“You could say I feel lost.”
Godfather Part II
Yoshua Bengio, a famed computer scientist who’s considered one of the three “godfathers” of artificial intelligence, is starting to feel a little blue about his life’s work, as AI — or at least its breathless hype — seems poised to spiral out of control.
In a new interview with the BBC, Bengio said that had he known how rapidly AI would develop, he would have prioritized safety over usefulness.
“You could say I feel lost,” Bengio told the outlet. “But you have to keep going and you have to engage, discuss, encourage others to think with you.”
The Canadian computer scientist’s comments come after he signed a disquieting open letter from industry leaders that warns of the “risk of extinction” that AI poses, along fellow AI godfather Geoffrey Hinton, who recently quit his job at Google after a similar personal reckoning.
Bad Actors
For now, one of Bengio’s foremost concerns is “bad actors” abusing AI, echoing Hinton’s sentiments from an earlier interview last month.
“It might be military, it might be terrorists, it might be somebody very angry, psychotic,” Bengio told the BBC. “And so if it’s easy to program these AI systems to ask them to do something very bad, this could be very dangerous.”
Bengio added that he believes governments need to keep tabs on companies developing AI in the same way they do “for any other sector like building airplanes or cars or pharmaceuticals.”
On top of that, Bengio suggested the radical idea that the people working on these AI systems need some form of certification in the way of “ethical training.”
“Computer scientists don’t usually get that, by the way,” he added.
Correcting Course
It’s usually never good when pioneering inventors liken their work to the atom bomb, which both Bengio and Hinton have done in their respective interviews.
But the remaining third “godfather,” Yann LeCun, apparently isn’t convinced that AI poses an existential threat to humanity, since he’s the only one of the trifecta not to sign the recent risk statement, not to mention being a vocal opponent of a proposed moratorium on AI development.
However squabbling humans decide to address the issue, Bengio, at least, thinks the challenge is surmountable.
“It’s never too late to improve,” Bengio said in the interview. “It’s exactly like climate change. We’ve put a lot of carbon in the atmosphere. And it would be better if we hadn’t, but let’s see what we can do now.”
More on AI: Eating Disorder Helpline Takes Down Chatbot After It Promotes Disordered Eating