Explained: Meta’s Make-A-Video AI system and what it means for users – Times of India

Meta Platforms Inc. has been developing and showcasing new AI and metaverse products for the past few years. The company that owns Facebook, Instagram, WhatsApp and other products and services has now introduced Make-A-Video, a new AI system tool that will allow users to turn their text commands or messages into short, high-quality video clips. This new AI tool can not only make original short videos from text prompts but can also make variations of existing videos. However, it is important to note that Meta’s Make-A-Video tool is currently not available to the public and is expected to be available for end users in the upcoming days.
What is the Make-A-Scene AI tool
The social media giant introduced another AI tool called — the Make-A-Scene AI system in July. Meta has explained that Make-A-Scene is a “multimodal generative AI method” that offers users “more control over the AI-generated content they create.” The company has even demonstrated how users can utilise this tool to “create photorealistic illustrations and storybook-quality art using words, lines of text and freeform sketches.”

How was the Make-A-Video AI system developed
The text-to-image synthesis used with image generators like Make-A-Scene helped Meta to develop the Make-A-Video AI system. Meta didn’t use labelled video data (like describing the depicted actions using captions) to train the new system.
Instead, the company used the image synthesis data (still images with describing captions) and applied it with unlabeled video training data to help the AI model to decode a text or image command. After the model deciphers the respective command, it can predict the behaviour of the image and can display the scene in motion for a short period like Live pictures on Apple iPhones.
What it means for users
From artworks to funny memes, the AI image generation tool already allows users to produce creative content. However, AI video generation is expected to open up more creative possibilities for the users. Meta’s Make-A-Video technology even has the potential (in the upcoming days) to help users to produce an entire movie with just text descriptions and advanced artificial intelligence.

Meta hasn’t shared any details on who will have access to the Make-A-Video AI system or when this technology becomes accessible to the general public, but the company has provided a sign-up form for interested users who are willing to try this technology in the upcoming days.
The company has also addressed the possibility of misusing photorealistic videos to cause certain social hazards. Hence, Meta will also add a watermark in all AI-generated video content made from Make-A-Video that informs the viewers that the video is not a recorded one, but footage generated with AI.