编码人声:DeepSeek 和 OpenAI 都没做到的技术,这群人先实现了 episode artwork

EPISODE · Feb 23, 2025 · 54 MIN

编码人声:DeepSeek 和 OpenAI 都没做到的技术,这群人先实现了

from 津津乐道中国版

目前,哪怕是 DeepSeek 或者 ChatGPT,都是人类说一句 AI 答一句。这种方式很「被动」,AI 只能等你来召唤它。本期节目邀请到声网大语言模型实验室的李忻玮,和实时 AI 创业者史业民。围绕「更主动交互的 AI」的话题,分享包括语音、视频等多模态的主动交互,以及背后更具体的技术细节和路径。如何让 AI 像自然对话那样,适时地主动说话、打断或是沉默?如何实现在你出门时候,AI 主动提醒你带伞?黄仁勋和李飞飞都在做的「世界模型」是什么?它将如何帮助 AI 理解和预测更真实的世界?期待本期节目对你有所启发。

Episode metadata supplied by the publisher feed · Published Feb 23, 2025

Embed this episode

Ready to play

编码人声:DeepSeek 和 OpenAI 都没做到的技术,这群人先实现了

0:00 54:30

No transcript for this episode yet

We transcribe on demand. Request one and we'll notify you when it's ready — usually under 10 minutes.

No similar episodes found.

Frequently Asked Questions

How long is this episode of 津津乐道中国版?

This episode is 54 minutes long.

When was this 津津乐道中国版 episode published?

This episode was published on February 23, 2025.

Can I download this 津津乐道中国版 episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.
URL copied to clipboard!