A statement published by the Centre of AI Safety reads: “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”
It’s signatories include OpenAI chief executive, Sam Altman, whose company is behind the increasingly popular ChatGPT, the three co-founders (Demis Hassabis, Mustafa Suleyman and Shane Legg) and a number of researchers from Google’s AI research lab Deepmind, and academics from top universities including Oxford, Cambridge, Stanford, Harvard, Yale, and MIT.
The centre’s webpage states the aim of the statement is to “open up discussion” about “concerns about some of advanced AI’s most severe risks”, which it says people find difficult to voice.
The i revealed recently that Prime Minister Rishi Sunak wanted new “guardrails” to protect against the potential dangers of AI, citing that there could be benefits if it was used safely. Regulation would need a co-ordinated approach from allies, he said.
Mr Sunak also met with Open AI’s Mr Altman, Deepmind’s Mr Hassabis, and Dario Amodei from Anthropic, a company focused on safe AI use. The quad discussed safe and responsible AI use and how it could be governed.
Despite being in charge of one of the world’s most popular AI technologies, Mr Altman has been quite firm and consistent in his warnings of possible dangers.
Just last week, he told an event at University College London that he thought AI technologies should be treated “as seriously as we treat … nuclear material”, particularly warning against “if someone does crack the code and build a super intelligence”.
Earlier this month, he urged the US Senate to regulate AI and in March he told ABC News he was “a little bit scared” of the technology.
However, some scientists think fears about AI are overblown.
Meta’s chief AI scientist, Yann LeCun, also a professor at NYU, regularly posts on social media against the idea that the current trajectory of AI is destined towards doom.
He acknowledges there are concerns and some risks, but also highlights there are benefits to be gained from the technology.
“AI systems may become more capable than you, but you’ll still be their boss,” he tweeted on 21 May. “If you feel threatened by having a staff — of humans or machines — that is smarter than you, you are not a good boss.”
Mr Altman is sometimes the target of Mr LeCun’s rebuttal of fears against AI, but it’s widely considered quite serious for the boss of such a technology to warn against their own product to some extent.
Even before getting to doomsday ideas, the rapid rise of AI technologies has been controversial.
University students have openly admitted to using Mr Altman’s ChatGPT to write essays, with some institutions banning the technology, and German magazine Die Aktuelle hit hot water in April after it published AI-generated quotes from Formula One legend Michael Schumacher.
Mr Altman’s concerns are much bigger. In his March comments to ABC News, he said he was worried about how the technologies could be used for “large-scale disinformation”, and even cyber-attacks as AI models learn to write computer code.
During a Senate hearing in May, he said his “worst fears are that we cause significant harm to the world”, noting that AI technologies are only going to get more powerful and intelligent, and the risk that AI could reach a point of functioning outside of human control.
More immediately, because of things like political elections, he said it presented a threat to democracy because AI models’ ability to manipulate and persuade, alongside the possibility of disinformation.
Twitter chief executive, Elon Musk, has also previously aired his thoughts about a lack of regulatory oversight of AI, saying he’s been calling for it for a decade.
And like Musk, fears about the technology have been raised by various other experts for years — AI is just in the public domain like never before.
French scientist Dr Birgitta Dresp-Langley, in an article published in the academic journal Frontiers of Artificial Intelligence this March, wrote about how AI is already being used in the weapons space but there was real potential for it to be used to create weapons of mass destruction in the future.
Parliament in the UK is aware of this too. Between 6 March and 8 May, the Artificial Intelligence in Weapons Systems Select Committee sought public feedback about the development, use and regulation of automated weapons systems (AWS).
At the time, Lord Lisvane acknowledged the use of AI in defence was controversial for AWS, and they planned to examine the ethics of these and if they risked escalating war more quickly.