Anthropic's internal AI 'spend' exceeds payroll only at retail token prices; 90% of new code is Claude-written
petrusenko_max · x · 2026-11-03
An operator pushes back on the viral claim that Anthropic's internal AI usage exceeds payroll: it only holds when self-consumed tokens are valued at retail prices, not actual spend, so citing it as proof of inference costs mis-models how bills work.
The signal worth tracking, the author argues, is that 90%+ of the code added that month was written by Claude — which reshapes code review workflows and hiring needs.
Related event: Anthropic's internal AI usage reportedly exceeds payroll at retail prices(2 posts)→
More from AGI Musings
- Prime Intellect resident poses the RL frontier problem: how do you reward unverifiable knowledge work? — xeophon · 2026-11-03
- Tim Lee: Pretraining is imitation learning, post-training is reinforcement learning — binarybits · 2026-11-03
- Robot sims should prioritize accurate physics over visuals, turning sim creation into an auto-research loop — wellecks · 2026-11-03
- Top-tier mechinterp research is 'dual use', researcher argues in AI safety circles — cephaloform · 2026-11-03
- LLM tools are mass-converting academic CVs into narrative-style prose — MarlosCMachado · 2026-11-03
- Causal ML decision-support system for acute heart failure enters first-in-human trial — ShalitUri · 2026-11-03