Equity Splits That Made Sense Then
Wasserman documents a recurring trap: co-founders splitting equity evenly to avoid conflict, then discovering they have locked in a permanent ambiguity about who decides when things get hard. In an AI-native company the stakes of that ambiguity are higher. Agents can ship product, draft strategy, and process customers at speed — but none of that motion resolves the question of who owns the judgment behind it. Unresolved authority at the founding level does not stay abstract; it surfaces the moment two founders disagree about what the agents should be optimizing for.
Rich vs. King in a Company Agents Help Run
The rich-versus-king tradeoff Wasserman names is the choice between maintaining control and maximizing outcome. Founders who choose control tend to build slower, hire later, and keep more equity. That calculus shifts when agents compress the time between idea and working prototype to hours. A founder with taste can now carry a product further alone — which means the temptation to delay co-founder conversations, board structure, and authority design is stronger than ever. Delaying those conversations does not make them cheaper. It makes them more expensive, because by then the agents have already been shaped by one person's defaults.
Judgment Is the Role That Cannot Be Distributed
Wasserman's data shows that the founding team's composition predicts outcomes more reliably than the idea itself. The reason holds in the agent era: software can now be directed in natural language, the bottleneck moves from typing to thinking, and what remains irreducibly human is deciding what to build, for whom, and where to stop. That is not a task you delegate to an agent or resolve by adding a co-founder late. It has to be owned explicitly, by a named person, from the start. The model writes the code. It does not settle the founder's dilemma.
