“There is a small chance that humanity will survive.”
terminator vision
Famously pessimistic AI researcher Eliezer Yudkowsky is back with new predictions about humanity’s future.
“If you drive me up the wall,” he said guardian In a fascinating new interview, he says: “I’m forced to put probabilities into things, but I feel like the timeline we have right now is more like five years than 50 years. Maybe two years. It could be 10 years.”
If you’re wondering what “remaining timeline” means in this context, guardian‘s Tom Lamont interpreted this as a “machine-induced apocalypse”, a “Terminator-style apocalypse”, or a “Matrix hellscape”.
“The difficulty is that people don’t realize it,” Yudkowski, founder of California’s Machine Intelligence Institute, told the newspaper. “There is a small chance that humanity will survive.”
bomb disposal squad
whole guardian This is a work worth reading. Lamont spoke with many luminaries in the field, from Brian Merchant to Molly Crabapple, and found that the conventional wisdom is that just because a new technology emerges, it’s not good for people. Some expressed skepticism about the idea that it is necessary to adopt a standard, even if it is a standard.
These days, much of that criticism focuses on AI. Why, critics argue, should we treat this technology as inevitable, even if it appears to be about to eliminate and destabilize large numbers of jobs? .
Or, in Yudkowsky’s case, when the technology is likely to pose an existential threat. His statement was the most provocative of the piece, which is not surprising given his background. AI watchers may remember, for example, last year when he called for bombing data centers to stop the rise of AI.
He has reconsidered that particular claim, he said. guardian — but only a little. Although he supports the idea of bombing data centers, he said he no longer believes nuclear weapons should be used to target data centers.
“I would choose more careful language now,” he told the newspaper.
Learn more about AI: Oppenheimer’s grandson signs letter saying AI threatens ‘life on Earth’