A man widely regarded as the godfather of artificial intelligence (AI) has quit his job to warn of the growing dangers posed by developments in the field.
Geoffrey Hinton, 75, announced in a statement to the New York Times that he has retired from Google, saying he now regrets his job.
He told his BBC that some of the dangers of AI chatbots are “pretty scary”.
“For now, as far as I know, they’re not smarter than us. But I think they’ll be soon.”
Dr Hinton also told the BBC:
“I think he’s 75 and it’s time to retire.”
Dr. Hinton’s groundbreaking work on neural networks and deep learning paved the way for his current AI systems like ChatGPT. A neural network in artificial intelligence is a system that resembles the human brain in the way it learns and processes information. This allows AI to learn from experience in the same way humans do. This is called deep learning.
The British-Canadian cognitive psychologist and computer scientist told the BBC that chatbots could soon surpass the information level of the human brain.
“Today, things like GPT-4 overwhelm people with the amount of general knowledge they have and are much smaller. In terms of reasoning, it’s not very good, but it’s already doing basic reasoning,” he said.
“And given the speed of progress, we expect things to get better pretty quickly, so we’ll have to deal with that.”
Is the world prepared for the coming AI storm?
AI reports he could affect 300 million jobs
New York Times article hints at “villains” trying to use Dr. AI for “bad things”.
When asked for details by the BBC, he replied:
“It’s kind of a worst case scenario, kind of a nightmare scenario. “For example, you could imagine that a bad guy like [Russian President Vladimir] Putin decided to give robots the ability to create their own subgoals.”
The scientist warned that this could eventually “generate sub-goals such as ‘I need to be stronger’.”
“I have come to the conclusion that the kind of intelligence we develop is very different from the intelligence we have.
“We are biological systems and these are digital systems. The big difference is that with digital systems there are many copies of the same set of weights of the same model all over the world.
“And all of these copies can be studied individually, but they can share their knowledge immediately. So it’s like having 10,000 people and if one learns something, everyone automatically knows. So these chatbots let him know more than just one person. ”
Matt Clifford, chairman of the UK’s Agency for Advanced Research and Inventions, told the BBC in his personal capacity that the presentation by his Ph.D. Hinton “highlights the speed at which AI capabilities are accelerating.”
“While this technology has significant benefits, it is critical that the world invest heavily and urgently in AI security and control.
Dr Hinton did not intend to criticize Google, stressing that it acted “very responsibly”.
“Actually, I want to say some good things about Google. And it’s more reliable when I’m not working at Google. “
In a statement, Google chief his scientist Jeff Dean said:
“We remain committed to the responsible use of AI. We are continually learning to understand new risks while innovating boldly.”