# YouTube 正在為用戶新增另一種利用人工智慧混音內容的方式。

*genai · news · 2026-05-20 · Social Media Today*

## Key points

- YouTube 正在整合 Google 全新的 Omni AI，用於進階 Shorts 影片影像生成與混音。
- 觀眾現在可以使用 AI 工具，生成並將自己或物件插入現有的 Shorts 影片中。
- 創作者可選擇退出視覺混音，這同時會移除其影片的任何現有混音版本。
- YouTube 的肖像偵測工具現已向所有用戶開放，以限制濫用和錯誤呈現。
- 由 Omni 支援的新功能擴展了 Extend with AI，允許更具體的影片編輯與混音。

This audio is auto-generated. Please let us know if you have feedback. YouTube is adding another way for users to remix content with artificial intelligence. The platform's new tools will enable Shorts viewers to add themselves, or any other elements they like, into clips in order to build alternative context. Using parent company Google’s new Omni AI image generation tools, YouTube users will now be able to easily generate alternate versions of Shorts videos. As explained by YouTube:“This update makes Shorts remixing easier for both viewers and creators – whether you want to swap in your pet, put a fresh twist on a video with stylistic changes, or add yourself to a scene, these tools let you join trends and conversations in a whole new way.” So now, if a viewer has an idea for an alternate element in a Shorts video, they’ll be able to generate that in-stream. The feature feels somewhat similar to the Extend with AI option that YouTube announced in September, which enabled viewers to generate an alternative ending to a Short via AI prompts. Using Extend with AI provided viewers with a selection of options that they could choose from to generate a different ending to a Shorts clip. Users could then post these alternative variations as their own clips, which is a lightweight way to generate content. This new Omni-powered option is like an expanded version of that, which will enable more specific editing of a Short, and could lead to alternate remix options. Though it may also lead to misrepresentation, which is very likely why YouTube has now expanded its likeness detection tools to all users.The capacity to change the context of Shorts may also lead to misuse, and YouTube is hoping to limit this by ensuring that creators are aware of any remix of their content, while also alerting users to any depictions of their image. Creators can also opt out of visual remixing, which will also remove any remixes already made from their content. So there are controls and failsafe measures built in. Even so, it still feels like a lazy means to create content, which is more likely to flood the app with slop than it is to drive new engagement trends. As with Extend with AI, the bigger question is, why would people want to do this? Sure, there’s no doubt some level of novelty to incorporating elements onto another user’s content, which could lead to absurdist and surreal takes. There will also be some level of creative value in this. But the real challenge of creativity is coming up with a great, original idea, not adding random takes into someone else’s content. So while some users will be excited to insert themselves into YouTube clips, in the majority of cases, these are going to be lazy, unoriginal, uninteresting depictions that only serve to push YouTube further away from original, inspired creation.

**Companies:** Google
**Countries:** United States

[Read the full story on Social Media Today](https://www.socialmediatoday.com/news/youtube-adds-new-ai-remix-feature-to-shorts/820802/)

---

Canonical: https://newsio.io/zh-TW/n/8ca48919-38ca-40d8-99d1-32cac03cca25/youtube-shorts
Summarized by Newsio from Social Media Today. https://newsio.io/how-it-works