View from Westminster Sign up for emails and get expert analysis delivered straight to your inbox.
Get your free View from Westminster email
As the AI Seoul Summit kicks off on Tuesday, some of the world’s leading AI scientists warn that not enough progress has been made since the first AI Safety Summit in the UK six months ago and We are calling for stronger action from world leaders.
Prime Minister Rishi Sunak will co-host a virtual summit with South Korean President Yoon Seok-yol on Tuesday, saying managing the risks posed by artificial intelligence is “one of the most important responsibilities” we face. said. by the government.
But in a new expert consensus paper published in Science, 25 of the world’s leading scientists in the field say governments have moved too slowly to regulate the rapidly evolving technology and that not enough progress has been made since the last summit.
They believe that world leaders will take seriously the possibility that more powerful general-purpose AI systems (capable of regularly outperforming humans) will be developed this decade or next, and respond accordingly. argues that it is necessary to do so.
Co-author of the paper, Professor Philip Toll from the University of Oxford’s School of Engineering, said: .
“This document provides a number of important recommendations for what businesses and governments need to address.”
In their paper, the experts say a rapid response authority for AI oversight needs to be established and funded far more significantly than many governments currently plan, while also mandating more rigorous risk assessments with enforceable consequences rather than the current voluntary and unspecified assessment model.
Experts include Turing Award winners, Nobel Prize winners, authors of standard AI textbooks, and others from major AI powerhouses such as the UK, US, China, and the EU.
Stuart Russell, a computer science professor at the University of California, Berkeley, and author of textbooks on AI, said: “This is a consensus document from leading experts, not a self-imposed code of conduct. We are calling for strict regulations.” By industry.
“The time has come to get serious about advanced AI systems. These are not toys. It would be downright foolhardy to increase their capabilities before we understand how to make them safe.
“Companies will complain that regulations are too hard to meet, that ‘regulation stifles innovation.’ That’s ridiculous. There are more regulations for sandwich shops than there are for AI companies.”
Ahead of the summit, the first edition of a new scientific report on AI safety – the first of its kind and commissioned at the AI Safety Summit in November – will present relevant experts The future of technology proved uncertain.
They argue that while AI has the potential to promote welfare, prosperity, and scientific research in the future, it could also be used to promote widespread disinformation and fraud, disrupt jobs, and reinforce inequality. He said that there is a sex.
In addition to highlighting the potential benefits and risks, the report also discusses various aspects of AI, including the current state of AI capabilities, how they will evolve over time, and the potential for extreme risks. It warns that there is no universal agreement among experts on this topic. It happens that we lose control of technology.
The interim report will be used as a starting point for discussions among world leaders, industry experts, researchers and tech giants at a recent two-day summit in Seoul.
This comes as the pace of innovation in the field shows no signs of slowing, with ChatGPT maker OpenAI, Google and Microsoft all announcing a range of new AI-powered tools and products ahead of the summit.
And Apple, another heavyweight in the tech world, plans to unveil its own AI in early June.