Questions mount over what an AI 'slowdown' would look like
BBC News ● Covered by 33 sources
Top AI bosses now want development slowed down. The hard part is who stops first when China and the US are both racing.
Based on reporting by BBC News — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Some of the loudest people in AI are suddenly talking like brake testers. Anthropic’s Dario Amodei called for a slowdown on Saturday, and Sam Altman of OpenAI plus Elon Musk said they agreed. Then Jacob Coxon, a former Anthropic researcher, told the BBC that staff building these systems were genuinely frightened about humanity’s future.
That sounds like a clean answer to a messy problem. It isn’t. The basic snag is geopolitical: America does not want to hand China an edge, China is not famous for volunteering to come second, and nobody wants to be the first company or country to stop. As Donald Trump put it on Sunday, whoever wins AI wins.
Even if the will existed, enforcement is a swamp. Amodei’s three-point plan calls for independent monitoring of models as they are built, industry-wide regulation and global regulation, but critics say nobody has explained what a slowdown actually means in practice. Ed Zitron of EZ Primary Research said the idea of putting a plant from METR inside every AI company and expecting democratic, global coordination did not mean much to him.
There is also the money problem. AI firms are burning through huge sums and natural resources, yet are not bringing in anywhere near as much revenue. Some users are already disappointed. Economists are talking about a levelling, or maybe a bubble bursting, and that would leave only a few survivors. If that happens, they could emerge as mega-corporations with even more power than the current giants.
In Britain, where AI is being pushed into the NHS, the economy and everyday work and study, a slowdown would not just be a Silicon Valley drama. A former government adviser told the BBC there is no plan B. Others worry that rushing into tighter rules could backfire because nobody can say with certainty how the technology will evolve. The result is a familiar mess: everyone wants safety, everyone wants growth, and everyone wants somebody else to go first.
My take — AI-written commentary, not fact-checked reporting
This is what happens when an industry builds itself into a national security race and then asks for trust on the side. The calls for a slowdown are sensible; the chance of getting one without real rules is close to zero. AI companies have spent years selling acceleration, so nobody should be shocked that the brakes now look imaginary.
Read more about this at: BBC News