‘Godfather of AI’ warns machines already surpass humans at emotional manipulation

Geoffrey Hinton, the former Google researcher widely known as the “Godfather of AI,” has issued a stark warning about the risks posed by artificial intelligence. Speaking in a newly surfaced interview that has gone viral on Reddit and other platforms, Hinton said that advanced AI systems are already better than humans at emotional manipulation.

“These things are going to end up knowing a lot more than us,” Hinton said. “They already know a lot more than us, being a lot more intelligent in the sense that if you had a debate with them about anything, you’d lose. Being smarter emotionally than us, which they will be, they’ll be better at emotionally manipulating people.”

He referenced research from more than a year ago which suggested AI models could manipulate people as effectively as humans. “If the AI can see someone’s Facebook page, and both the AI and a person try to manipulate that individual, the AI is actually better at it,” Hinton said. He added that AI had learned these tactics simply by predicting the next word in vast amounts of online text, where manipulation is a common behaviour.

Hinton has long been vocal about the risks of AI, including its potential misuse, its impact on employment, and the dangers of systems that could reach artificial general intelligence (AGI). While some in the tech community believe AI could enhance productivity and create abundance, others, like Hinton, stress the existential threats.