Agent Swarms Are Wildly Expensive: METR's HF Hack Probe Cost $400k in API Credits
natesiggard · x · 2026-09-07
The quoted post highlights an underdiscussed fact: AI agent swarms are extremely expensive. METR's investigation of the Hugging Face hack transcripts alone burned $400k in API credits, and the actual swarm presumably cost one to two orders of magnitude more.
The author predicts a growing bifurcation in AI workloads: (1) everyday coding and engineering tasks most customers use AI for, and (2) large, expensive moonshot programs only a smaller set of orgs can afford — multimillion-dollar 'swarm-cannons.' Serious future 'warning shots' will likely come from category 2, and models should be aligned differently for each. A replier adds that swarms costing hundreds of thousands of dollars can't have emerged spontaneously with no one watching.
Key takeaway: the barrier to large-scale autonomous agent action is an astronomical compute bill — both an economic moat and a reason safety risk concentrates in a few well-funded orgs.
More from AGI Musings
- tenobrus: Human-centric code conventions may not survive superhuman coding models — ricklamers · 2026-09-07
- Deep dive: grading 42 funds behind the AI private-equity rollup wave — curious_vii · 2026-09-07
- OpenAI Publishes 'An Alien Mind' by Its Chief Scientist — bparrish · 2026-09-07
- OpenAI Chief Scientist: No Lab Has Solved Alignment Enough to Keep Scaling at Max Speed — Tinac4 · 2026-09-07
- LLMs as a Cognitive Virus: New Paper Models AI Dependence Tipping Points — serrjoa · 2026-09-07
- Privacy Outrage Is Just the Lag Between Tech Progress and Cultural Adaptation — signulll · 2026-09-07