Dessa, a speech-to-speech AI model developed by Block, is capable of producing more natural and human-like speech by processing audio directly without transcription.

Block, the payments firm formerly known as Square, has released Dessa, a speech-to-speech AI model trained from scratch on nothing but human conversation. It handles audio directly instead of transcribing to text, so it can listen and speak at the same time, giving replies a more natural cadence and fewer of the robotic pauses common in voice assistants. Block says it runs locally on a MacBook Pro M4 Max and that a small group of AI safety researchers already have early access. https://gizmodo.com/check-out-the-many-uncanny-human-ish-voices-from-blocks-new-ai-model-2000830237 #AI #GenAI #AIresearch

Block unveils Dessa, a "true speech-to-speech" AI model trained on "nothing but human conversation" that it says runs locally on a MacBook Pro M4 Max or better (Jay Peters/The Verge) https://www.theverge.com/ai-artificial-intelligence/1017906/jack-dorsey-block-dessa-ai-voice-speech http://www.techmeme.com/261024/p5#a261024p5