I’m going to say something that might get me in trouble. The entire AI-driven storytelling industry is a house of cards. It’s a bold statement, I know. Especially from someone who has built a career on this technology. But after eight years in the trenches, two successful exits, and writing checks to over 200 startups like Anthropic and OpenAI, I’ve seen where the bodies are buried. I went from burning cash like a crypto bro in a bull market to building an 8-figure business by breaking every rule in the book. This isn’t another sanitized success story. This is the raw, unfiltered truth of what actually works.
My story in this space began almost a decade ago, long before “AI” was the answer to every question in a pitch deck. Back then, using algorithms for creative work was a niche, almost academic, pursuit. My first real swing at it, MovieLaLa, was a classic Silicon Valley tale of ambition meeting a brick wall of reality. We had a brilliant team, some of the sharpest engineers I’ve ever had the pleasure of working with, and a vision to change how people discover movies. We also burned through millions in VC money chasing that dream. We built complex models and wrote beautiful code. The problem? We were in love with the tech, not the story. We were engineers building for other engineers, and we completely forgot about the only person who mattered: the audience. The project was a technical masterpiece and a commercial disaster. A painful, expensive, but vital lesson.
The House of Cards: What’s Wrong with AI Storytelling
That failure was a wake-up call. It forced me to see the deep-rooted problems in the AI creator space. So many companies are repeating my mistakes, hypnotized by the promise of AI while ignoring the basics of human connection. They throw around terms like “generative models” and “neural networks” but forget that storytelling is an ancient art. The tech is just a tool. A very powerful tool, yes, but still just a tool, like a pen or a camera. It is not the story.
The core of the issue is a complete misunderstanding of the goal. Most teams are trying to build an AI that can write the perfect script or compose a flawless symphony. They’re chasing a ghost, trying to replicate human creativity with code. It’s a fool’s errand. You can’t automate soul. What you get is a tidal wave of generic, lifeless content that just feels… off. It’s technically perfect and emotionally empty. It’s the uncanny valley of creativity, and people can sense it. They might not have the words for it, but they know it’s not real.
I remember a pitch I sat through a few years back. A startup wanted to create an AI that wrote personalized novels. They’d raised a seed round and were beaming with pride over their text-generation model. They showed me a demo, a short detective story set in some generic cyberpunk city. The writing was clean, the grammar impeccable. But it had no heart. It was a paint-by-numbers mystery, hitting every predictable plot point without a single surprising or genuinely emotional moment. I asked them one simple question: “Would you read this for fun?” The founders just looked at each other. In that silence, I saw the fatal flaw. They were so obsessed with the technical puzzle that they forgot to make something a human being would actually want to read. They had built a solution for a problem that didn’t exist.
Breaking the Rules: From Tech-First to Story-First
After the MovieLaLa implosion, I took a step back. I still believed AI could change everything, but the current path was a dead end. I had to unlearn everything I thought I knew about building a tech company. I had to stop thinking like an engineer and start thinking like a storyteller. This led me to a set of core principles that became the bedrock of my next company, the one that would eventually hit the 8-figure mark.
My first and most important rule is this: AI should be a collaborator, not a creator. This is the mental shift that changes the game. Instead of trying to build an AI that replaces the artist, we should be building tools that supercharge their creativity. Think of it as a power-up for the human imagination. The AI can generate a flood of ideas, explore a million possibilities, and handle the grunt work, but the final vision, the emotional core, has to come from a person.
We were religious about this principle. In our project, which was all about creating animated content, we never asked the AI to write a script. We’d ask it to generate a thousand different loglines from a simple premise. A human writer would then dig through that pile of ideas, find a diamond in the rough, and build a story around it. We didn’t ask the AI to design a character. We had it generate hundreds of visual styles and concepts that our artists used as a jumping-off point. The AI was our creative engine, a tireless intern that never slept, but the human artists were always the ones with their hands on the wheel. This was the complete opposite of the industry’s obsession with full automation. It was slower, messier, and forced our creative and technical teams into a deep, sometimes chaotic, collaboration. But it worked.
