Rishi Sunak’s killer robots summit was a success

We should proclaim the unsung success of Rishi Sunak’s artificial intelligence summit. It is easy to mock the prime minister’s fanboy interview with Elon Musk, and we can certainly have fun at Sunak’s naivety in saying that the British ought to be ready to give up “the security of a regular pay cheque ... and be comfortable with failure”. But the summit itself showed that Britain still matters, despite Brexit.

It showed that Sunak was prepared, for the sake of the national interest, to take the personal risk of looking like a doofus. It was in the national interest to organise a global conference on the dangers of artificial intelligence (AI), even though it was widely predicted that it would fail. No one will come, it was said. They won’t be able to agree anything useful, everyone agreed.

But they did come, and enough of them at a senior enough level, including Kamala Harris, vice president of the US, and Wu Zhaohui, a vice-minister of science and technology in the Chinese government, to make it count. All the independent specialist observers I could find said – often to their surprise – that the agreement, the Bletchley Declaration, was worthwhile. Sunak said to Musk, in their conversation at the end of the summit, that one of the questions he wanted to ask was “how do you transpose” that risk-taking culture “from places like Silicon Valley across the world”? He was talking about himself, a transposition from California, confident enough to risk failure in organising the summit, and indeed to take the risk of engaging not just with China but with Musk. The same argument applied to both. We may not like them, but China and Musk are both powers on the world stage, and if Britain wants to be a leader in tech it needs to engage with them.

Leaving the cringier parts of their discussion to one side, Musk repaid Sunak’s investment in him handsomely. San Francisco and Greater London are the “two leading locations on Earth” for AI, he said, adding that Britain is “doing very well”. He said: “Thanks for this summit. I think it will go down in history as quite important.” And he praised Sunak for inviting China. “Having them here, I think (is) essential really. If they’re not participants, it’s pointless.” Take that, Liz Truss, China “hawk”. Musk was there for the whole two-day event at Bletchley Park, the HQ of the Allied code-breaking operation during the Second World War, along with Sam Altman, boss of OpenAI, and the representatives of 28 countries and the EU.

Personally, I think Musk’s musings about us having to find something meaningful to do with our lives when AI has abolished work are a form of neo-Luddism. Clever people have been predicting that technology would mean the end of employment for centuries now. John Maynard Keynes in 1930 predicted that we would be working 15 hours a week by 2030.

Similarly, I am on the sceptical end of the “killer robots will kill us all” spectrum. I haven’t seen ChatGPT do anything more useful than write good, quality pastiches. But as Musk said, although he thinks that “on balance, AI will be a force for good”, “the probability of it going bad is not zero per cent”. So it makes sense for governments to try to keep up with it to “safeguard the interest of the public”.

In which case, in normal times, you might expect the British media to say that the prime minister had a good week. You might expect commentators to contrast Sunak’s successful summit with the shambolism on display in the Covid inquiry hearings. You might expect long blog posts about how we British undervalue quiet competence, and berating us for having been taken in by a prime minister who could do jokes. Well, we get a lot of second, but not so much of the first.