How new AI video tech could help brands produce long-form content

Brands have long believed they have the panache for creating long-form, episodic content, from Coca-Cola’s 2022 holiday series to Chick-fil-A’s annual animated film. But new developments in AI could make this style of video marketing a more common, and less costly, experiment. Showrunner is a content generation platform that uses AI to create full-length TV shows—from visuals to dialogue to directorial style. The tool’s official launch to testers last month garnered significant attention in the entertainment industry, in part for the impressive clips that the founding studio, The Simulation, teased on social media. With the power to quickly generate this content through mere text prompts, Showrunner is receiving interest from Hollywood studios, The Simulation wrote on X. Soon, brands at large could begin to seek out the platform, as well as other platforms in the burgeoning AI-powered video space, to develop longer-form content that has previously been difficult to justify. This format typically requires substantial expenses and impediments toward scale, said Craig Elimeliah, chief creative officer at Code and Theory. The higher ROI of short-form content has also fueled hesitancy with long-form. But with Showrunner’s AI technology cutting out much of the costs and delivering a film far more quickly, “now we’ve got a conversation,” Elimeliah said. Check out the latest Ad Age events and award programs Showrunner’s videos are built from a text prompt, with the option for users to edit outputs across numerous criteria after they’re created. While the platform is currently confined to producing animated films, it has already proved an impressive ability to do so in various styles, including those eerily close to branded IP. “Exit Valley,” for example, is a satirical comedy on Silicon Valley that resembles “South Park.” “Pixels” is another forthcoming show that resembles Disney’s “Cars” franchise. The Simulation did not respond to a request for comment. Also read: OpenAI’s Sora tool reveals divide over AI and creativity For brands, Showrunner could help to build community through episodic material, said Elimeliah. Brands including Dr Pepper have previously done this through non-AI means, such as with “Fansville,” the soda brand’s multi-year campaign around college football. Developments in AI, said Elimeliah, turn these efforts into more malleable assets via the quick and easy process of editing. The technology can also incorporate proprietary data, allowing for content that is more emotionally resonant with a brand’s specific audiences. Entertainment brands appear to be Showrunner’s primary enterprise client. “What we’re seeing is that Hollywood is excited to explore a world where some episodes of popular shows are made by fans, so long as the revenue and credits make sense and are appropriate,” The Simulation posted on X. Media giants including Fox have previously attempted fan-created content through NFT technology, but have yet to see substantial success. As for a more traditional brand such as Dr Pepper, Showrunner and other AI platforms do not yet have a place in its “Fansville” plans, chief marketing officer Andrew Springate told Ad Age. “But that’s something we’ll certainly experiment with and get an understanding for the right opportunity,” he added. AI as a video genre Other developments in AI video creation could change how brands leverage long-form content. Sora, OpenAI’s text-to-video platform, is widely considered to be at the cutting edge of the space. Such a platform could improve how brands create long-form content because it can instantly provide a visual representation of any concept, said Victor Potrel, senior VP of partnerships and creator services at digital studio TheSoul Publishing. At least one brand has already released a Sora-created production: Toys R Us, which last week debuted a one-minute campaign video that tells the story of how its founder developed the brand. That campaign was met with backlash from the creative community, with some criticizing the video quality. Previously, brands developing long-form content have needed to brainstorm a multitude of concepts before selecting just one to follow—a result of budget and time constraints. But with Sora, “testing an idea is much easier and faster with limited means,” Potrel said. One criticism of generative AI is that the technology can allow for more content, but not necessarily better content—sometimes even noticeably terrible content. And given the quality of long-form material that consumers are used to in their daily consumption, this format already has a particularly steep curve. But Potrel sees the distinctive qualities of AI-generated content as a benefit, not a drawback. “AI creation is like a genre in itself,” he said, in which consumers knowingly take part in AI experiments to engage with material that feels wholly different from that of other categories. Marketers such as Avocados From Mexico and Burger King have already embraced this genre when it comes to image and text generation. AFM asked fans to input guacamole ingredients and outputted an image of the concoction, while the burger chain did the same with a burger recipe. Also read: Snickers’ AI clone of a soccer legend will give advice to fans This same interactivity can be applied to AI video content, potentially enabling fans to create their own TV episodes with branded IP, said Potrel. Brands that are hesitant to put their IP in consumers’ hands can still take advantage of this new category by creating long-form video that is transparent about the role of AI, so as to establish expectations about the quality of the content, Potrel added. Consumers have already shown an interest in long-form video content that is created by AI, even when this material is of far inferior quality. Last year, thousands of people watched an AI-generated version of “Seinfeld” that perpetually streamed on Twitch, seemingly dazzled by the ways that the technology was continually stringing together content. And streaming services are already testing out the form. TCLtv+, the online streamer for consumer electronics company TCL, is releasing a rom-com almost entirely produced by text-to-video platform Runway and text-to-image platform Midjourney. Dubbed “Next Stop Paris,” the film will be roughly the length of a TV episode. Copyright infringement on the rise As AI automates more duties in the video creation process, production companies could be impacted. Animators, for example, are no longer needed by companies using Showrunner, Potrel said, unless, of course, they want a style that the platform is incapable of producing. The notion of eventual replacement is already sowing fear within the production industry, as are issues over copyright. In March, an Under Armour ad created with AI by production company Tool ignited controversy because it used footage previously shot by an uninvolved director. Copyright infringement has already reared its head in longer AI-generated video content. As part of an effort to demonstrate Showrunner to the public last year, The Simulation released a 22-minute episode of “South Park” generated by the platform. The studio created the episode for research purposes only, and therefore did not fall into any trouble with the franchise’s owner in Paramount. But the experiment, launched at the height of the WGA strike, showed how easily brands’ IP can be re-created by AI platforms. Announcing our paper on Generative TV & Showrunner Agents!Create episodes of TV shows with a prompt - SHOW-1 will write, animate, direct, voice, edit for you.We used South Park FOR RESEARCH ONLY - we won't be releasing ability to make your own South Park episodes -not our IP! pic.twitter.com/6P2WQd8SvY— The Simulation (@fablesimulation) July 18, 2023 Another text-to-video platform, Luma’s Dream Machine, stirred problems this month when a promotional video for one of its users’ studios was found to include Mike Wazowski, a character from Disney’s “Monster Inc.” movie, without authorized permission. Upon closer inspection, Mike from the actual Monster's Inc. Is in the cartoon. Straight up ripping it off. https://t.co/gOPbppcXHj pic.twitter.com/DOW9l4Nsxv— Jacob Davison (@JacobDavison_) June 16, 2024 Despite consistent copyright violations in the AI video space, Code and Theory’s Elimeliah is not worried about the safety of these tools. He sees assurances of safety as being part of the maturation of value propositions for AI developers, such as through indemnification policies, which companies like Amazon have already introduced. The policy, along with a watermarking practice, protects clients from claims of third-party IP infringement. As for longer-form videos, Elimeliah is also focused on the technology’s ability to improve from where it is now. The quality that is currently demonstrated by platforms like Showunner seems to be besides the point. “The shows might be shit, but the idea is not,” he said.

How new AI video tech could help brands produce long-form content

Featured Stories