Cohere's research director: pre-training ROI is slowing, focus shifts to test-time compute
sarahookr · x · 2026-09-02
Sara Hooker, Cohere's head of research, argues that pre-training ROI is slowing and the focus is shifting to test-time compute (TTC)—compute spent during inference/reasoning rather than training.
The compute profile needed for inference is very different: modern GPUs and TPUs are organized around maximizing matrix multiply performance and delivering data fast for throughput, both of which dominate deep neural networks—and both of these assumptions now change under TTC.
More from AGI Musings
- When AI can clone your product by sight, paid software's moat is service and trust — sujingshen · 2026-10-18
- Anthropic borrows lawyers' ethics for Claude agents that closed 186 deals with no human approval — rohanpaul_ai · 2026-10-18
- Tim Sweeney: AI will unlock a new golden age of games too costly to build before — TinfoilTricorn · 2026-10-18
- Researcher argues we may solve AI-human alignment before humans align with each other — 3scorciav · 2026-10-18
- EigenGender: private-industry coup isn't the default, but worth guarding against — EigenGender · 2026-10-18
- Writing instructor's meltdown: "LLMs are literally just spicy autocomplete" — Decent-Decision-9028 · 2026-10-18