Nabeel Hyatt

Harry Stebbings × Nabeel Hyatt

AI is pushing venture from spreadsheet-solvable puzzles back toward first-principles judgment, as no one knows what a model will do next week.Hyatt argues that promotion-driven markups and checklist investing reward short-term packaging, while enduring value depends on founders’ speed, taste, product decisions, and capacity to reinvent.Anthropic’s bull case is a full-stack learning loop connecting models, interfaces, and expert data exhaust, but DeepSeek and rapid cloning leave vertical agents needing hard, decade-long problems and multiple future acts.