The AI Wave Will Create Enormous Roadkill

David Frankel, founder of Founder Collective, has been investing at seed stage for 18 years—through internet, SaaS, mobile, and now AI. He was the first check in Uber, Kupong, SeatGeek, PilPack, Olo, Shield AI, and Suno. In conversation with Harry Stebbings, he explains why the current moment is unlike anything before, why he refuses to raise a larger fund, and what most people misunderstand about seed investing today.

Why Seed Is Crowded but Not Dead

Frankel describes seed as "commoditized" and "incredibly tough." The problem is structural: funds in the $50–100 million range are too big to write $200K collaborative checks but too small to lead $8–10 million seed rounds. They get squeezed from both sides. Meanwhile, multi-stage firms are using their growth funds to write large seed checks as "call options"—insurance policies that give them a seat at the table without deep commitment.

But Frankel argues seed remains viable for disciplined firms. His data shows that of the top 500 companies created in the last 25 years, the median exit is $2.6 billion. Owning 5% of one such company returns an entire fund. "You can wait and wait and wait, and then you just see someone—a founder or a team—and you go, I have to be there." That conviction, he says, is the drug that makes early-stage investing work.