OpenAI reportedly orchestrates 10k internal agents without quality loss, while users' sub-agents 'just burn tokens'
RexDouglass · x · 2026-09-16
- TokenGremlin highlights a contrast: OpenAI allegedly runs an internal system that orchestrates 10,000 agents without losing quality, yet the user-facing version is described as a setup where orchestrating even 3 sub-agents "almost certainly just burns tokens for nothing."
- alzzyd draws a further implication: OpenAI may itself be using those 10k agents inefficiently — a clever researcher could potentially solve problems like Navier-Stokes with a much smaller token budget.
More from AGI Musings
- Anthropic and OpenAI spend only ~1% of capex on AI safety, analysis shows — AlexTensor · 2026-09-16
- Timnit Gebru slams AI safety community as an 'apocalyptic death cult' — examachine · 2026-09-16
- Why aren't labs investing real money into alignment research, Redditor asks — Stabile_Feldmaus · 2026-09-16
- If Safety Work Matters So Much, Why Isn't It Open Source? — beffjezos · 2026-09-16
- Beff Jezos: AGI Power Overconcentration Is the Real Threat, Resist — beffjezos · 2026-09-16
- Gates Foundation Commits $1 Billion to Equitable AI Access in Education, Health and Agriculture — emmanuelvivier · 2026-09-16