Thinking Machines wants to build an AI that actually listens while it talks
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Thinking Machines Lab—founded by former OpenAI CTO Mira Murati—announced “interaction models” that enable full‑duplex AI conversation.
- TML‑Interaction‑Small delivers responses in 0.40 seconds, a speed the company says matches natural human dialogue and beats comparable OpenAI and Google models.
- Thinking Machines Lab plans a limited research preview of the model in the next few months, with a broader release slated for later this year.
- TechCrunch reported on the launch, highlighting the model’s ability to interrupt users and likening the experience to a phone call rather than a text exchange.
Why it matters: Enterprises seeking near‑real‑time conversational AI gain a faster option, while OpenAI and Google risk losing market share as Thinking Machines pushes a 0.40‑second full‑duplex model toward broader release later this year.