Another rule we lived by was to embrace the chaos. The hunt for technical perfection leads to sterile content. Real art is messy. It has flaws and quirks and moments that make you go, “What the hell was that?” We discovered that some of our best ideas came from the AI’s “mistakes.” A weird visual glitch or a nonsensical line of dialogue wasn’t a bug to be fixed; it was a happy accident to be explored. One of our most popular animated shorts came from an AI-generated image that was a total misinterpretation of our prompt. It was a bizarre, dreamlike image that made no logical sense, but it was incredibly evocative. Our creative team saw it and built an entire story around that single, strange image. That short went viral, racking up over 10 million views in a week, not because it was perfect, but because it was weird and surprising.
Inside the 8-Figure Project: How We Did It
So, what was this mysterious 8-figure project? It was a platform for creating and distributing short-form, AI-assisted animated content. A new kind of animation studio built for the speed of the internet. We started with a small team of 20 people, a mix of engineers, artists, and writers, and a $2 million seed investment. Within three years, we were pulling in over $10 million in annual revenue with a 40% profit margin. And we did it with zero marketing spend. Our content was our marketing.
The secret wasn’t some magical algorithm. It was our workflow. It was the culture of collaboration we fostered between our human talent and our AI systems. Our process was a four-step dance:
- Human-led Ideation: Our creative team would kick around a theme or a high-level concept. Something simple, like “a story about a lonely robot who finds a friend.”
- AI-powered Exploration: We’d feed that concept to our internal tools. The AI would then spit out hundreds of story beats, character designs, and environmental concepts. This wasn’t a single button press. It was an interactive jam session. Our artists would guide the AI, tweak the parameters, and mash different outputs together.
- Human Curation and Creation: The creative team would then sift through the AI’s output, picking the best bits to use as building blocks. A writer would weave the story beats into a script. An artist would refine a character design. An animator would bring it to life. The AI provided the clay, but the human touch did the sculpting.
- Rapid Iteration and Distribution: Because our workflow was so fast, we could produce a high-quality animated short in days, not months. We released a constant stream of content, saw what the audience responded to, and immediately doubled down on what worked. We had a direct, real-time feedback loop with our viewers, something a traditional studio would kill for.
One of our biggest hits was a series about a grumpy gnome living in a smart home. The initial idea was basic, but the AI helped us explore it in hundreds of hilarious ways. It suggested absurd scenarios, like the gnome getting into a philosophical argument with a condescending AI assistant or trying to use a smart toaster that demanded a firmware update. The series was funny, relatable, and deeply human, even though a machine helped birth it. That series alone brought in over $2 million in revenue from merchandise and licensing deals.
The Uncomfortable Truth
After all these years, I’ve learned that the secret to success in AI-driven storytelling has almost nothing to do with AI. It has everything to do with understanding people. It’s about knowing what makes them laugh, what makes them cry, and what makes them hit the share button. The technology is an incredibly powerful tool, but that’s all it is. A tool. The companies that win in the long run will be the ones that put the story first. They will be the ones who use AI not to replace human creativity, but to amplify it.
The house of cards is the belief that technology can solve a human problem. It’s the fantasy that a sufficiently complex algorithm can manufacture emotion and meaning. It can’t. The future of this industry belongs to the artists, the writers, and the creators who are brave enough to grab these new tools and use them to tell their own, unique stories. The rest is just noise. The uncomfortable truth is that you still need to be a great storyteller. The AI can’t save you from a bad idea.
Frequently Asked Questions
What experience informs this perspective?
This perspective comes from over a decade of building companies in Silicon Valley, two successful exits (RemoteTeam to Gusto, MovieLaLa to Gfycat), and investing in 200+ startups including Anthropic, OpenAI, and Scale AI. I write about what I've lived.
Do all experts agree with this view?
No, and that's fine. The best ideas in business are often contrarian. I share my perspective based on my experience and data, but I encourage you to seek out opposing viewpoints and form your own conclusions.
How can I apply this thinking to my own situation?
Start by identifying the core principle behind the opinion, not the specific example. Then ask yourself: does this principle apply to my context? If yes, test it in a small, low-risk way before going all in.