Skip to main content

Stephen Hawking fears competent machines, not malicious ones

Stephen Hawking fears competent machines, not malicious ones

Share this story

Like any good scientist, Stephen Hawking has taken his time. A Reddit AMA (Ask Me Anything) that collected questions from the site's users back in July is today being answered comprehensively by the world's most famous living physicist. Unsurprisingly, given his high profile and the extended period allowed for questions, this is already the third-largest AMA that Reddit has hosted, with over 9,000 comments submitted.

Artificial intelligence has been the predetermined focus for the AMA, building on the open letter that Hawking signed this summer — along with other public figures like Noam Chomsky and Elon Musk — urging for greater human oversight when developing AI systems. Today he expands on the argumentation of that letter, which is primarily concerned with autonomous weaponry of the sort already in deployment by the US military (aka drones). Looking further off into the future, Hawking says it's not malicious machines that would be the problem, but competent ones. "A super intelligent AI will be extremely good at accomplishing its goals," says Hawking, "and if those goals aren’t aligned with ours, we’re in trouble." In other words, if AI does wipe out the human race, it would more likely be a form of collateral damage than some villainous eradication plot. This is why Hawking and others continue to insist on maintaining human control over AI.

There are more lighthearted aspects to the AMA as well, such as the professor confessing his fondness for The Big Bang Theory and Rod Stewart's "Have I Told You Lately." Check out the full disclosure at the link below.