Artificial intelligence will destroy humanity with a 95% probability – ex-Google engineer
3 August 13:25
This opinion was expressed by Nate Soares, a former Google and Microsoft engineer and now president of the Machine Intelligence Research Institute.
This was reported by The Times, according to "Komersant Ukrainian".
“If we don’t turn off the current course, there is almost no chance of avoiding death. We are rushing to the cliff at 100 km/h,” he said.
Nobel Prize winner Jeffrey Hinton, mathematician Joshua Bengio, and the heads of OpenAI, Anthropic, and Google DeepMind agree with him.
All of them signed a statement saying that “reducing the risk of humanity’s extinction through AI should be a global priority alongside pandemics and the nuclear threat.”
The scientists’ concerns are related to the possible emergence of super-intelligent AI (ASI) capable of deception, planning, and getting out of control. Even now, AI can already lie, and its internal processes are often incomprehensible to people.
Some experts believe that the risk of humanity’s “gradual exhaustion” rather than complete extinction is also real: in a world where decisions are made by machines, there will simply be no room for humans.
Meanwhile, Jeffrey Nielson, the “godfather of AI,” estimates the chances of AI at 10-20% within 30 years. Holly Elmore from PauseAI estimates 15-20% in the next 10 years and up to 75% in the 50-year perspective.
