In a sunlit room in El Segundo, two investors who have spent years betting on technological upheaval tried to answer a question that now haunts nearly every venture capitalist and founder: how do you invest when the ground is shifting beneath your feet?

Carter Reum of M13 and Chang Xu of Basis Set Ventures agree that artificial intelligence has created a market unlike anything they have seen. Xu describes it as both a bubble and not a bubble. Revenue curves that once looked extraordinary now seem pedestrian. Companies can leap from idea to tens of millions in annual recurring revenue in a couple of years, with tiny teams and positive cash flow. In that world, sky-high valuations can be rational, but only if investors resist the temptation to assume every startup will follow the same trajectory.

Reum argues that the pattern is familiar, even if the slope is steeper. Previous waves — the car, the iPhone, the cloud — also triggered panic about jobs and power. What is different now, he says, is that today’s upstarts are not just competing with one another, but with the most capitalized, technically sophisticated incumbents in history. For once, the giants may actually have the edge: they own the data, the distribution and the talent. That makes picking winners harder, and the downside of being wrong more brutal.

Both investors have adapted by tightening their frameworks. Xu focuses on what is technically defensible, above and below the AI layer. Infrastructure built for humans is being rebuilt for AI agents, from databases to version control. At the application layer, she looks for products with deep, durable differentiation rather than thin wrappers around foundation models.

Reum, meanwhile, looks for friction as a moat. Regulated sectors like emergency services and healthcare move slowly enough that hyperscalers will not rush in for relatively modest outcomes. He also urges founders to operate with a microscope in one eye and a telescope in the other: execute ruthlessly week to week, but constantly scan for shifts that could render a business obsolete overnight.

Xu divides markets into velocity and depth. In velocity markets, speed of execution is everything and fast followers are relentless. In depth markets, biology, regulation or physics impose hard limits; there, patient, technically ambitious teams can still build enduring moats.

For both, the most compelling opportunities lie not in today’s obvious AI plays, but in the second and third ripples — the strange ideas that do not look like businesses yet, in sectors where taste, culture and creativity matter as much as code.