> Beyond the initial 2-3 minutes of novelty, what do AI generated videos really have to offer when there is no shortage of people making professional, high quality content on competing platforms?
I don't know where they got September from; Sora launched in Feb 2024[0] which was a bit before people had become tired of awful AI-generated content. There was real belief that people would be willing to spend all day scrolling a social network with infinite AI-generated content. See the similar hype with Suno AI, which started a whole "musicians are obsolete" movement before becoming mostly irrelevant.
I think Sora 2 produced quite good videos, at least of a certain type. It was very good at producing convincing low-resolution cellphone footage. Unfortunately you had to have a very creative mind to get anything interesting out of it, as the copyright and content restrictions were a big "no fun allowed" clause, which contributed to its demise. Everything on the main Sora page was the same "cute animals doing something wholesome and unexpected" video.
My "favorite" part was how the post-generation checks would self-report. e.g. It was impossible to make a video of an angry chef with a British accent because Sora would always overfit it to Gordon Ramsey, and flag its own generated video after it was created!
[0] https://news.ycombinator.com/item?id=39386156 - only one mention of "AI slop" in the entire thread, though partial credit goes to "movieslop".
To nitpick a tiny bit, from Wikipedia[0]:
> In February 2024, OpenAI previewed examples of its output to the public,[1] with the first generation of Sora released publicly for ChatGPT Plus and ChatGPT Pro users in the US and Canada in December 2024[2][3] and the second generation, Sora 2, was released to select users in the US and Canada at the end of September 2025.
[0] https://en.wikipedia.org/wiki/Sora_(text-to-video_model)