Top AI users hit 3.6x output efficiency as agent usage follows a power law
peterjliu · x · 2026-11-19
peterjliu observes a steep power law in AI usage (notably finance): top users generate more output yet are 3x more token-efficient — "output-maxxing, not token-maxxing." Data shows agent-generated files per credit: since the Oct 5 V2 launch, top 1% users climbed to 3.6x, top 10% to 1.3x, all active users only 0.9x. The differentiator is tooling — those running many parallel, asynchronous long-running agents see explosive productivity gains while the gap widens.
Related event: Financial AI usage follows power law: top users vastly outproduce peers(2 posts)→
More from AGI Musings
- Anthropic Employees Debates Slowing AI as Historic IPO Nears, NYT Reports — JeffLadish · 2026-11-19
- NYT: Anthropic Shows Organizational Fractures Ahead of Historic IPO — kylekabasares · 2026-11-19
- The biggest advantage of a top lab isn't money — it's saving you years — madeofAjala · 2026-11-19
- Proposal: A nonprofit to give visa safety to AI lab whistleblowers — AaronBergman18 · 2026-11-19
- "No innovation ever comes out of the finance department": ex-Covariant researcher on AI ROI — HanchungLee · 2026-11-19
- The case for externalized memory as the real fix for continual learning — daniel_mac8 · 2026-11-19