Alexandr Wang says product ideas often start where builders get blocked

ycombinator · x · 2026-07-27

First-principles thinking at YC Startup School

The post highlights Alexandr Wang’s Startup School talk around first-principles thinking and how it can be discovered through curiosity rather than simply adopted as a slogan.

A concrete example from the talk: while building models at MIT, Wang could easily find training and evaluation frameworks online, but there was no one-click source for data. That bottleneck pushed him to found Scale AI. The takeaway is that being blocked is often the best signal for what to build, because product-market fit usually starts with a real constraint around efficacy, price, or the absence of a solution.

Related event: Alexandr Wang Shares AI Startup Insights at YC(4 posts)→

Original post →

More from Companies & People

Companies & People channel →