Founder Advice and AI Startup Landscape

Mitchell Hashimoto’s new way of writing code

Topic 16 of 17

Published Feb 25, 2026 · Clip 1:50:41–1:55:02 (4:21)

Mitchell tells aspiring founders to account for survivorship bias, imagine 10 years of commitment, and have enough hubris to believe they'll do it better than anyone — without being blind to change. On advising startups, he notes AI startups face unprecedented pressure to move fast, while other questions (remote, open source) remain the same. He sees the engineer role shifting toward capacity to experiment and handle vague tasks — building full demos solo — while productionization feels unchanged; he warns companies blindly copying AI firms' fast shipping. Pre-seed dynamics changed too: you should be able to build the prototype before raising a seed.