Britain should embrace the AI revolution

Rishi Sunak’s big speech this week was easily lampooned. Having accused Keir Starmer of ‘doomsterism’, the Prime Minister warned that Britain’s most dangerous years lay ahead, and talked of the threat from ‘colluding authoritarian states’.

Less attention was paid to the part of his speech about artificial intelligence, which was in fact genuinely optimistic. As well as bringing greater freedom, choice and opportunity, AI could double productivity ‘in the next decade’, he said.

Imagine, he went on, a world in which every teacher is free to spend more time with struggling students, and in which a single picture of your eyes can not only detect blindness but also predict other diseases such as heart failure and Parkinson’s. Given the pace of technological innovation, this is entirely plausible.

OpenAI released its new model this week, GPT-4o, which responds to voice, images and videos, not just typed commands. The first iteration of ChatGPT was met with amazement and then panic: how long might it be before the bots lead us into war? The reaction this time was closer to impatience. If AI can be used to treat cancer, Alzheimer’s or Parkinson’s, surely we should be doing everything we can to accelerate the process.

The scenario Sunak posited – that AI doubles productivity in a decade – would transform British society. It could also lead to the fastest uplift in wages in modern history. The UK’s notoriously stagnant productivity is a trap which means the average worker is being paid no more, in real terms, than 15 years ago. It didn’t help that Sunak’s furlough scheme, welcome at the time, slowed the vital process of economic generation. This has kept more workers bogged down in unproductive companies that would otherwise have gone bust.