Paris, France – AI video generation is poised to become a new favorite in the advertising industry, offering a low-cost, high-efficiency alternative to traditional video production methods. This trend is exemplified by the recent release of OpenAI’s Sora, a tool that allows users to create high-resolution videos in minutes by inputting prompts into platforms like Runway or Synthesia.
Josh Kahn, a sports entertainment filmmaker who has created content for LeBron James and the Chicago Bulls, has been experimenting with AI video tools to envision the future of the Olympics. Using the latest version of Runway, Kahn created a one-minute video showcasing a futuristic Los Angeles for the 3028 Olympics. The video features a rising sea level, a football stadium on top of skyscrapers, and a beach volleyball court in a central port.
While the video is more of a demonstration of AI’s potential than a blueprint for urban planning, it highlights the significant potential of AI video generation in the advertising industry. Companies and content creators can leverage this technology to quickly and inexpensively produce videos, saving time and resources.
Alex Mashrabov, an AI video expert and founder of Higgsfield AI, agrees that AI video generation has limitations but acknowledges its potential. Good dialogue content is hard to generate through AI because it often depends on subtle facial expressions and body language, he says. However, he believes that the technology has a strong potential in areas such as product advertising, where AI-generated videos can be much cheaper than traditional methods.
Temu, a major supply chain company, is already using AI-generated videos for direct product advertisements. Even if an AI model needs a lot of prompts to generate a usable ad, shooting with real people, cameras, and equipment could be hundreds of times more expensive, Mashrabov says. With technology improving, this kind of application might become one of the first large-scale applications of AI video generation.
Challenges and Limitations
Despite its potential, AI video generation still faces several challenges. Kahn notes that each shot requires a new prompt, making it difficult to maintain consistency in color, sunlight angle, and architectural design. Additionally, AI models struggle with creating realistic human-like details, such as close-ups of faces.
Mashrabov acknowledges that the success rate of AI video generation is often low, with a ratio of one successful video to every 20-100 attempts. However, he believes that the technology has a strong potential in areas such as product advertising, where even moderately successful videos can be valuable.
Future Prospects
As AI technology continues to evolve, AI video generation is expected to become even more sophisticated and efficient. The potential applications are vast, ranging from advertising to entertainment and even virtual reality. While there are still limitations to the technology, the future of AI video generation looks promising, offering new opportunities for creativity and efficiency in the advertising industry.
Views: 0