Observer: Moonshot's Compute Instincts Belong at Google, DeepSeek's Rookies Built the Most Efficient LLM
teortaxesTex · x · 2026-10-24
X user teortaxesTex argues Moonshot's talent, starting with its founder, has instincts about compute utilization "more appropriate inside Google," and notes that Wenfeng chose to hire mainland novices who "built the most efficient LLM known to man." The quoted thread adds that Moonshot has data, architecture research talent and large-scale training/serving experience, and debates whether a cheaper K3.5 pivot would just be outdone by the next GLM. A substantive take on Chinese lab talent density and the efficiency gap with US labs.
More from Companies & People
- Tesla starts production of Optimus humanoid robots, targets 20,000 units per week — LexiLove · 2026-10-24
- Ben Horowitz on winning as a founder: AI lowers the barrier, apps may stop mattering — a16z Podcast · 2026-10-24
- OpenAI's Berlin DevDay community hackathon kicks off, with a non-technical founder track — paw_lean · 2026-10-24
- Use Your Own Org Data to Predict Slow Cases Before Customers Complain — louisdorard · 2026-10-24
- Baidu's Apollo Go and Lyft's Freenow unveil RT6 robotaxi in London, targeting 2027 driverless launch — Baidu_Inc · 2026-10-24
- Paris DevDay community hackathon kicks off, 100-person call crashes chat — paw_lean · 2026-10-24