Thinking Machines Lab, founded by Mira Murati, former CTO of OpenAI, has unveiled interaction models designed to revolutionize the way we interact with AI. Unlike current systems that require a back-and-forth exchange, their new model can process your input and generate responses simultaneously, mimicking human conversation.

Called TML-Interaction-Small, this full-duplex system responds in 0.4 seconds – roughly equivalent to natural human interaction speed, faster than its competitors at OpenAI and Google. However, the technology is still in a research phase, with limited releases planned for later this year.

The idea of native interactivity within AI models is intriguing; it challenges traditional approaches where responses are generated only after input has been fully processed. Critics may question whether this new approach truly enhances user experience or just adds complexity.

Murati's vision is clear: to build an AI that engages with humans more naturally, akin to a real-time conversation rather than the current model of sequential exchanges. This could have profound implications for how we use and perceive AI in everyday life.