It is a beautiful drone(无人机) shot, popular kind of video you might see in a travel video. However, there is no drone camera, and you needn't travel in person because the video was generated by AI, which is from a new tool programmed by Open AI called Sora.
It's not perfect, but it's a large step up from what we've seen before. What most people don't know is Sora can do more than just create videos from scratch. It can combine separate videos into one scene. And its videos can be up a minute long and in 1080p resolution.
(compare) with Runway ML, the difference with Sora is that it's coherent. Previous system have a (character) morphing(图像渐变) quality as the video (program). With Sora, that's vastly reduced or gone altogether. The (object) remain stable even when covered by things in the foreground. We've seen something similar to this in research since 2019, but what is new is the ability (combine) two videos together in one scene. It can also make up different camera angles of a single scene with just one prompt.
Sora, like digital audio workstation (envelop) recording studios several years ago, must be a trend soon.
