Jack Dorsey's payments company Block · has unveiled a voice AI model that sounds remarkably human in company demonstrations, prompting reactions ranging from amazement at its conversational abilities to warnings that such technology could make phone scams harder to detect.
Called Dessa, the model was trained entirely on human speech rather than text and can listen while speaking, allowing it to handle interruptions, laughter and changes in tone during a conversation, according to the Block’s announcement.
“Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence,” Dorsey wrote on X.
In one demonstration, Dessa laughs in the middle of a sentence; in another, it reacts to a surprise with what sounds like genuine emotion. Block says the conversations were recorded live without anyone directing the model's responses.
Unlike systems that first convert speech into text before generating a spoken response, Dessa processes audio directly. Block said it trained Dessa from scratch using a single Nvidia GB200 NVL72 computing system. The model is compact enough to run locally on a MacBook Pro equipped with an M4 Max chip or better, according to the firm’s statement.
For now, Dessa remains a research preview rather than a publicly available product.
Dessa’s demo quickly drew attention on social media, with some users praising how natural the interactions sounded.
X user Alexander Hoff said he had never “heard anything artificial that sounded so… human,” while user Hesamation described it as “straight up sci-fi stuff.
Source link
