coindesk.comarticleBlock's new voice AI sounds eerily human, sparking awe and scam fears
Dessa is a voice AI model from Block, trained from scratch on human speech only, with no text stage. It listens while speaking, so it handles interruptions, laughter and tone changes. Block says it trained on one Nvidia GB200 NVL72 system and that it runs locally on a MacBook Pro with M4 Max or better. Model size, benchmarks and release date are undisclosed. Early access goes to a small group of AI safety researchers. Reactions split between praise and scam fears.
TL;DR Block previews Dessa, a speech-native voice model that skips text, runs on an M4 Max MacBook Pro, and drew both awe and scam concerns.
Organisations named here
- Block
- Square
- Cash App
- Bitkey
- Proto
- Dessa
- NVIDIA
- CoinDesk
- Binance
- GiggleAcademy
- Apple
- Changpeng Zhao
- Jack Dorsey
- Justine Bateman
Technologies named here
- Dessa
- Nvidia GB200 NVL72
- MacBook Pro with M4 Max chip
- Apple M4 Max
- MacBook Pro
Techniques named here
- Direct audio speech modelling
Related in models
