Rapid AI improvements have stirred many observers’ fears about the tech and its implications for society. There’s even some “doomerism . . . coming from inside the industry,” Michael says on this week’s second edition of The Editors. 

“You’ve also seen a few things like resignations of people from . . . the safety departments of these companies saying stuff like, ‘I’m retreating to the U.K. to study poetry in anonymity while we have time left.’” he says. Michael, however, has “generally been surprised at how warm I felt towards the development of AI.” While Michael has his worries, he says, “It’s just not clear to me that this is a job-killing technology. So for instance, you might feel . . . ‘This is horrifying that a computer can study every law document ever written and then it can produce law documents. What’s going to happen to all the lawyers?’ Well, the lawyers are going to have to read a lot of law documents that have been produced by AI. There’s still human interaction here.”

Michael is more worried about data breaches, saying, “We’ve really had a ton of notorious damaging hacks already in the history of the internet: the Equifax breach, the Yahoo breach, Stuxnet worm, Colonial pipeline. The chances for something like that on a much more massive scale or disruptive scale does seem high to me, at least in the early days.

“You wouldn’t want extremely powerful industrial machines just like being set into random motion in your neighborhood,” he says. “You wouldn’t expect a good outcome. So I’m worried about something like that, which is not an apocalypse, but just, you know, major accidents.

“I think right now is the time it’s to be bullish on this technology, not to be terrified of it.”

The Editors podcast is recorded on Tuesdays and Fridays every week and is available wherever you listen to podcasts.

© National Review