As concerns about the development of increasingly powerful AI grow, artificial intelligence pioneer Geoffrey Hinton has said that there is a 10% chance of AI causing the extinction of the human race within the next decade, adding that that’s “not unreasonable.”
During the interview he was asked about a recent statement by Anthropic researcher Evan Hubinger, who believes there is a more than 10% chance that AI will cause human extinction in the next decade. Hinton said the estimate was a “reasonable” one.
But the danger could still occur if AI doesn’t actually go after humans, Hinton, referred to as the “godfather of AI,” said. He said some of the potential applications are having AI used in the development of biological or computer viruses, and conducting advanced cyberattacks.
He also highlighted that the probability of such unprecedented technology is more of a shot in the dark, and not backed by real evidence. Hinton has previously predicted that there is a 10-20% possibility of AI leading to the extinction of the human race in 30 years.
The warning coincides with a resurgence in the debate about the safety of the new advanced AI. Anthropic alignment researcher Evan Hubinger recently stated that the company doesn’t yet have a solution for the problem of keeping future superintelligent systems under human control, and former Anthropic researcher Jacob Coxon cautioned AI companies were rushing to make systems that could self-improve.
In 2023, Hinton left Google to be able to speak out more openly about the potential dangers of AI. He has been a vocal advocate for caution, especially as AI systems grow more sophisticated and capable than human intelligence.
