AI bots like ChatGPT could herald an ‘alien invasion’ with the capacity to ‘wipe out humanity’
AI bots like ChatGPT could herald an ‘alien invasion’ with the ability to ‘exterminate humanity’, artificial intelligence professor warns
AI bots like ChatGBT could herald an “alien invasion” with the ability to wipe out humanity, a leading artificial intelligence professor warned.
Stuart Russell, who laid the groundwork for new technology, argued that ministers were complacent about its risks.
The British professor, who has advised Downing Street and the White House, said he is annoyed that ministers have been calling for light regulation of the AI industry despite warnings from officials that the products could pose a huge threat.
Russell recently signed an open letter warning that AI labs were “stuck in an out-of-control race to develop and deploy increasingly powerful digital minds that no one, not even their creators, can understand,” according to The times.
“We’ve made progress without expecting it,” he said, adding that it’s possible that a system similar to ChatGBT could be part of a super-intelligent machine that could withstand any attempt to control it.
AI bots like ChatGBT could usher in an ‘alien invasion’ with the ability to wipe out humanity, a leading artificial intelligence professor warned
Stuart Russell (pictured), who laid the groundwork for new technology, argued ministers were complacent about its risks
A spokesman for the UK government said: ‘We recognize that as artificial intelligence systems become more powerful, they will create new challenges and risks. That is why the international technology strategy includes a public commitment to start a global dialogue on AI risks.
“We will have to act and adjust the way we regulate technologies such as AI, which play an increasingly vital role in our economy and society.”
For about 10 years now, Russell has been calling for a ban on using AI to make autonomous weapons.
He believes one of the risks is that a powerful AI gets a goal and chooses a method to achieve that goal, but with unexpected side effects that are devastating to humanity.
He used the example of a world where AI is challenged to solve climate change, and decides that the best way is to rid the Earth of humans.
Russel explained that the benefits of using AI responsibly can be great – one way to do this is to create systems that are only good at one thing, as opposed to an AGI that can pick up on any human skill.
“It’s an incredibly useful thing. But it’s not about taking over the world, is it?’, he said.