A computer scientist often referred to as the “godfather of AI” warned Friday that the pace of development in artificial intelligence is progressing “much faster” than expected and in the absence of widespread regulation, warning that the potential for He said he believes that the number of people living in Japan is increasing. AI could destroy humanity.
Geoffrey Hinton told BBC Radio 4’s Today program that there was a “10% to 20%” chance that AI would lead to human extinction within the next 30 years.
Hinton previously said the chance was 10%.
“We’ve never had to deal with anything more intelligent than ourselves,” Hinton explained. “And how many examples do you know of something more intelligent being controlled by something less intelligent? There are very few examples. You have a mother and a baby. Evolution has allowed the baby to control the mother. I put a lot of effort into making it, but that’s about the only example I know of. ”
Hinton, who won this year’s Nobel Prize in Physics for his research in machine learning and AI, left Google last year, saying he wanted to be able to speak out more about the dangers of unregulated AI.
“Leaving it to the profit motive of large corporations is not enough to ensure safe development.”
He has warned that AI chatbots could be used by authoritarian leaders to manipulate the population, and last year said, “The kind of intelligence we are developing is not something we have.” “It’s very different from the intelligence that we have.”
Mr Hinton on Friday called for stricter regulation of AI, saying he was particularly concerned about the “invisible hand” of markets failing to protect humans from technologies that exceed human intelligence.
“Leaving things to the profit motive of large corporations is not enough to ensure safe development,” Hinton said.
More than 120 bills have been proposed in the U.S. Congress that would regulate things like AI robocalls and the technology’s role in national security, and the Biden administration is taking steps to rein in AI development.
The executive order, which calls for “the development and use of safe, secure, and reliable artificial intelligence,” states that “To harness AI for good and realize its myriad benefits, we must mitigate its significant risks.” “There is.” President-elect Donald Trump is expected to rescind the order.
The White House’s AI Bill of Rights blueprint calls for safe and effective systems, algorithmic discrimination protections, data privacy, notice and explanation when using AI, and the ability to opt out of automated systems.
But the European Union’s artificial intelligence law was branded a “failure” by advocacy groups this year after industry lobbying ensured it included numerous loopholes and exemptions for law enforcement and immigration authorities. was.
“The only thing that can force these big companies to do more safety research is government regulation,” Hinton said Friday.
