New AI can 'think' while you're still talking, just like a person preparing their next sentence.
March 19, 2026
Original Paper
The Silent Thought: Modeling Internal Cognition in Full-Duplex Spoken Dialogue Models via Latent Reasoning
arXiv · 2603.17837
The Takeaway
Most AI systems wait for a user to stop talking before they start processing, which creates an unnatural lag. This 'FLAIR' method simulates human subconscious cognition by running a reasoning process while it listens, allowing for fluid, instant conversations that feel more like talking to a person than a machine.
From the abstract
During conversational interactions, humans subconsciously engage in concurrent thinking while listening to a speaker. Although this internal cognitive processing may not always manifest as explicit linguistic structures, it is instrumental in formulating high-quality responses. Inspired by this cognitive phenomenon, we propose a novel Full-duplex LAtent and Internal Reasoning method named FLAIR that conducts latent thinking simultaneously with speech perception. Unlike conventional "thinking" me