There you go, we AI music videos. It was a certainty, but now it’s a reality.
OpenAI launched Sora - its text-to-video model, a few months back. Since then, it has partnered with many artists and video studios, giving them priority access to the model. One of those artists, Paul Trillo dropped the first officially commissioned video, made with Sora.
How was the video made?
Sora can produce 5-10-second clips from a text input given to it. Trillo made this video by generating hundreds of such clips, selecting 55, and then joining them together. Here is the launch tweet with more details about the production process.
Is Sora publically available?
No. Sora is still in the testing phase.