Sora is definitely the focus of recent attention from relevant people and interested parties in the technology field. This large artificial intelligence video model launched by OpenAI, the manufacturer of ChatGPT, has the advantage over similar products on the market that it can produce one-minute long videos with accurate subjects and detailed details based on the user's prompt words, and can present complex Camera movement and multiple characters with continuity.
OpenAI announces first video generation model.
OpenAI claims that “the model understands not only what the user has asked for in the prompt, but also how those things exist in the physical world.” things exist in the physical world) and treat it as a "world simulator".
As a simulator, Sora still has many limitations. For example, when it comes to the physical interaction of characters or animals, it is difficult not to make mistakes, but I believe its work team will gradually optimize and solve these problems.
When Sora appeared on the scene, voices that denounced the film and television industry naturally appeared. Just like when ChatGPT first came out, public opinion predicted that journalists, writers, screenwriters and other literary work would be doomed. "Sora has brought a biting chill to the film and television industry." When seeing these comments, the reporter felt that it was time to have a casual chat with my friends who are film and television practitioners.
Early stage of film and television: Everyone has a creative drawing board?
"In the field of film and television pre-production, Sora will become a transformative force." Ms. Zhou, a senior storyboard artist, said seriously. Tools like Sora can help film and television creators use highly efficient composite videos to tell their ideas. With the use of Sora in the early stage of film and television, the era of carefully hand-drawn storyboards and concept drawings may be over.
But as Sora realizes the automation of concept presentation in the pre-production process of film and television, are the job prospects of traditional storyboard artists, concept designers, etc. really bleak? In Ms. Zhou's view, this is not necessarily the case. In her understanding, the use of Sora is still based on the prompt written by the user. Whether the prompt is written accurately determines whether the video generation can accurately convey the concept, which plays a decisive role. It still works on people, so its tool properties have not changed.
"Since it is a tool, what changes is the way of creation, not the life of the creator." Ms. Zhou said humbly that Sora can make the storyboard dynamic, which is something she can't do as an artist for the time being. "But for conceptual design, I think excellent conceptual design will not be easily replaced by tools such as Midjourney and Sora for the time being, because their work has high personal creative attributes."
Regarding the prediction that “in the near future, there will be massive unemployment in the early stage of film and television,” Ms. Zhou shook her head. She believed that tools like Sora will inevitably bring challenges to the early stage of film and television. “But any job must face challenges and keep learning. , maybe in the future, early-stage workers with strong learning willingness and ability can integrate tools like Sora into their work processes, and even come up with more ideas to improve work efficiency and work quality through learning programming, large models, algorithms, etc. method. As for practitioners who have stopped making progress, if AI does not eliminate them, they will also be eliminated by better and hard-working people, right?"
"In a broader sense, I am actually looking forward to having a tool that can break down the creative barriers brought by skills such as painting and animation. Anyone with ideas can use tools like Sora to bring good ideas to their mind. I think it’s a good thing to be more visual and communicate better with the management and production side.” Although Ms. Zhou is confident about the prospects of the industry, she said she will not be in the film and television industry next year and will join MCN (for the Internet). A multi-channel network providing services to video creators. "But the reason why I left this industry was not because I was afraid of being replaced by Sora, but because I was fed up with the stupid xing of Party A," Ms. Zhou lamented, "Maybe AI would be good for Party A."
Mid-term film and television period: Is there no substitute for "star power"?
The middle stage of film and television usually refers to the filming stage of film and television works. Video generation tools such as Sora do not seem to have much room for use at this stage. "No matter how vivid and beautiful the faces generated by AI are, currently the audience still only recognizes celebrities, and management and platforms only pay production fees for celebrities." Young executive producer Xiao Lu said with a smile, "My boss (chief producer) In the past, when I was most bothered about deciding on the cast (deciding the cast), I had to beg my grandfather and grandma to deal with some difficult A-list (stars), and I had to drastically change the script according to their requirements. But now it seems that the stars may be traditional film and television The future 'moat' of the industry."
Reporter asked: What if celebrities grant portrait rights to the AI platform? Maybe the filming of film and television dramas can be directly generated using a large number of celebrity portraits? Xiaolu was stunned, "Isn't it possible? The legal and economic risks involved are too great, and it will be difficult to implement in the short term. And by then, celebrities are afraid that they will be replaced by AI, right? Unless they use their portrait rights to generate video buttons Commercial charges apply."
But what if even the definition of celebrity changes? Even as early as ten years ago, AI idols, anchors, singers, etc. had appeared in large numbers on Chinese and foreign Internet platforms. With the further development of AI technology and Internet social platforms, it may not be impossible for several AI superstars to unify global aesthetics. By then, will AI stars also use generative video technology like Sora to present more possibilities for film and television drama creation?
Xiaolu listened and then added, "If the future really becomes like this...are the performance awards at Chinese and foreign film and television festivals still meaningful? Will AI stars participate in the competition?" His words stopped the reporter.
The first episode of the sixth season of "Black Mirror" features celebrities and ordinary people whose images have been licensed
Film and television post-production: elites calm, grassroots carnival?
Film and television post-production is a broad general term, which includes editing, special effects, color correction, dubbing and other types of work. Among these types of work, many practitioners currently believe that the one with the most commercial potential for Sora is special effects.
However, Junjun (pseudonym), who is engaged in editing and has recently tried to switch to special effects production, believes that Sora will not threaten medium and large-scale film production in the short term. "Why do special effects have to be done frame by frame? Because the human eye is It is something that is very difficult to deceive. And if you look at the development trend of global film special effects in the past two decades, you will know that they are becoming more and more sophisticated, and they are becoming more and more refined, because the audience's aesthetic and excitement points are getting higher and higher. If Once your project is determined to use the spectacle of special effects as its selling point, it determines that you must invest enough money, energy and manpower in special effects.”
