Elon Musk will head to Downing Street for talks with Rishi Sunak today following the prime minister’s AI safety summit.
The billionaire owner of SpaceX and Tesla jetted in for the event at Bletchley Park, which began on Wednesday with attending countries backing an agreement on the need to manage risks posed by the technology.
China and the US, the world’s leading AI powers, were among 28 countries to endorse the Bletchley Declaration.
It said nations should work together to research the safety of so-called frontier AI models, which some experts – including Musk – believe could one day threaten humanity.
“It’s a risk,” he told Sky News on day one of the summit.
PM’s AI balancing act
The Bletchley Declaration said any threats are “best addressed through international cooperation”, and also set out plans for more global summits next year.
Mr Sunak said the agreement was a “landmark achievement”.
But there was little sign of a concrete approach to regulation or any suggestions of a pause in AI’s development, which experts including Musk called for earlier this year.
It also did little to satisfy critics who warned the prime minister ahead of the summit he was too focused on hypothetical future threats, rather than present dangers like job losses and misinformation.
In a joint statement after the declaration was published, leading AI experts and civil society organisations warned politicians were not showing enough urgency to regulate.
Technology Secretary Michelle Donelan has defended the government’s approach at the summit, saying more hypothetical risks were still ones “we shouldn’t take lightly”.
She said the government was seeking to “strike the right balance” between safety and innovation.
Leading AI firms Anthropic and ChatGPT maker OpenAI have opened international offices in the UK, she added, proving the government was taking the right approach.
China keeps close control of its AI companies, will the West be able to do the same?
Elon Musk might have brought some stardust to this summit, but a more quietly significant presence was the Chinese government.
Although AI safety has been discussed in places like the UN, this is the first time China has sat round a table to discuss the issue with their American and European counterparts.
The UK government faced criticism from some of its own MPs for inviting China. The truth is, any honest effort to mitigate the risks of AI has to be a global one.
If, as some have suggested, super-intelligent AIs of the future might represent the same existential risk as nuclear weapons did in the 20th century, only a similar level of international agreement can keep us safe.
According to Professor Yi Zing, an AI researcher at the state-run Chinese Academy of Sciences, China has already developed AIs equally as powerful – and potentially as problematic – as GPT4 and its rivals in the West.
The major difference of course, is that the Chinese state keeps close control over its AI companies – and can ensure it benefits from any advances they make.
For regulators in the West it’s not so easy. Can they persuade increasingly powerful AI firms to allow them meaningful access to their AI models to ensure they are safe? And what can they do if they conclude they are not? Progress on that is a key objective of the second day of this summit.
When is the Sunak-Musk meeting?
The meeting between the prime minister and Musk will take place after the summit has officially closed.
Thursday will see Mr Sunak convene a small group of governments, companies, and experts, while the technology secretary will meet again with her international counterparts.
His talks with Musk will take place in Downing Street, and be livestreamed on X (formerly Twitter).
Musk has hosted politicians on the platform before, notably a glitch-filled discussion with Ron DeSantis when the Florida governor launched his US presidency bid.
Musk and Mr Sunak have been divided on the need for AI regulation, with the former telling the US Congress in September there was “overwhelming consensus” for it.
Mr Sunak on the other hand has expressed caution, saying too much oversight would stifle innovation.
But Musk – the world’s richest man – changed his tune somewhat ahead of his UK trip, voicing his opposition to sweeping safeguards unveiled by US President Joe Biden earlier this week.
It included requiring AI companies to share safety data with the government before releasing their models publicly.
Speaking at the UK summit, Musk suggested he would prefer a “third-party referee” to regulate the sector.