The discontinuation of the video generation AI “Sora” was surprising news for many. Why did such a revolutionary technology disappear so quickly?

In short, this is not simply a service shutdown. It’s an event that revealed the “unique nature of the video domain.”

First, Sora was technically superior. It could generate high-quality videos from text, allowing anyone to easily create “professional-looking videos.” This meant democratizing video production, which previously required specialized skills. However, what awaited them was a wall that technology alone could not overcome.

One was the cost issue. Video generation is overwhelmingly more computationally demanding than text or image generation. As usage increased, costs ballooned, making it difficult to sustain as a business.

Another was the issue of copyright and ethics. Video is not just data. It’s “meaningful content” that includes people, stories, and their relationship to reality. This isn’t simply a matter of technological innovation. Video has the power to influence society and move people’s perceptions and emotions. That’s why it requires more caution in its handling than other media.
If AI-generated video resembles an existing work, is it a coincidence or an infringement? If it generates faces or voices similar to real people, is it expression or a problem? Society and the law still don’t have clear answers to these questions.
Even more important is the issue of “truthfulness.” Unlike text and images, videos “appear real.” In other words, AI-generated video always carries the risk of being mistaken for “fact” by viewers. This is precisely a problem that television has grappled with for many years. That’s why television stations continue to be cautious with AI.
Fact-checking.
Editing responsibility.
Broadcasting ethics.
Only after going through these processes does video become “something that can be released to the world.”
Sora’s shutdown is not a defeat for AI. Rather, this incident demonstrated that the field of video is a far more “responsible medium” than one might imagine. And from this, a paradox emerges.

Precisely because we now live in an era where anyone can create video thanks to AI, the weight of the question, “Who will release that video to the world?” has increased. In other words, the value of video is shifting from “Can it be created?” to “Can it be released responsibly?” From this perspective, television is by no means an outdated medium.

Rather, it is beginning to possess value as the “last filter” in the AI ​​era. In an age where anyone can broadcast, who will take responsibility?

The end of Sora confronted us with this question. And the responsibility for that judgment rests not with technology, but with humans. This is simultaneously an act of protecting the very credibility of the media. And it leads to the future.

※Translating Japanese articles into English with AI