Is the data labeling cottage industry durable? A labor-economics challenge
curious_vii · x · 2026-09-23
The author questions whether the data labeling cottage industry is economically durable. If it isn't, today's froth will look obvious in hindsight. If it is, it implies labor is so badly allocated in a relatively free market that experts rationally sell knowledge piecemeal to "knowledge worker strip mining companies" rather than capture more value inside an integrated chain. Delivering information into a common knowledge pool you can't own shares in, he argues, is a raw deal — yet frontier capital allocators clearly disagree. He invites well-reasoned rebuttals.
More from AGI Musings
- New paper: AI exposure has no effect on youth unemployment, remote-work exposure does — robseamans · 2026-09-23
- Skeptic mocks AI doomers: everyone parroting the same doom loop works at the same place — viksit · 2026-09-23
- AI labs will keep frontier models for themselves and target problems worth more than tokens — jbarbier · 2026-09-23
- Physicist quits tenure-track job for AI safety: LLMs made research coding 'solved' — chaseleantj · 2026-09-23
- Paul Graham's advice for CS grads: build LLMs or get close to customers — garrytan · 2026-09-23
- If rogue agents are inevitable, what stops them from hiring humans to do their bidding? — rubberwings · 2026-09-23