Insider: AI post-training data business booms despite having no moat
himanshustwts · x · 2026-11-15
A job-seeker in post-training roles explains why the AI data business keeps booming despite seeming moat-less: "quality" is hard to define, labs judge data by eye or via vendor-trained small models, and the same data can work well at one lab but poorly at another. Reposter adds labs often don't know what they actually want — "it is a race."
More from Companies & People
- Why No One Can Replicate Jensen Huang's Asian GPU Supply Chain — _onionesque · 2026-11-15
- We Now Have a Spacecraft That Can Take Us to Mars and Bring Us Back — XFreeze · 2026-11-15
- Why Did Intelligence Officials Wait a Year Before Flagging Jensen Huang Rumor? — kevinnbass · 2026-11-15
- Burkov Releases 'The Hundred-Page Language Models Book' With Hands-On PyTorch, From $25 — burkov · 2026-11-15
- Researcher maps Jensen Huang media storm: 2,644 stories, 79 retracted lies, open data repo — kevinnbass · 2026-11-15
- Thorsten Ball breaks down Amp's only meeting: the one-hour Monday alignment ritual — Register Spill (Thorsten Ball) · 2026-11-15