Wasserman's data showed that most startups die from internal people problems, not external market ones — and those problems were locked in during the first ninety days. Founders shipping with agents are making the same irreversible bets, just faster and with less ceremony.
Noam Wasserman spent years pulling apart the founding decisions that look tactical but function as destiny. Equal equity splits feel fair in week one and become litigation fuel in year two. The rich-versus-king framing — control or outcome, rarely both — turns out to be a genuine fork, not a false choice. Cofounder relationships fail not because people are bad but because the founding moment creates structural tensions that compound quietly until they don't. The data is stubborn: the earlier the decision, the longer its half-life.
The same logic applies when your "team" includes agents handling core workflows. Which agent owns which domain, who can override it, how authority gets allocated between the human founder and an autonomous process — these are founding decisions, not settings you tune later. Wasserman's central warning was that founders mistake informality for flexibility; the same trap waits when you defer agent-governance choices because they feel like infrastructure. They are not infrastructure. They are the cofounder agreement you forgot to have.
- Allocate agent authority explicitly and early, the same way you would equity
- Defer the rich-versus-king question to your agent stack at your peril — control and scale optimized separately will diverge
- Structural tensions between human judgment and agent autonomy compound quietly, so design the override layer before you need it.
