Jack Dorsey's Block made an AI model that talks to you like an awkward new friend
Block's Dessa is a full-duplex speech model trained entirely on human conversation, not text—a novel architecture bet.
- 01Block unveiled Dessa, a speech-to-speech AI model built from scratch on human conversational audio rather than text corpora.
- 02The full-duplex design enables simultaneous speaking and listening, closer to real dialogue than turn-based voice assistants.
- 03Early demos show mostly natural exchanges but also notable quirks—the model misidentified a user as "Ronaldo." For professional-services and CFO-suite leaders evaluating voice AI for client intake, advisory or managed-services delivery, Dessa signals a maturing but still rough frontier in conversational AI infrastructure.
Block's Dessa is a full-duplex speech model trained entirely on human conversation, not text—a novel architecture bet.
Block unveiled Dessa, a speech-to-speech AI model built from scratch on human conversational audio rather than text corpora. The full-duplex design enables simultaneous speaking and listening, closer to real dialogue than turn-based voice assistants. Early demos show mostly natural exchanges but also notable quirks—the model misidentified a user as "Ronaldo." For professional-services and CFO-suite leaders evaluating voice AI for client intake, advisory or managed-services delivery, Dessa signals a maturing but still rough frontier in conversational AI infrastructure.
Watch: Whether Dessa's speech-native architecture produces measurably lower error rates in domain-specific professional conversations compared with text-first models—that gap will determine enterprise services relevance.
Jack Dorsey's Block announced an AI model called Dessa that's trained "from scratch on nothing but human conversation," and to show off the model, Block shared chats it has had with humans that can occasionally be a little awkward and uncanny. Take this one , for example, shared by Dessa lead Ragavan Thurairatnam. The conversation between the model and the speaker does have a mostly natural back and forth (though for some reason, the model initially mistakes the speaker for somebody named "Ronaldo").
Read the full article at theverge.comJack Dorsey's Block announced an AI model called Dessa that's trained "from scratch on nothing but human conversation," and to show off the model, Block shared chats it has had with humans that can occasionally be a little awkward and uncanny. Take this one , for example, shared by Dessa lead Ragavan Thurairatnam. The conversation between the model and the speaker does have a mostly natural back and forth (though for some reason, the model initially mistakes the speaker for somebody named "Ronaldo"). We made the first AI model you can have a realistic conversation with that was trained entirely on speech! Dessa is a full-duplex speech-to-sp … Read the full story at The Verge.
Don't miss tomorrow's
The Daily Pulse in your inbox each morning — sourced and linked.