Jack Dorsey's Block Unveils Dessa: An AI Voice Model That Learned to Talk From Sound Alone

Jack Dorsey

Jack Dorsey's financial technology company Block has unveiled Dessa, a new artificial intelligence voice model trained entirely on human conversations — with no text and no pretrained models. Unlike conventional voice assistants that convert speech into text first, Dessa processes audio directly, listening and speaking at the same time.

Block says the full-duplex model can recognise and express emotions, sarcasm, laughter, frustration and sadness, preserving the tone and feeling that text-based systems lose. It can laugh mid-sentence, adjust its volume when asked, and decide for itself what to say and when to respond — behaviours it was never explicitly trained for.

Remarkably, Block says Dessa learned language from hundreds of times fewer words than text-based models, and was trained on a single NVIDIA GB200 NVL72 system. The company is giving a small group of AI safety researchers early access to hunt for risks it has not anticipated, calling a model that learned to talk from sound alone 'new territory'.

Source: read more

Comments

Popular posts from this blog

Fairbairn's Walk-Off Field Goal Seals Wembley Thriller as Texans Edge Jaguars 34-31

Anthropic IPO: Public S-1 Filing Expected Today Ahead of $2 Trillion Listing

Netflix Beats Earnings but Slashes Ad Forecast — Stock Tumbles After Hours