Thinking Machines Lab Unveils "Full Duplex" AI Model, Challenging Industry Giants
Mira Murati's new AI startup, Thinking Machines Lab, has released a research preview of its latest innovation: a model called TML-Interaction-Small that aims to redefine real-time conversation with artificial intelligence.
"The model operates in 'full duplex,' allowing simultaneous input processing and response generation, similar to a phone conversation."
Unlike traditional AI models that require a user to finish speaking before generating a reply, this new system can process incoming information and produce a response at the same time. This design is intended to create a more natural, uninterrupted dialogue flow.
Blazing Fast Response Times
The company is making bold claims regarding speed. According to the press release, TML-Interaction-Small delivers a response time of just 0.40 seconds.
This latency is reportedly comparable to the pace of natural human conversation and is faster than the current offerings from major competitors like OpenAI and Google.
From Research to Reality
It is important to note that this is a research preview rather than a commercial product launch. While the company has published benchmarks suggesting superior performance, real-world user experience has not yet been validated by the public.
What to Expect Next:
- Limited Research Preview: Expected within the next few months.
- Wider Public Release: Planned for later this year.