Scientists say human survival is threatened by advances in artificial intelligence

The computer scientists who helped lay the foundations for the artificial intelligence (AI) technology currently in use are now warning of its dangers, though they don’t agree on what it is or how to avoid it.

The survival of humanity is at stake when “intelligent things can fool us,” warned Jeffrey Hinton, considered the “godfather” of AI, during a conference this Wednesday at the Massachusetts Institute of Technology (MIT), in the United States. .

“It could keep us going for a while to keep the factories running. But after that, maybe not,” he stressed.

After parting ways with Google to speak more freely, Hinton, 75, said he recently changed his mind about the reasoning capabilities of computer systems, which he has explored all his life.

“We learned this from us, reading every novel that ever existed and everything that Machiavelli ever wrote, how to manipulate people,” Hinton emphasized, addressing the audience at the conference.

“Even if they can’t pull the levers directly, they can certainly make us pull the levers,” he added.

As for possible solutions, Geoffrey Hinton stressed that he does not have any, and he is not sure if they exist, according to the Associated Press (AP).

Yoshua Bengio, another AI pioneer and co-winner with Hinton of the top computer science prize, told the AP on Wednesday that he “pretty much agrees” with Hinton’s concerns, although he is wary of simply saying humanity is “condemned” there will be no help.

“The main difference, I would say, is that he is a bit of a pessimist and I am more optimistic,” said Bengio, a professor at the University of Montreal, Canada.

“I think the dangers – short-term and long-term – are very serious and should be taken seriously not only by some researchers, but also by governments and the public,” he warned.

There are several signs that governments are taking notice, such as the White House case, which called the CEOs of Google, Microsoft and OpenAI, the creator of ChatGPT, to a meeting with Vice President Kamala Harris today. authorities as a frank discussion of how to mitigate the short and long term risks of this technology.

European authorities are also accelerating negotiations on a comprehensive new AI regulation.

Others fear that talk of future dangers posed by superhuman machines that don’t exist is a distraction from trying to establish practical safeguards for existing AI products that are unregulated and have been shown to cause problems in the real world.

Margaret Mitchell, former head of Google’s artificial intelligence ethics team, said she didn’t understand why Hinton didn’t speak out during the decade he held a leadership position at Google.

Bengio has expressed concern for years about the short-term risks associated with artificial intelligence, including labor market disruptions, automated weapons, and the dangers of biased datasets.

But these worries have intensified recently, prompting Bengio to join other scientists and tech company leaders like Elon Musk and Apple co-founder Steve Wozniak in calling for a six-month hiatus from developing the world’s most powerful artificial intelligence systems than the latest OpenAI model. GPT-4.