Block's Dessa Voice Model Sounds Eerily Human

Jack Dorsey's Block has released Dessa, a full-duplex speech model trained entirely on human speech using a single NVIDIA GPU rack. Unlike most voice AI, Dessa skips text in its pipeline, embedding emotion, pauses, laughs, and natural cadence directly into the model. It can talk and listen simultaneously and run locally on a MacBook M4 Max or better. Block claims current voice AI discards emotional data by converting speech to text first. There is no public access yet; Block is seeking safety researchers to test the model before wider release.