a more intelligent thing being controlled by a less intelligent thing? There are very few examples. There is a mother and a child. Evolution has worked hard to allow the baby to control the mother, but this is the only example I know of. I like to think of it like this: let’s try to imagine ourselves as a three-year-old child and AI as an adult.”
In 2023, Hinton resigned from Google, where he worked on AI development. A choice dictated by the desire to speak openly about the risks posed by the development of unconstrained artificial intelligence. His concerns are known to all: the professor emeritus at the University of Toronto fears that systems more intelligent than humans could represent an existential threat to our species by bypassing every obstacle and every control by man. “I didn’t think we would get to where we are now” he admitted: “I thought we would get there, but in a more distant future. In twenty years we will develop an artificial intelligence that is more intelligent than humans. And that’s a very scary hypothesis.”
But not only that.
Hinton believes that leaving AI in the hands of large companies, motivated solely by profit, is not enough to guarantee that they will develop it safely: “The only thing that can force those companies to do more research into AI safety is government regulation of the industry.”.