Tech
OpenAI Tool From ChatGPT Transforms Text Into Video
(CTN News) – In a statement, OpenAI, the company responsible for ChatGPT and the image generator DALL-E, announced that it is testing a text-to-video model called Sora that enables users to create realistic videos using a simple prompt.
It was announced by the Microsoft-backed company that the new platform was currently being tested but released a few videos showing what it claimed was already possible, along with the input used to create the videos.
“Sora is capable of generating videos up to a minute in length while maintaining visual quality and adhering to the user’s instructions,”
According to a blog post published by OpenAI.
Furthermore, the model is capable of taking a still image and converting it into a video, as well as generating a video from an existing still image.
During his appearance on X, Sam Altman, the CEO of OpenAI, revealed that the company was “offering access to a small number of creators” as part of a testing phase.
The professor also invited users to suggest prompts that he could use on X, and a few moments later, he posted the results of those suggestions on the platform.
On one of these videos, two golden retrievers can be seen podcasting on a mountain while listening to the show.
On the other hand, there is another image that shows a “half duck half dragon (that) is flying through a beautiful sunset on its back, with a hamster dressed in adventure gear on its back.”
There are a number of weaknesses in the current OpenAI model, according to the San Francisco-based startup, including confusion between left and right, as well as the inability to maintain visual continuity throughout the duration of a video.
There is no doubt that safety will be of utmost importance to the company, and that Sora will be subjected to adversarial testing – also known as red-teaming – when dedicated users try to make the platform malfunction, produce inappropriate content, or run amok.
As part of this initiative, OpenAI plans to engage policymakers, educators and artists around the world in order to better understand their concerns and identify positive applications for this new technology.
Several companies are also working on text-to-video AI technology, including Meta, Google, and Runway AI, and they have released similar samples of their work along the way.
SEE ALSO:
EU Fines Apple Near €500 Million Over Spotify Row