OpenAI's Sora turns text prompts into photorealistic videos


It seems generative AI can do anything these days. There are AI image generators, text chatbots, and now videos are being created by AI. Earlier today, OpenAI unveiled its new Sora AI model. But how exactly does it work?

According to the announcement, Sora is a text-to-video AI model that can create videos of up to 60 seconds in length. These videos can feature highly detailed scenes, complex camera motion, and multiple characters capable of showing emotion. Of course, the complexity and detail of a generated video depends on the provided prompt.

An example showcased on Twitter demonstrated a very descriptive prompt used to generate a video of a snowy street in Tokyo. As you can see, it looks like an actual video captured by a camera. It even looked like footage from a drone, given the angle it shows. However, it's quite easy to see how the video could be generated. If you look closely, you'll notice parts of the video where the angles shift and cause objects to fade from view.

Having said that, it's still incredible that video creation is now a domain of AI models. Would it threaten the livelihoods of video editors? Probably not yet, but what do you think? Share your thoughts about the news in the comments below!