When the FCA sent Crowdcube a cease-and-desist, we made our case in the press instead arguing that they were trying to stifle startup innovation. The response was incredible. A week later they put up a white flag reached out to us to ask us to help shape the regulations together, and those rules still govern online retail investment in the UK today. I was part of Crowdcube’s founding team and lead Operations and Commercial raising £1.5bn for thousands of companies including Monzo, Revolut and Freetrade with Crowdcube valued at £140m+. I later co-founded a Lakestar backed Open Banking company and also a Mastercard-backed Gen Z focused fintech-for-good to help young people build better money habits.

None of it was AI safety but it’s a topic I’ve been concerned about since 2017, when I wrote an MBA research paper called ‘The Long View on AI,’ flagging value alignment, the black-box problem and winner-takes-all dynamics as the three things that would matter most. Nine years on, all three look worse, and the labs are sprinting while the institutions meant to govern them are still walking. I’m not anti-AI. I build with these models daily and I’m in awe of their power but I’m a pragmatist who thinks the safety and governance work sits a long way behind the capability, and I worry the path we’re on will not be good for humanity.