Startup Advice from Sam Altman — YC Lecture Series

Startup Advice Briefly" and the associated YC lecture material are Altman's most condensed early-career writing — produced during his time as YC president and encoding the mental models he used to evaluate thousands of companies. His most influential framework: "make something people want" is a necessary but insufficient condition; you also need to want to do the thing yourself (so you don't stop when it gets hard) and you need to understand why now is the right time (market timing). His tactical advice for early-stage founders is unusually specific: launch as fast as possible because your intuitions about what users want are almost always wrong and fast iteration beats better planning; hire people who are slightly above your current needs because you'll be hiring for the company in 18 months, not today; and talk to users more than you think you need to because most startup failure comes from building something that isn't actually useful. Altman's YC period produced the "Do Things That Don't Scale" doctrine (via Paul Graham, but operationalized at YC during Altman's presidency), the Launch YC program, and the shift toward global batch diversity. His observations about what patterns predict success — missionaries vs. mercenaries, internal vs. external locus of control, genuine vs. performed passion — are the observational foundation of his later writing about what makes people successful at OpenAI. The lecture series is where Altman's pattern recognition was most directly articulated.