AI data labeling gold rush: Mercor founders, 22, become youngest self-made billionaires as data goes stale at 70% accuracy
FinanceYF5 · x · 2026-10-10
A thread by FinanceYF5 breaks down why AI training data is the only product that expires the moment it works:
- The paradox: once a model hits 70% accuracy on a task, labs drop the task (@EpochAIResearch) — "once you've trained, you've trained" (@athomasq).
- The money: Mercor's three founders became the youngest self-made billionaires at 22 (Forbes); Surge's Edwin Chen never raised a dollar and is worth $18B; Meta paid $14.3B for 49% of Scale AI and took its founder; dead startups' old Slack chats sell for $10k–$100k.
- Buyers building in-house: Google hired Mechanize's founder and team; xAI cut 500 generalist labelers.
- Supply-side contamination: LoRA co-creator Edward Hu joined Mercor, whose first model is built on Alibaba's Qwen; a 2023 study found a third+ of crowd workers used AI to write "human" answers; the next teachers: factory workers with head cameras.
More from AGI Musings
- GPT-6 and Opus 5 beat Montezuma's Revenge, but Metaculus's AGI bet still isn't resolved — emollick · 2026-10-10
- From ICLR volunteer to top CS PhD: Jeande's full-circle meeting with Sasha Rush — Jeande_d · 2026-10-10
- Mathematicians push back on 'solved problems = progress' as AI tests the metric — _onionesque · 2026-10-10
- Sayash Kapoor: Why Superhuman AI Won't Transfer to Every Task — sayashk · 2026-10-10
- Debunking the claim that AI systems persist nothing across interactions — TinfoilTricorn · 2026-10-10
- COLM Best Paper: Int-Bench Shows LLM Assistants Intervene Too Early and Too Often — maxhkw · 2026-10-10