Selling data to AI labs is like selling to the government: few buyers, brutal RFPs
himanshustwts · x · 2026-10-05
- Quoted tweet: selling data to AI labs mirrors government contracting — a handful of buyers, vendor vetting via researchers and shared Slack channels, endless RFPs, competing against 20 companies per project, but loyal repeat business once you're credible (until research directions shift or a better vendor displaces you).
- The main post argues data work should be treated as applied research rather than ops-heavy execution: in the commodity-model era, labs have already compressed uncertainty; with broad, complex data and good intuition, you can find the real bottleneck and whether solving it shifts model capability or economics. Differentiation and learning loops are what's hard to replace.
More from AGI Musings
- Should AI tell you if a core belief is psychologically necessary but false? — PierceLilholt · 2026-10-05
- AI agents are about to flood the workforce, and no one's ready: WIRED — Dr_Singularity · 2026-10-05
- Eric Horvitz Revisits His 2001 Human-AI Vision: Core Capabilities Now Exist — erichorvitz · 2026-10-05
- Sam Altman: the biggest mistake talented people make is not thinking about what to work on — _AustinCalvert_ · 2026-10-05
- AI decompilation era: PS5 is 80% ported to PC and 'closed source' is ending — almmaasoglu · 2026-10-05
- The Curve talk to tackle acceleration emergency and superintelligence policy agenda — jachiam0 · 2026-10-05