Moving Images

As with other AI in recent months, AI video has made extreme strides in terms of quality and usability. While there is still a little work to be done, AI video could soon become an integral part of a production workflow.

While there are a few categories when we talk about AI and video. One, which is already built into some sites and software, is the ability for AI to edit the video. To take clips from existing video and put it together. Automatically removing certain aspect of the videos as specified by the user, combining similar videos, even if shot at different times based on simple text prompts. Not to mention smart color, smart titles and enhanced audio.

But there is another side to the AI video spectrum. Completely AI generated video. This is a video clip that is created completely out of only a text prompt or a still image. The user merely telling the AI what they want to see and AI, using its vast learning, creates a full video.

At the time of this writing, these videos are short (about 5 seconds) and not quiet ready for full production use.

However, also at the time of writing this, several AI companies have announced models that produce much more fluent, longer videos.

Adobe has announced built in tools to its production suite to use AI to remove or add objects into videos all with a simple click and a prompt. What used to take hours of special effect work, can now, or soon, be done in seconds.

AI video will open new avenues for production. Allowing for highly detailed special effects and edits to be done for a fraction of the cost and time. Video producers will be able to spend more time being creative and less time adjusting each frame.

UPDATE: With OpenAI's announcement of SORA, AI video is going to all new places.

READ ALL ABOUT IT