A British DARPA to fund high-risk, high-reward scientific research

ARIA (the Advanced Research and Invention Agency) was a good idea, underfunded and overcautious. The model it was based on, DARPA in the United States, has a track record that speaks for itself: the internet, GPS, mRNA vaccine technology, and autonomous vehicles all trace their origins to DARPA-funded research.

The key to DARPA is not just funding. It is programme managers with the autonomy to make big bets without committee approval, and the freedom to fail without career consequences. The British version needs the same. Hire exceptional scientists, give them budgets, let them fund the most ambitious ideas they can find, and accept that most of them will fail. The ones that succeed will more than pay for the rest.

Fund it at £1 billion per year, which is still a fraction of what the US invests. Protect it from Treasury meddling and departmental interference. The returns from breakthrough science are enormous but they take time, and they require a tolerance for failure that the British public sector has never been comfortable with.