On his social media page, Elon Musk announced the release of an updated version of the Grok Imagine artificial intelligence model designed for creating images and video clips. Now the system more accurately synchronises lip movements with speech and adds realistic sounds. This has brought AI-generated fake videos to a level at which they are almost indistinguishable from real footage. Ixbt.com reports.
In the video clip he posted, Musk particularly emphasised that there is not a single real frame in it, and all images are entirely the product of the neural network. A few days earlier, the ability to animate images and combine them into a single scene was added to the Grok Imagine model. The xAI company has also introduced a function for intelligently continuing video clips.
In addition, the xAI team presented the Grok 4.3 update, in which the main focus is placed on practical working tools. The company also announced that its Grok Speech to Text (STT) API interface is now open to the public. At the beginning of April, Grok Imagine had learned to create high-quality cinematic videos based on short text queries.
Read “Zamin” on Telegram!
