There is still buzz surrounding the resignation of Geoffrey Hinton, widely regarded as the spiritual father of artificial intelligence, particularly as the 75-year-old linked his decision to a statement yesterday in which he regretted his past work. years.

The renowned British-Canadian expert in cognitive psychology and computer science not only lamented, but also revealed the dangers of artificial intelligence he touched on during his long practical journey, which shocked the tech world and even the stock market. of giant corporations.

“very scary”

The person confirmed that some of the dangers of chatbots for artificial intelligence have become “very scary”, noting that today they are not smarter than humans, but they could change soon and they can be exploited. “Bad parties”, as he put it with the BBC.

He said these programs can now automatically generate a lot of text, making them more effective spam (ie hacking) programs.

He said he worries about the existential risk that it could also allow authoritarian leaders to manipulate their constituencies, and things like that, and what happens when these things are smarter than we are.

Different intelligence

He expressed the conclusion that the intelligence we create is very different from the intelligence humans have, “That’s why if you have 10,000 people, if someone learns something, everyone automatically knows. And bots can understand that this chat is more than any person.”

He explained that he worries that people will soon be unable to discern what is true from the images, videos and AI-generated text circulating on the internet. Realistic images, including one of Pope Francis wearing a puffy coat by Balenciaga, went viral last March.

He also worries that AI will replace paralegals, personal assistants and other hard-working jobs, and more in the future.

Watch out, it’s dangerous!

Notably, Hinton isn’t the only high-profile expert to express fears that technology could seriously harm humanity. Last month, Elon Musk said he disagreed with Google co-founder Larry Page. Grandpa didn’t take the safety of artificial intelligence seriously enough.

Valerie Pisano, CEO of the Quebec Institute of Artificial Intelligence (MILA), felt that a reckless approach to security in AI systems would not be tolerated anywhere else.

And the “spiritual father of artificial intelligence” announced his departure from his job at Google yesterday, Tuesday, after years of successful partnership. It’s the reason I quit working for its giant corporation.”

The announcement came as a surprise to the tech world, especially since Hinton’s work in neural networks is central to the artificial intelligence systems that power many contemporary technology products.

Read more