ai has continued to have an impact on Hollywood. Among them, the new AI ML released by OpenAI in February - sora has attracted much attention. With the advancement of large-scale models, film and television screenwriters have also gone on strike due to common demands, which profoundly reflects the increasingly significant influence of AI in the creative field. Likewise, the film industry is increasingly concerned about such technology.
Tyler Perry, a Hollywood super producer and director who owns one of the largest studios in the United States, became the first victim of sora. This plan has been in preparation for four years. It is expected to greatly increase the size of the location and build 12 new sound stages. However, everything has come to an abrupt end now.
Compared with past AI video production tools, sora has the ability to generate one-minute high-quality videos based on prompt words and realize lens switching and composition adjustment functions. At the same time, it can perfectly display background details related to the theme. The generated videos have A high sense of reality, just like a reflection of the real world. With sora, are there changes in the future of the film industry? Hollywood is in turmoil.
Hollywood is facing severe pain: sora storm
Recently, 's plan for twelve buildings costing approximately US$800 million and covering an area of 330 acres was quickly halted due to sora. This is the conclusion reached by the famous American director and actor Tyler Perry after observing the performance of sora. He said: "sora will fully cover our industry."
For example, sora can show in a video work that a group of elephants are Running in the woods, this uncanny ability to understand the timing of complex instructions brings it one step closer to a tool truly suited to storytelling.
On the other hand, although openai admitted that sora still needs to be improved in handling complex occlusion problems, Paul Trillo found that even under constant motion, the text in the video generated by sora remained consistent, and This shows that sora may not only rely solely on diffusion effects, but also integrate three-dimensional animated environments and special effects elements.
It is understood that a research report by cvl economics in January this year revealed that in interviews with 300 senior executives in the entertainment industry, up to 75% of the respondents believed that AI technology may reduce or even replace many traditional jobs, especially the impact to sound engineers, voice actors, visual effects producers, and other post-production workers. predicts that this trend may affect nearly 204,000 employee roles in the next three years.
In 2021, Disney established a dedicated artificial intelligence group to explore the use of AI to reduce the cost of original film and streaming content production, and apply it to its entire industry chain. In the same year, netflix, a large visual effects studio, also officially hired artificial intelligence expert . They seek individuals with extensive experience in neural network image generation, familiarity with the principles of stable diffusion ecosystems, and a deep understanding of cue word engineering. It can be seen that Netflix is also actively deploying AI in visual special effects production.
Since the emergence of sora, there have been some similar AI video generation and editing models in the industry, such as runway and pika. They can generate short clips from text prompts and 2D images, create animations from still images, repair, replace and add visual elements, and even spread to external space beyond the scope of the original frame, thereby presenting "pictures outside the frame" and enriching video content. . Such various improvements have made people's demands for video richer and more diverse.
In fact, some Hollywood companies have applied runway technology to projects and won Oscar awards. Looking at sora with , it further improves and breaks through the existing level, including improving the quality and depth of the video to make it close to the real environment; in addition, the length of the video can also be up to 1 minute, fully demonstrating the content covered by the text. information. In comparison, Sora's competitor Runway also launched its second-generation model in August this year, but its longest generation time was only 18 seconds.
"AI's Hollywood" may not be smooth sailing
Although sora, a powerful AI tool, has the potential to use it to generate movies. Although these videos can be spliced into a movie, make sure the characters and scenes are consistent.However, if the video derived from text prompts and conditional parameters cannot maintain the consistency of the character in the next scene, it will not be able to generate a longer-lasting "artificial intelligence story."
Currently, the widespread application of sora in Hollywood is facing many challenges. In addition to coherence issues, this model often misunderstands real-world shapes and behaviors, resulting in significant "physical deviations" in image and video output. Not long ago, the sora model "rollover accident" that was hotly discussed on the Internet, such as chairs suddenly moving, characters "inexplicably missing" and other phenomena not only made people laugh, but also highlighted the difficulties faced by sora in the process of making movies.
On the other hand, copyright legal issues are also an important factor hindering the application of sora in Hollywood. Presenting AI-generated materials on the screen may cause copyright disputes. Even if the AI is trained based only on relevant materials, there is still a risk of potential infringement.
Recently, openai has not released sora to all users. Some people in the industry praised its infinite creativity after experiencing it, while also pointing out its loopholes and risks of abuse. is similar to the deepfake incident last year. If Sora is fully opened to the outside world, it may also produce a large amount of false information and rumors, causing harm to society. When Google released the AI video model lumiere in January, it also stated frankly: "While using this technology, there is indeed a risk of fabricating false or harmful content and suffering unfair use."
Sophia feels that despite this, sora will be fully open to the public and social media in the future After the media creators open up, a large number of video works will surely emerge in this field. However, not everyone is as clear about copyright-related issues as creative or advertising professionals, and relevant policy supervision is still very necessary.
