After 75+ early-stage investments — including first cheques into companies now worth over $1bn — I've watched the profile of a winning founder change more in the past three years than in the previous ten. The best companies I see today aren't companies that use AI. They're companies that would be unrecognisable without it.
AI-first is an operating model, not a product description
Plenty of pitches have AI in the product. The founders I back have AI in the company — in how they ship, sell, support, and decide. Their cost structure and tempo look nothing like the last generation's. A ten-person team can now do work that took many times that headcount when I was scaling Uber and Revolut, and the founders who have internalised that don't hire their way out of problems. They build systems.
International ambition from day one
The second thing I look for is international ambition from the start. AI has collapsed the mechanical costs of expansion — localisation, support, market research — so going global is no longer a Series C project. It's a founding decision. The AI-first companies that win won't dominate a home market first and venture out later; they'll be built for many markets from the first commit. That was the hardest, most expensive work we did at Uber and Revolut. It's now available to any seed-stage team that chooses it.
What hasn't changed
The moat still isn't the model. Distribution, workflow depth, switching costs — the boring, compounding advantages — decide who wins, same as ever. So I still ask the old questions. Why will this customer still be here in five years if a better model ships tomorrow? What's the unfair advantage that can't be bought?
Where I spend my time
I work directly with a small number of these founders each year, across fundraising, commercial scale, and international expansion. The pattern is consistent: the technology is rarely the constraint. The scarce skills are narrative, commercial discipline, and knowing which market to enter when. Operator problems, in other words — the ones I've spent 25 years on.
Every platform shift mints a new generation of category leaders, and this one's are being founded right now — AI-first, global from day one. My job is to find them before the rest of the market agrees.