Elon Musk and the White House want to learn more about the risks of AI
The man often referred to as the “Godfather” of AI is, he says, inundated with requests for help from a billionaire Elon Musk, Democrat Bernie Sanders and the White House just days after leaving Google to warn the world of the dangers of digital intelligence.
Dr. Jeffrey Hinton received the highest honor in computer science, the Turing Award, in 2018 for his work on «deep learning». As The Guardian recalls, the technology now at the heart of the AI revolution came about as a result of Hinton's efforts to understand the human brain – efforts that convinced him that the digital brain might be about to supplant the biological one.
But the London-born 75-year-old psychologist and computer scientist may not be giving the advice the powers that be want to hear this.
“The U.S. government inevitably has a lot of national security concerns. And I tend to disagree with them,” Dr. Hinton told The Guardian. – For example, I'm sure the Department of Defense believes that the only reliable source for this material is the U.S. Department of Defense – the only group of people that actually uses nuclear weapons”.
«I'm a socialist,» Hinton added. “I think that private ownership of the media and 'computing facilities' — this is not good. If you look at what Google does in the context of the capitalist system, you will see that it behaves as responsibly as you might expect it to. But this does not mean that the company is trying to maximize utility for all people: it is required by law to maximize utility for its shareholders, which is a completely different matter”.
Hinton expressed his concern that advances in artificial intelligence could lead to the end of civilization within 20 years.
But when it comes to giving specific advice, he is at a loss for words. «I'm not a political person,» he says. “I’m just a person who suddenly realized that there was a danger that something really bad would happen. I wish I had a good solution, like, «Just stop burning carbon and you'll be fine.» But I don't see such a simple solution.»
Last year, rapid advances in artificial intelligence models convinced Hinton to take seriously the threat that «digital intelligence» may one day supplant the human.
“Over the past 50 years, I have been trying to create computer models that could take in things in much the same way that the brain does, to better understand how the brain learns things. But just recently I decided that maybe these big models are actually much better than the brain. We need to think hard about this now, and if there is anything we can do. The reason I'm not so optimistic is that I don't know of any example of smarter things being controlled by less smart things. You need to imagine something that is smarter than us to the same extent that we are smarter than a frog. It's all very nice to say, «OK, don't connect them to the Internet,» but as long as they're talking to us, they can get us to do something.
Hinton has other concerns about the rapid growth in the power of artificial intelligence models, citing the influence of Cambridge Analytica advocate Robert Mercer on political campaigns on both sides of the Atlantic: “Bob Mercer and Peter Brown, when they worked at IBM on translation, understood the power of having a large amount of data. Without Bob Mercer, Trump might well not have been elected yet. And Bob must have understood the manipulative power that big data can give you, which is why I think it's already backfired”.
Authoritarian governments, according to Hinton, are the biggest red flag , which suggests that humanity will not be able to realize the risks associated with artificial intelligence until it is too late: “This material helps authoritarian governments destroy the truth or manipulate the electorate. And when you have to deal with these threats, in a situation where Americans can't even agree not to give assault rifles to teenage boys, it's not hard to think about. In Uvalde [the 2022 Texas elementary school massacre of 21] there were 200 cops who didn't dare to come in the door because the guy on the other side had an assault rifle and he was shooting at kids. And yet they can't decide not to ban assault weapons. So a completely dysfunctional political system like this is simply not up to the task of dealing with these threats.