Thinking Machines Lab, led by former OpenAI CTO Mira Murati, has introduced "Interactive Models," a research preview designed to revolutionize voice AI interactions. Unlike traditional models that operate on a turn-based system, these new models can generate responses while simultaneously receiving user input, creating a more natural, phone call-like conversation experience. This "full-duplex" technology allows the AI to continuously understand context and prepare responses, with the TML-Interaction-Small model achieving a response time of approximately 0.40 seconds. Currently, the technology is in a research preview phase and not yet available for public use. Thinking Machines plans a limited research preview in the coming months, with a broader rollout expected later this year. The company aims to integrate interactive capabilities directly into the model, potentially transforming the use of voice assistants and conversational AI products. However, real-world performance and user experience remain to be tested once the product is publicly released.