Introducing Sora, Open AI’s new artificial intelligence model. OpenAI, whose founding partners include Elon Musk, became famous with ChatGPT, an artificial intelligence model that answers questions asked to it. The company is now here with its new artificial intelligence model, Sora, which can produce videos with text commands.
By understanding text commands, Sora can produce highly realistic and creative video images up to 60 seconds long. More than one character can appear in the video images produced by Sora. It also uses accurate details about the desired scene or subject in the images it produces.
OpenAI company states in its statement that Sora’s understanding of language is very deep, so he can produce interesting characters that can express vivid emotions. In addition to producing videos with text commands, Sora can take an existing image and enhance it or create a new video by filling in the missing frames. The company also emphasized in its statement that Sora will serve as a basis for models that can better understand the real world and imitate it.
OpenAI also stated that work on Sora continues and that the artificial intelligence model has weaknesses, especially in the cause-effect relationship. As an example of this, a video produced by Sora was shown. There was no bite mark on the cookie in the video where a bite was taken in the next scene.
The company also stated in its statements that Sora was tested to ensure that it did not produce harmful or inappropriate content before being made available to the general public. The model is currently only available to a small group of visual artists, designers and filmmakers to get feedback on how it can be improved to be most useful.
The scene of a woman wearing glasses walking on the streets of Tokyo in the evening, which we see in the video below, was produced by Sora. You can watch other videos produced by Sora on the company’s own website.
Views: 329
