Google has introduced a new feature for YouTube creators that was announced some time ago: the ability to generate avatars, synthetic versions of the humans who have traditionally featured in their videos. In this way, any user will soon be able to create Shorts in which the main actor is an avatar, or include it within an already created video.
And I’m not sure that this is going to be a good idea.
As Google itself explains, with this measure—which will be rolled out gradually—the intention is to “facilitate and improve accuracy when adding your image to your videos”. Avatars create a digital version of yourself so that you may generate videos that look and sound like you, “in a secure and confidential way”. To create your AI avatar, you will be able to use the YouTube mobile application or do it directly from the YouTube Create app. As YouTube clarifies, you will only need to be of legal age and own a preexisting YouTube channel.
To prevent confusion, all videos featuring these types of avatars will include visible watermarks and digital labels such as SynthID and C2PA to indicate that the content is generated by AI.
Of course, a somewhat questionable scenario may have occurred to you: could you generate an avatar using someone else’s image and create videos with it?
The short answer is no. Thankfully.
YouTube assures that only you will be able to use your avatar to create original videos; you can delete it whenever you wish, and all resulting content will be labeled as AI-generated. To guarantee this, the social network has enabled a secure live selfie capture process, in which you will record your face and your voice to create your avatar. You will only need to create it once, whether from the YouTube application or the YouTube Create application.
Once your avatar is ready, you will simply need to write your prompt, with which you can generate a video where your avatar appears. To follow the example given by YouTube, something like “I am flying high in the sky, above the clouds, saying I can fly!”
Wonderful.
Additionally, you may also create your avatar and add it directly to a preexisting Short. To do this, you just need to tap on the “remix” and “reimagine” features. Then, you simply have to select “add myself to this scene.” You can regenerate or delete your avatar at any time. Once you have uploaded videos featuring your avatar, you may choose to delete the video or restrict others from modifying it.
Moreover, YouTube states that both avatars and videos are managed individually. In other words, if you delete your avatar it will not remove the videos where it appears. Similarly, deleting a video with your avatar will not delete your avatar from your account.
It seems evident that with this move, YouTube—that is, Google—is seeking to boost video production by giving creators the opportunity to generate a video at any time, regardless of illness, travel, lack of time or, perhaps, simple laziness, which might otherwise keep them from pressing the record button. This obsession with content volume may be understandable at a time when short formats are booming—the same ones that have allowed the platform to withstand the pressure from TikTok in this format.
However, I am not convinced this is the best path forward, and it is contradictory to the philosophy that Google maintains for all its content, whether text (as seen with core updates in SEO) or video: favoring authenticity, credibility, and the personal touch.
On the other hand, greatly increasing the volume of content on YouTube will obviously have a clear impact, as we have already seen in the case of Instagram. Most likely, the metrics for each video will decline. Quite simply, users’ viewing time is finite, and even if it gradually increases, it is unlikely to be sufficient to be distributed among so much new content.
This move also arrives shortly after OpenAI decided to terminate its Sora initiative and its AI-generated microvideos, partly due to the need for Sam Altman’s company to focus on its core business and partly because of the negative repercussions caused by the ease with which hyper-realistic videos—impossible to distinguish from reality—could be created.
YouTube has learned from this lesson and, to avoid ending up like Sora, has implemented SynthID and C2PA (invisible watermarks and metadata) from the very beginning. Basically, Google is saying: “We are launching this, but we are going to make it so clear that no one can claim we are deceiving them.”
Image: Gemini
Your email address will not be published. Required fields are marked *
Δ