- Gates said "even in the next 2 to 3 years, you'll see significant negative effects in in those two areas", naming cyber attacks and the job market (49:25).
- His big concern is a 20-year transition period "where it'll have winners and losers and a lot of disruption", a length he picked "as a reasonably safe number" (42:00, 42:19).
- Gates said the notion that liability laws are how we manage AI safety "will be looked back on as as kind of absurd" (53:53).
Bill Gates, co-founder of Microsoft and chair of the Gates Foundation, was interviewed by Nilay Patel on the Decoder podcast. The video, "Bill Gates says to prepare for 20 years of tumultuous AI disruption", was published on the Decoder YouTube channel on 22 October 2026.
Asked whether he had become a doomer, Gates said "No, I'm not a doomer" (00:02). Of the response so far, he said it "hasn't been adequate to the speed that that the challenge is arising" (01:24). On timing, he said "some of the things like cyber attacks and the job market thing, you know, even in the next 2 to 3 years, you'll see significant negative effects in in those two areas" (49:25). Of the longer horizon he said "My big concern is this 20-year transition period, where it'll have winners and losers and a lot of disruption" (42:00), and that he picked 20 years "as a reasonably safe number" (42:19).
Patel read out a line from Gates's letter calling for a response greater than the US national security response after 9/11 and asked about an international body (49:49); Gates answered "Absolutely." (50:05). On liability, Gates said "The notion that liability laws are how we're going to manage AI safety will be looked back on as as kind of absurd" (53:53). On open-weight models he said "I do think allowing people through open weights to remove all safeguards" (51:18) "is irresponsible" (51:25), and "I would hope that uh models above a certain capability uh that we find a way to require the safeguards and monitoring to be maintained in those models" (54:53).
What it means for you Our view
This is Gates's view, not a forecast we can check. He says the response so far "hasn't been adequate to the speed that that the challenge is arising", calls allowing people to remove all safeguards through open weights irresponsible, and hopes models above a certain capability will be required to keep safeguards and monitoring. His time frame, two to three years, falls inside most planning cycles. Check whether your security review and your workforce plan for the next two to three years assume AI changes nothing, and write down what would change them.
TAGSDecodercyber attacksjobs
