Miles Brundage: Codex ultrafast mode is overkill for most tasks vs running more in parallel
Miles_Brundage · x · 2026-10-01
Former OpenAI policy lead Miles Brundage discussed compute economics with @herbiebradley, who said he used to regularly exhaust his Codex budget on standard mode, but the new fast/ultrafast modes now feel excessive for most tasks—spending the same FLOPs on parallel work is a better deal. Brundage replied that "excessiveness is relative to the amount of FLOPs available per dollar," framing the tradeoff as a pure cost-per-compute question.
Related event: Codex Ultra Fast Mode Burns Through Budgets, Sparking Compute Divide Debate(4 posts)→
More from Models
- Researchers flag AI "delusional spiraling": sycophantic models amplify users' false beliefs — QuintinPope5 · 2026-10-01
- You.com, NVIDIA and CoreWeave bring live web search into RL training, starting with Nemotron 3.5 Lightning — RichardSocher · 2026-10-01
- Gemini 4 Argon Posts Strong Result on AA Coding Agent Index via Antigravity — prateeky2806 · 2026-10-01
- Anthropic: GLM-5.3 builds Chrome exploits nearly as well as unreleased Claude Mythos, with lax safeguards — matthew_d_green · 2026-10-01
- Full Artificial Analysis Intelligence Index results for Solar Mini 4 — ArtificialAnlys · 2026-10-01
- Upstage's Solar Mini 4 scores 24 on AA Intelligence Index, but costs ~5x GPT-6 Luna per task — ArtificialAnlys · 2026-10-01