• by Laurie Sullivan  @lauriesullivan, December 10, 2024

Advertisers have now gained a text-to-video AI model that enables them to generate videos from text prompts, animate images, and remix existing videos.

Sora, the technology, debuted from OpenAI for ChatGPT Plus subscribers as a “turbo” product, and officially launched as part of its "12 Days of OpenAI" event.

OpenAI describes the visual generation tool as bringing "your imagination to life."

The company initially teased the model earlier this year, and on Monday demonstrated the tool in a video posted to YouTube (see above). 

OpenAI CEO Sam Altman explained why it is important to make tools for creatives, and said the world had been missing something important with the AI systems only being able to generate text. 

"We want our AIs to understand and generate video," Altman said. “It will change the way we use computers.”

The tool is available today in the U.S. and most other countries. It is a completely new product experience for video generation.

OpenAI expects this early version to make mistakes, but in the long run it will become very useful to augment projects.

Developers at the company highlighted the “storyboards” feature that allows users to generate videos based on a sequence of prompts and timelines with the ability to turn photos into videos. OpenAI also demonstrated a “remix” tool that enables users to change Sora’s output with a text prompt that can blend two scenes together with AI.

The storyboards can be seen as a way for the entertainment industry to preview any type of story for movies or television sequels, or a tool to give an author the tools to think out a storyline in detail. It will not have the ability for moviemakers to make an entire movie, for now.

Exploration -- part of the experience -- is a product section on the site with video feeds shared by the community to show what is possible. It is intended to share inspiration and knowledge on how the videos were created.

Videos generated with Sora will have visible watermarks and C2PA metadata to indicate they have been made with AI.

Before uploading an image or video to Sora, OpenAI prompts require the creator to click a box for certain safeguards to ensure that the video does not feature anyone under the age of 18, and that it does not contain any explicit or violent content or copyrighted material.