cHHillee: MFU tuning isn't a moat, but demand aggregation in ML infra is
cHHillee · x · 2026-09-22
cHHillee pushes back on the idea that kernel-level MFU gains are a significant advantage, arguing ML infra instead benefits from demand aggregation—the most efficient setup for a model like Kimi K3 might involve PD disaggregation plus wide EP. Patrick Toulme agrees GPU renting is a moat but disputes that MFU-tuning software is one, since future models may need far less tuning depending on how RSI plays out. cHHillee adds that large-scale RL posttraining infrastructure is resilient to future AI improvements, requiring substantial GPU-hours and token investment to replicate.
More from Companies & People
- Dev argues Meta, not OpenAI, will bring AI agents to mainstream users — _nateraw · 2026-09-22
- AI Slop Is a Productivity Tax: Execs Say Enterprises Are Drowning — sebpaquet · 2026-09-22
- Notion hires Nathan to build a space where people and agents think together — daniel_mac8 · 2026-09-22
- Parent Moves Daughter to Alpha School: 2 Hours of AI-Driven Academics a Day — LamarDealMaker · 2026-09-22
- Shopify CEO Warns of "Slop Grenades": Unreviewed AI Output Creating Work for Everyone — ATTlKA · 2026-09-22
- Lenny Rachitsky on managing Nikita Bier: book him on Intro just to get a 1:1 — lennysan · 2026-09-22