Nicknamed the “godfather of artificial intelligence”, Geoffrey Hinton has not hidden his antipathy for the boss of OpenAI.
Few Nobel Prize winners end up regretting their work. Geoffrey Hinton, co-winner with John Hopfield of the Nobel Prize in Physics for their contribution to the field of artificial intelligence, is one of them. Nicknamed the “godfather of AI”, the researcher never ceases to warn about the potential deviations of the models.
In his sights, the bane of those who criticize the sudden explosion of ChatGPT: Sam Altman, the boss of OpenAI. During a series of questions and answers with the University of Toronto on October 8, the very venerable Geoffrey Hinton paid the all-powerful boss with a scathing remark.
“I would like to thank my students. I was particularly lucky to have had many brilliant students, much brighter than me,” he said. “They’ve done incredible things. And I’m especially proud that one of my students fired Sam Altman.”
He then refers to Ilya Sutskever, one of the founders of OpenAI, who tried, in vain, to overthrow Sam Altman last year. Although the Board of Directors had succeeded in firing him, it was ultimately forced, under pressure from employees and shareholders, to reinstate him. Ilya Sutskever has since left the company, like many others.
Profits over security
The origin of the conflict is summarized by Geoffrey Hinton: “over time, it turned out that Sam Altman was much less concerned about safety than he was about profits. And I think that’s a shame.”
“OpenAI was set up with a strong focus on security. The first objective was to develop artificial general intelligence and ensure that it would be secure,” he stressed.
Founded in 2015 by Sam Altman and Elon Musk among others, OpenAI has effectively grown from a non-profit organization to a lucrative startup valued at $150 billion, thanks to Sam Altman. To the point that strong dissensions appear within the company over the level of security put in place.
“My concern is that AI can also lead to bad things,” insists Geoffrey Hinton. “Especially when we make things smarter than ourselves. No one really knows if we’ll be able to control them.”
Thomas Leroy Journalist BFM Business
