BritGPT: Should the UK Nationalise AI?
A proposed £11 billion UK national AI initiative, including £10 billion for cloud infrastructure, to boost public-focused AI research in areas like medicine and clean energy, as advocated by Labour for the Long Term.
As we consider how the UK might develop a world-leading national AI policy, an especially interesting idea posed a couple of years ago was the concept of ‘BritGpt‘.
Not the chatbot with the etiquette typical of a British gentleman, but instead a headline brand for a national AI policy.
A thinktank affiliated with the Labour Party, Labour for the Long Term, urged UK Labour leader Keir Starmer to commit £11bn to developing “BritGPT”—a national AI system—and a “Great British cloud” infrastructure, aiming to counter market failures and prioritize public goods like medical research, clean energy, and AI safety rather than direct competition with Silicon Valley.
The initiative would involve £10bn for physical cloud infrastructure, akin to Labour’s proposed Great British Energy, with funding potentially drawn from the party’s £28bn annual climate investment pledge. The report warns that the government’s existing £1bn AI commitment is insufficient, risking greater UK reliance on US tech giants amid escalating AI development costs—such as £10bn for datacentres and hundreds of millions for models like GPT-4.
Benefits highlighted include bolstering British firms, fostering a robust AI ecosystem, and addressing underfunded areas like AI safety, which private entities often neglect for profit reasons. The proposal draws context from AI pioneer Geoffrey Hinton’s cautions on commercial AI risks and examples like Google’s restrained research amid competitive pressures. Haydn Belfield of Cambridge’s Leverhulme Centre stresses AI’s transformative power and the peril of the UK falling behind without bold investment.
