Redakcja Choose TV
YouTube is stepping up its AI game and allowing users to create photorealistic avatars that look and speak just like them. The new feature is coming to Shorts and enables users to record videos without actually appearing in front of the camera. All that's needed is a short recording of the user's face and voice, and the system will generate a digital version of the user that can then be used in clips. This is an extension of earlier AI tools from Google, but now a key element of realistic voice is being added.
Creating an avatar is simple, but the effect looks increasingly like a real person
The process is straightforward: you record a selfie video and read a few sentences, and the system builds your avatar. Then you can generate short clips, up to 8 seconds long, that can be combined into longer materials. The most important thing is that the avatar maintains a consistent appearance and voice, which makes it resemble a real person rather than a typical animation. This can be a huge convenience for creators who do not want to record themselves but still want to be "present" in their materials.
YouTube secures the feature but still opens the door to a new kind of content
The platform is introducing security measures for avatars that have AI markings (e.g. SynthID), and only the account owner has access to their creation. Nevertheless, the change is significant, as creating content no longer requires a camera, lighting, or even a physical presence in front of the screen. This could accelerate the production of Shorts and change the way people build their online presence, especially in short formats.
YouTube is heading towards a future where you don't have to record yourself, just your digital counterpart. It's convenient, but also a little unsettling.
source: 9to5google
