For years, generative AI has promised to make content creation easier. But Brennan Erbz, cofounder and CEO of Flik, believes the industry has been solving the wrong problem.Most AI companies, he argues, have focused on building better creative tools. Flik is betting that creators don’t just need tools; they need infrastructure. The startup is building what Erbz describes as a vertically integrated platform where creators can produce, publish and monetize AI-generated content without stitching together multiple products.“The gap between having a model and having a media company is enormous, and nobody is closing it,” Erbz said. “That’s the vision of Flik; it’s like a platform designed to help storytellers create and distribute content.”It’s an ambitious vision that has attracted attention from the likes of Y Combinator. The company is currently raising another round as it expands both its production platform and its upcoming streaming service.A nightmare that became a companyThe idea for Flik didn’t begin with AI. It began with a dream.“I had this nightmare once where I woke up in a world that didn’t have stories in it. No books, no films, no music, nothing,” Erbz said. “Stories are how people connect, how they make sense of things, how they build something that matters to other people.”For Erbz, the problem wasn’t that people lacked ideas.“The thing that always kept more stories from existing wasn’t a lack of ideas. Everyone has ideas,” he said. “It was that turning an idea into something real... required resources most people didn’t have.”When generative AI models reached a point where they could produce polished media, Erbz saw an opportunity.“If you could make it cheap enough and easy enough for anyone to tell their stories, you wouldn’t just get more content,” he said. “You’d get entire businesses built around stories, creators who could actually own what they make, build an audience, and make a living from it.”Building a production factory instead of another AI tool Erbz draws a sharp distinction between Flik and many of today’s AI media startups. “Many platforms described as AI studios focus on a particular part of the production process, such as video, voice or templates,” he said.Instead, Flik aims to automate the entire production pipeline.“We built a factory," Erbz said. “You give it the idea and it produces the finished thing: the script, the characters, the edit, the score, the whole product, with no human approving each individual step.”That philosophy extended to Flik’s engineering effort.“The filesystem, the multi-agent orchestration, the production skills, the video editor, none of that existed off the shelf, so we had to build it all from scratch,” Erbz said. The company spent roughly two years and a team of about ten people building the platform.The economics, he says, fundamentally change who can create professional-quality media. “We produce a finished minute of video for around fifteen dollars in compute. The traditional number is one to five thousand,” Erbz said. “When production stops being the expensive part, the entry barrier that protected incumbents for eighty years goes away.”Growing beyond creatorsToday, Flik’s largest customer base consists of creators and YouTube channels using AI to increase output without expanding headcount. Brands and agencies are beginning to adopt the platform as well.“We had a branding agency produce several finished commercials in a single quarter on the platform, which is the kind of volume that used to take a production company a year,” he said.Usage has accelerated alongside adoption.“People are generating 18,000 hours of content a month on the platform,” Erbz said. “The average session runs about 18 minutes... because it tells you these are people actually building things, not just poking at a demo and leaving.”The next layer: DistributionWhile production is Flik’s current business, Erbz sees it as only the first piece of a much larger platform.“The next thing people will see is Flik streaming, a place to publish, watch, and monetize what the factory makes,” he said.Rather than forcing creators to jump between editing software, publishing platforms and monetization systems, Flik intends to keep everything inside one ecosystem.“The reason a creator would choose Flik streaming over YouTube is that you produce and distribute on the same stack,” Erbz said. “There’s no separate editing software, no separate distribution platform, no cut to a platform that doesn’t help you make anything.”He believes starting with production gives Flik an unusual advantage.“The cold-start problem is real and the way we solve it is that the factory is already producing content and bringing in an audience, so the platform launches with supply already on it.”Safety as a competitive advantageAs AI video becomes increasingly realistic, deepfakes remain one of the industry’s biggest concerns. Erbz says Flik has intentionally built safeguards directly into the product.“We block real faces and specific-person prompts at generation, and we have copyright tools for post-generation to make sure anything you make is cleared,” he said.That approach has also shaped the company’s work with public figures.“When a real person wants to work with us, it’s a consented partnership with real terms,” Erbz said. “The Peyton Meyer clip wasn’t an unauthorized generation. He came to us, we made it together, and his IP is his.”Although Meyer is currently the only publicly announced celebrity collaborator, Erbz suggested that additional partnerships are already in progress.Betting on the full stackErbz believes Flik may benefit from continued improvements in the foundation models developed across the wider AI industry.“The models commodify and we swap for the best available, so continued investment in stronger foundation models may also improve the capabilities available to our platform,” he said. “We can’t lose a model race we’re not running.”Instead, Flik is betting its long-term advantage comes from owning the workflow around creation.“You produce on the factory, you distribute on the platform, you monetize through revenue splits, you own your IP, and the data from all of it compounds so the next thing you make is better informed than the last,” Erbz said.For Erbz, that’s a much bigger opportunity than building another AI model.“The moat is that we solve the full problem, not one piece of it.”VentureBeat newsroom and editorial staff were not involved in the creation of this content.