"But in low-budget films, especially independent films, I am quite optimistic that generated videos like Sora may bring new creative possibilities to young independent filmmakers and even grassroots creators." And in his view, although Currently, the videos produced by Sora are up to one minute in length, but he predicts that with OpenAI’s iteration speed, Sora will not be far away from producing continuous videos that are tens of minutes long.
Does that mean that aspiring but uninvested young filmmakers will no longer need expensive equipment and teams to realize the special effects of their works? Maybe just by writing a prompt to Sora, they can realize the gorgeous fantasy short film they want? "According to some of the generated videos I've seen so far, there are quite a lot of errors. The level of surprise can't be compared to the frame-by-frame surprises of our special effects production, so it can't be said to be 'gorgeous', but it provides a very likely possibility. Zero-cost creation tools and support for various visual styles allow creators to try out whatever they want, rather than being limited by money and time costs." He added, "But you can't rely on AI to generate the entire work. , maybe it can be realized technically in the future, but in that case, is this considered your work, or is it considered an AI work? Does your creation only provide a prompt word?"
"Generally speaking, I am very optimistic that lowering the technical barriers to special effects will liberate the productivity and creativity of more creators." Junjun said, "But it is too early to say this. Sora has not yet been opened to the public, and there are still The copyright issue of AI-generated videos is currently almost a legal blind spot around the world."
Facial details rendered by Sora
User-generated content (UGC): Will film and television be “finished” sooner or later?
"Sora's ability to quickly generate high-quality videos is consistent with the current fast pace of social media consumption." Annie (pseudonym) is a staff member of an MCN agency in Hangzhou. She has long been responsible for the video content planning of many Internet celebrities under the company. "Rather than playing a role in the film and television industry, I am more optimistic that tools like Sora will shine in short video content. After all, the film and television industry is already a sunset industry."
Sora will allow individuals to create video effects and content that only professional filmmakers can accomplish like never before. Sora will undoubtedly exacerbate the ambiguity between UGC and filmmaking content. "The most exciting thing about this technology is that it allows everyone to share their unique perspectives with the world. This is a decentralized era where everyone has the desire and power to express themselves. Tools like Sora can enable everyone to share their unique perspectives with the world." Groups can share their experiences and perspectives more vividly and effectively, even if their expressions are not part of the dominant narrative. Will this make this era more inclusive and diverse?" Promote empowerment and self-actualization for all , there is no doubt that UGC has a very serious social mission in Annie’s eyes.
The reporter reminded her that if she wanted to use Sora to tell her own experience, would it be unavoidable to give AI access to my private materials, such as home videos and personal photos, in the future? Are there huge privacy leak risks and social threats? Regarding this question, Annie said that this chaos is not in the future, but has already happened, "because there is no privacy on the Internet, and big data is always grabbing privacy." Her expression was very categorical. Regarding copyright risks, many negative cases can already be seen on UGC content platforms such as Douyin and Kuaishou. For example, AI face-changing fraud, rumors, etc.
OpenAI, like several other generative AI companies, has faced multiple lawsuits accusing it of using copyrighted material to train its products. OpenAI reports that they will be taking some important security measures before officially launching Sora and taking precautions to prevent the technology from being used to create misleading content. But whether all this will work for the best, who knows?
With the help of AI, will a new genre of film and television be born?
When he learned that short video practitioners said that the film and television industry was going to be over, screenwriter Ronaldinho scoffed: "Short video makers are a bit forgetful of their ancestors when they say this. How much technology did you steal from film and television drama creation and production in the first place? How many films and TV shows have you taken from it? Doing commentary on a work and making a second creation to make a profit is a crime?"
“I don’t mean to look down on short videos. I also use Douyin and Kuaishou myself, but to be honest, it is the most difficult for most short video content creators to establish barriers in front of AI-generated videos. Nowadays, many Internet celebrities may be able to be AI can be easily replaced." Ronaldinho said firmly, "On the contrary, in the film and television industry, its core content creation is the hardest to be replaced, and that is text creation."
In Ronaldinho's view, Sora's current "text-video" generation method has determined the decisive role of text creation in its production. "Similar to the gameplay of Midjourney, you are playing with the accuracy of text expression. If you want AI to generate the video you want, you must constantly calibrate the accuracy of the text prompts." Ronaldinho said frankly, "The importance of text creation is even more prominent. So in my opinion, in the entire industry at the moment, we screenwriters are currently the most difficult job to be replaced.”
"The birth of film and television art is the product of technological progress. Historically, the invention of any technology has brought opportunities to the film and television industry: from the film age to the digital age, from 2D to 3D." He believes that Sora is like the film and television industry Like any technological revolution in history, it will improve production efficiency, update production, and may even create new types, styles, and genres.
"First of all, AI-generated videos have changed the meaning of media. Perhaps one day, we will doubt that any video is generated and no longer believe in what we see - the change in audience psychology may bring new ideas to film and television creation. Maybe." Xiao Luo said, "In addition, we are currently talking about Sora more as a 'tool', but if we regard it as a 'world simulator' claimed by OpenAI, maybe one day, the more mature Sora World-building capabilities can open up unimaginable vistas for visual storytelling, unleashing countless voices to tell stories never imagined before."
Like it or not, we stand on the edge between science fiction and reality. Like all technological advancements, in the future, AI-generated video tools like Sora will be a double-edged sword, bringing hope and danger to the field of film and television production. It may break the status quo and reshape the industry. The most risky and Those who learn the spirit may hold the key to unlocking the huge potential of AI tooling and shaping the future of the film and television industry and other fields.
