Kimi CEO's Advisor Russ Salakhutdinov: No Secret Architecture in Frontier AI, Moat Is Data and Engineering
rsalakhu · x · 2026-08-01
In the Forward Deployed podcast, Russ Salakhutdinov, PhD advisor to Kimi CEO Zhilin Yang and professor at CMU, shares deep insights into AI. He studied under Hinton, sold his startup to Apple, and worked at Meta's Superintelligence Lab. He reveals that frontier AI labs have no secret architecture; the real moat is data, engineering, and infrastructure. He predicts LLMs will become commodities and notes that many startups like Cursor are quietly running on Chinese open-source models. He also discusses why computer-use agents still hit only 60% success (when 99.9% is needed) and whether AI can beat prediction markets.
Related event: Kimi CEO's Advisor Says No Secret Architectures in Frontier AI(3 posts)→
More from AGI Musings
- A missing parameter in scaling laws: H, the human effort behind learning signals — lateinteraction · 2026-10-16
- First-ever paper on multi-agent swarm safety lands NeurIPS spotlight — LauraRuis · 2026-10-16
- Yoav Artzi on peer review: paper flooding still pays off as review system strains under load — yoavartzi · 2026-10-16
- Letta founder: work in fields defined by problems, not methods — lateinteraction · 2026-10-16
- Why pure mathematicians will soon flood into AI x-risk debates — Dr_Atoosa · 2026-10-16
- RSI began when model outputs augmented next-gen training data, agents just escalated it — akbirthko · 2026-10-16