Recently, the video model Sora released by Openai is very significant. It can directly text the video to generate video, which can create a 60 -second video, which contains high detailed scenes, complex cameras, complex cameras Multiple characters of sports and vibrant emotions.
This model is very important. 360 Chairman Zhou Hongyi expressed his view on this. He believes that the emergence of Sora indicates that the realization of general artificial intelligence (AGI) may be possible. It is not far away, and the competition gap between China and the United States in the AI field may be further widening.
Of course, here we will not discuss the competition in the Sino -US AI field.What the author is considering what impact this video generates will bring.
As the saying goes, it is true.As a Homo man, vision is the main channel for accepting information, and therefore, vision is the main channel for our connection with the objective world.At the same time, we have also developed a lot of related qualities. For example, through vision, we will have a lot of emotions, can relax or nervous, and can be pleasant or angry.This is also the core reason for the development of the art and television film industry.
The emergence of the video has indeed triggered a lot of social changes. As early as the industrial era, the role of art was not just to record life and display financial resources. Whats more important is to mobilize people to take similar behaviors through specific art methods, whether worship, Chong Shen or shaping issues.After the industrial age, in the era of television, a large number of videos have also produced many changes.In the 1960s of the US presidential election, the first live broadcast debate was adopted. Kennedy, with a better image temperament, defeated Nixon. People no longer examine the logic behind the political program through newspapers. the difference.
Compared with text and photos, the video can present information more completely, although it can be changed by editing In contrast, videos often have higher credibility.Video is also more attractive than photos, and it is easier to induce human emotions.Specific narrative rhythm and lens language can bring different effects and cause different social emotions.If it is stimulated, the power of the video is more and more common than text and photos.
Madam Chen Lichun, Madam of South Vietnam, is also a typical case
But the content of the video itself is often not easy False, performance requires actors, venues, scripts, directors, later, etc. The cost is not low, and it is relatively easy to identify real and drama.Public events depend on the surveillance video and live videos of evidence, which are authoritative and objective.
Even so, in reality, a large number of short videos have been difficult to distinguish between authenticity.And if the entire video is made of AI quickly, how do we know whether this happened?How to establish trust?Will there be false videos in the future to completely fictional facts, such as false results or accidents, false speeches, and harsh events of false public figures?In contrast, some film and television workers are unemployed and stars have been made.
Therefore, I personally think that the harm brought by the automatic generation of AI videos will be much larger than the large language model of ChatGPT. The text of the ChatGPT is more rational and easier to identify.But humans are too dependent on visual understanding of reality and emotional inspiration.If AI can already produce videos in this way, then human beings may completely believe in completely fictional events, and have a completely irrelevant judgment and action.Once some inexplicable emotions are stimulated by a large -scale, it is likely to bring serious social consequences.
Considering that ChatGPT already has a serious nonsense, creating a case and concept that does not exist at all, SORA has the ability to create some social affairs. Essence
But with human self -control, this problem is afraid of no solution.
