Dev predicts 10k-100k tok/s inference will make frontier models real-time robot policies
eigenron · x · 2026-09-07
eigenron predicts that within 1-2 years, fast inference on future chips like WSE-5/6 or next-gen Etched silicon will make general-purpose frontier models viable as real-time robot policies without task-specific training — roughly GPT-8 at 10k-100k tok/s per stream. He tested something recently that blew his mind and argues latency, not raw throughput, is the bigger bottleneck.
More from AGI Musings
- OpenAI computer-use agents near superhuman speed, say the team building them — a16z · 2026-10-26
- Kai-Fu Lee publishes AI Native, a roadmap for building AI-native companies — kaifulee · 2026-10-26
- RCT: ChatGPT temporarily boosts consultants' data science skills, gains vanish without it — EricTopol · 2026-10-26
- Dev on AI games backlash: reject AI slop, not the tool itself — ezshine · 2026-10-26
- Worry less about AI than about the bad actors wielding it — zakkohane · 2026-10-26
- If an agent can rewrite itself, does your consent still mean anything? — tallmetommy · 2026-10-26