Another sky, 'Sora'.

A few filmmakers were asked to test out OpenAI's new generative video model, Sora, when it was unveiled last month. The results, which the business released this week, are seven bizarre short films that demonstrate how quickly generative video is evolving.

The first wave of text-to-video models debuted in late 2022 and came from startups in the video-tech space, Runway, Google, and Meta, among others. Although it was a clever approach, the results were short, pixelated, and had a grainy quality.

After a whole year, the finest of Sora's high-definition, photorealistic work has become so impressive that some rapt spectators are calling for the demise of Hollywood. The newest models from Runway can create short films that are on par with those produced by major motion picture companies. The companies that created two of the most well-liked text-to-image models, Midjourney and Stability AI, are now focusing on video as well.

Many businesses are vying for market share as a result of these discoveries. The majority are discovering that business on the fly. When using these tools, Gary Lipkowitz, CEO of Vyond, a company that offers a point-and-click platform for creating brief animated films, says, "I'll regularly scream, 'Holy cow, that is wicked cool." "How can you use this at work, though?"

Regardless of the response to that query, it will undoubtedly disrupt a broad spectrum of industries and alter the responsibilities of several experts, ranging from animators to advertising. Concerns about abuse are also increasing. It will be simpler than ever to flood the internet with propaganda and nonconsensual porn, thanks to the ubiquitous capacity to create false videos. We are able to predict its arrival. The issue is that there isn't a viable solution.