Geoffrey Hinton worries that “bad actors” won’t be stopped from using AI.
Geoffrey Hinton, a Google engineer who was known as the “godfather” of AI, quit his job and is now warning about the risks of developing AI further.
Hinton worked at Google for more than ten years, and he is the one who made a big technological breakthrough in 2012 that is the basis for AIs like ChatGPT. In a letter to the New York Times, he said he was leaving Google because he now regretted his work there.
He told the paper, “I make myself feel better by telling myself that if I hadn’t done it, someone else would have.”
Hinton went on to say about AI, “It’s hard to see how you could stop bad people from using it for bad things.”
In 2012, when Hinton was working with two graduate students in Toronto, they made a big step forward in AI. The NYT says that the three people were able to make an algorithm that could look at shots and find familiar things like dogs and cars.
The program was just the start of what modern AIs like OpenAI’s ChatGPT and Google’s Bard AI can do. Shortly after the success, Google paid $44 million for the company Hinton built around the program.
Ilya Sutskever, a graduate student who worked on the project with Hinton, is now the chief scientist at OpenAI.
Hinton said that the changes since 2012 are amazing, but they may only be the tip of the storm.
“Look at how the industry was five years ago and how it is now,” he said. “Take the difference and pass it along. “That’s spooky.”
Hinton’s worries are similar to those of more than 1,000 tech leaders who called for a short halt to AI development in a public letter earlier this year. At the time, Hinton did not sign the letter, and he says now that he did not want to attack Google while he was working there. Since then, Hinton has left the company. On Thursday, he talked on the phone with Sundar Pichai, the CEO of Google.
“We are still committed to AI in a sensible way. Jeff Dean, Google’s chief scientist, told the Times that the company is constantly learning about new risks and taking chances with new ideas.