Cursor CTO Reframes Bezos' Two-Pizza Rule for AI Teams

Business Insider reported that Cursor field CTO David Pan wrote on X that Jeff Bezos' two-pizza rule should be revised for the AI era. The article says Pan called the original metaphor influential for engineering organizations, but argued that two pizzas can now represent too large a team when AI tools increase individual throughput. Business Insider also notes that Pan previously worked at Amazon and that he recently joined Cursor as field CTO.

AWS's own two-pizza team explainer defines the model as a team small enough to be fed by two pizzas, ideally fewer than 10 people, with less communication overhead and stronger single-threaded ownership. That context matters because Pan's comment is a critique of a heuristic, not evidence of a universal staffing rule. AI may reduce some coordination costs, but it can also increase the need for review, telemetry and governance when smaller teams ship more automated work.

The useful operating lesson is to update team design around interfaces and accountability. Smaller AI-enabled teams need clearer service ownership, stronger tests, better incident paths, shared prompts or agent policies, and explicit review thresholds for generated code. Otherwise, fewer people can mean faster output with weaker institutional memory.

Watch whether AI-native software companies formalize smaller squad structures around agentic development workflows, and whether larger enterprises copy the pattern or keep two-pizza teams as a ceiling. The measurable signal will be delivery quality, incident rates and customer-facing velocity, not the pizza metaphor itself.