Jack Dorsey 旗下 Block 发布语音对话模型 Dessa，仅用人类对话训练
Jack Dorsey’s Block made an AI model that talks to you like an awkward new friend
- 1Block 发布语音对话模型 Dessa，从零只用人类对话训练，先向少数 AI 安全研究者开放
- 2官方称其为真正的语音到语音模型，可在 MacBook Pro M4 Max 或更高配置本地运行
- 3官方展示的样例对话自然但不时出戏，如认错说话人、聊电影时突然跑题
- 4官网样例中 Dessa 与 OpenAI GPT-Live-1 语音模型对话显得更自然
Block 的 Dessa 绕开文本中间层直接做语音到语音，主打情绪与节奏感知，但目前仍停留在官方演示与安全研究者内测阶段。
- 影响：语音到语音路线直接用语音训练，省去转文本环节，官方称能跟上人类节奏并感知情绪；模型仅向少数安全研究者开放，企业现阶段无法试用或采购，现有样例均由官方挑选。
- 参考：可考虑整理现有语音客服在情绪识别和应答节奏上的失败案例，形成内部基准；等安全研究者或独立评测给出反馈后，再评估纯语音模型是否值得试点。
- 依据：语音到语音、仅用人类对话训练、M4 Max 本地运行等说法均为官方口径，暂无独立验证；展示样例由 Dessa 团队挑选，早期访问仅限一小群 AI 安全研究者。
Jay Peters
is a senior reporter covering technology, gaming, and more. He joined The Verge in 2019 after nearly two years at Techmeme.
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”).
Here’s a video of Dessa playing a guessing game over voice:
The Dessa website features a few other examples, like a back-and-forth about Christopher Nolan’s films (including Dessa going on an unexpected tangent about Bryan Singer’s Superman films), a person telling Dessa about taking it on a surprise trip to Paris (to which Dessa, as the website describes, “then answers with what sounds like teary gratitude”), and a conversation between OpenAI’s GPT-Live-1 voice model and Dessa (where Dessa comes across as more natural). You can hear clips from some of those conversations in this video.
On the website, the Dessa team says that Dessa is a “true speech-to-speech model, and everything it knows came from human speech,” and it can run locally on a MacBook Pro M4 Max “or better.” As for why they made it, the team gives the three reasons “curiosity (aka fun), better understanding between humans and AI, and safety,” and provided some more detail:
Babies learn to talk without ever seeing a word written down. We wanted to see if a model could too. It can. It’s just very hard.
Speech is the most natural interface to any intelligence, super or otherwise, and today’s voice AI wastes it. It flattens what we say into text and can’t keep up with human timing.
AI keeps gaining power and independence, and it’s still nearly deaf to the emotion in our voices. Imagine being in serious pain while a voice AI has no idea and just keeps going. That’s the safety problem. Dessa is a step toward fixing it.
The Dessa team says it will be giving “a small group of AI safety researchers” early access to the model.
Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.
- Jay Peters
来源：The Verge · AI · theverge.com