- Meta’s new tool makes videos with just a few simple commands.
- All videos produced by Make-A-Video will have a watermark.
- Creating films from text has historically been difficult.
According to Meta CEO Mark Zuckerberg, the company has built an AI that can create films using simple verbal commands. This AI has been utilised to create a new tool for creating art.
The CEO called the tool’s results “very excellent” and “extremely astounding.”
Make-A-Video, the system’s artificial intelligence, adds its awareness of motion in the actual world to traditional text-to-image synthesis.
According to Zuckerberg’s comments, it creates a video for you after you offer a text description.
Creating video has always been extremely difficult since the AI must not only produce pixels but also predict how they will change over time.
This problem is overcome by introducing “unsupervised learning” as a feature.
The IT behemoth has not yet revealed when the tool would be made available to the general public or whether there will be any prerequisites, such as signing up for the system or subscribing. The company does pledge to offer demonstrations soon.
Given the possibility for disinformation and deep fakes, there are always moral quandaries surrounding AI.
As a workaround, Meta suggested putting a watermark to Make-A-Video videos.
For the time being, we will continue to scrutinise, test, and evaluate Make-A-Video to ensure that each stage of distribution is secure and thoughtful, according to Meta.