Mira Murati: Next-gen models never stop listening, chatbots will look like fax machines
Chris_Armstrong · x · 2026-11-07
Former OpenAI CTO Mira Murati argues today's models are turn-based — "while they're thinking they're deaf and blind" — while the next generation will continuously take in audio, text and video and respond in 20-millisecond chunks. Her evolution path: prompt box → chatbot → agents → AI working next to you in real time. The key shift: instead of waiting for you to finish typing, an interaction model sees and hears what you're doing and can cut in the moment you go wrong — but almost nobody is ready to work with an AI that never stops paying attention.
The quoted article, "Harness, Loop and Graph: The Enhanced Edition of Agent Engineering," breaks down why agents pass every test yet fail to fix the issue: the broken part isn't the model or the prompt but the surrounding layers. It dissects harness, loop, and graph — what each owns, how they nest, and which to fix first — built in Claude Code, Codex, and Grok Bot.
More from AGI Musings
- Anthropic study: AI could absorb 27-52% of researchers' time, median 38% across fields — emollick · 2026-11-07
- Nearly 30% of ICLR 2027 reviews mention "appendix", fueling AI-written review fears — nsaphra · 2026-11-07
- "A year or two from now you'll just tell Astra 9 to make your humanoid do it," says SF builder — carlosdponx · 2026-11-07
- Stanford's Anshul Kundaje: ML conference peer review is broken and unfixable — anshulkundaje · 2026-11-07
- Eric Drexler: AI options, not optimism — participants weigh options, not odds — sebkrier · 2026-11-07
- Musk lays out SpaceX vision of deploying up to 1 TW of space-based AI compute per year — XFreeze · 2026-11-07