I’m going to say something that might get me in trouble. Most of the generative video content you see online is garbage. It’s a flashy, incoherent mess of pixels that has more in common with a 90s screensaver than actual storytelling. And I should know. I’ve spent the last two years, and a not-insignificant amount of money, going deep down the rabbit hole of this technology. I’m not just an observer; as an investor in companies like Anthropic, OpenAI, and Scale AI, I have a front-row seat to the revolution. But investing and creating are two different beasts. To truly understand generative video, I had to get my hands dirty.
It started with a simple question: could I, a non-filmmaker, create a compelling short film using only AI tools? Two years later, I have my answer. And it’s a lot more complicated than a simple yes or no. Along the way, I’ve made every mistake in the book. I’ve wasted countless hours on dead-end workflows. I’ve celebrated breakthroughs that turned out to be fool's gold. But I’ve also had moments of pure magic, where the machine and I created something that felt genuinely new, something that wouldn’t have been possible any other way. These are the 7 most critical, non-obvious lessons I learned. This is the stuff I wish someone had told me when I started. It would have saved me a lot of time and frustration.
1. Your Best Ideas Will Come from the AI’s “Mistakes”
When you first start with tools like DALL-E or other image generators, your instinct is to get the perfect image. You write a long, detailed prompt, and you get frustrated when the AI spits back something weird. A character with three arms, a sky that’s green, a car that’s melting into the pavement. My first instinct was to curse the machine and start over. But then, I had a breakthrough. I was trying to generate a shot of a futuristic city, and the AI gave me a building that was twisted like a piece of licorice. It was a “mistake.” But it was also beautiful. And it sparked an idea. What if the city in my story was alive? What if the buildings could change shape based on the emotions of the inhabitants? That one “mistake” became the cornerstone of my entire film.
Now, I actively look for the happy accidents. I’ll intentionally write vague prompts to see what the AI comes up with. I’ll generate hundreds of images and look for the outliers, the weird ones, the ones that make me scratch my head. That’s where the gold is. The AI isn’t just a tool for executing your vision. It’s a creative partner. And sometimes, its best ideas are disguised as mistakes. Don’t be so quick to hit the delete button. Embrace the weird. It’s the source of true originality in this new medium.
2. Prompting Is Not Just About Words, It’s About Rhythm
Everyone talks about prompt engineering. They’ll tell you to be specific, to use descriptive adjectives, to reference famous artists. That’s all true, but it’s only half the story. The real secret to getting great results, especially with video, is understanding the rhythm of your prompts. A video is a sequence of images. And the prompts for those images need to have a certain flow, a certain cadence. You can’t just write a series of disconnected prompts and expect to get a coherent video.
I learned this the hard way. My first attempts at creating a scene were a disaster. The character’s clothes would change from one shot to the next. The lighting would be completely different. The background would morph into something else entirely. It was a continuity nightmare. The problem was, I was thinking in terms of individual images, not in terms of a sequence. I had to learn to write prompts that built on each other, that created a sense of momentum. I started thinking like a songwriter, with verses and choruses. My prompts became a kind of poetry, with repeated phrases and subtle variations. It’s a completely different way of thinking, and it’s what separates the amateurs from the pros.
3. Story Is Still King, and AI Can’t Save a Bad One
This is the most important lesson of all. You can have the most beautiful, technically perfect AI-generated video in the world, but if your story is weak, no one will care. I’ve seen so many people get seduced by the technology. They spend all their time trying to create the most mind-blowing visuals, and they forget to tell a good story. It’s a trap. And I fell into it myself.
My first film was a visual feast. It had impossible camera angles, stunning landscapes, and characters that looked like they had stepped out of a dream. But it was also a narrative mess. There was no emotional core, no character development, no real stakes. It was just a series of pretty pictures. I showed it to a friend, a seasoned screenwriter, and he was not impressed. “You’ve got a great sizzle reel,” he said, “but where’s the steak?” He was right. I had to go back to the drawing board. I had to learn the fundamentals of storytelling. I had to create characters that people could connect with. I had to build a world that felt real, even if it was fantastical. The AI can’t do that for you. It can’t create a compelling narrative from scratch. That’s still your job. And it’s the hardest, but most rewarding, part of the process.
4. The “Right” Tool Is the One You Know Best
I get asked all the time, “What’s the best AI video generator?” It’s the wrong question. There is no “best” tool. There are dozens of them, and they all have their strengths and weaknesses. Some are better for realistic images, others are better for abstract animation. Some are easy to use, others have a steep learning curve. The truth is, the best tool is the one you know inside and out. The one where you understand its quirks, its limitations, and its hidden features.
I’ve seen people jump from one tool to another, always chasing the latest and greatest thing. They spend more time learning new interfaces than they do creating. That’s a recipe for failure. My advice? Pick one or two tools and go deep. Master them. Push them to their limits. I’ve made a conscious decision to focus on a few key AI design tools. I know their every nuance. I know how to get the results I want, and I know how to work around their limitations. That’s a much more valuable skill than having a superficial knowledge of a dozen different tools.
5. Monetization Is Not Obvious, So Think Like a Bootstrapper
Everyone is talking about how AI is going to change the world, but very few people are talking about how to actually make money with it. Especially in the creative fields. The old models are breaking down, and the new ones haven’t been built yet. So what’s an entrepreneur to do? My advice is to think like a bootstrapper. Don’t wait for someone to hand you a business model. Create your own.
I’ve seen people have success with a variety of approaches. Some are creating and selling their own short films. Others are using AI to create marketing videos for businesses. Some are even building their own AI-powered tools and services. The common thread is that they are all scrappy, resourceful, and willing to experiment. They are not afraid to fail. They are constantly testing new ideas, and they are not afraid to pivot when something isn’t working. This is not a time for big, top-down strategies. It’s a time for nimble, bottom-up innovation. The people who are going to win in this new world are the ones who are willing to get their hands dirty and figure it out as they go.
6. The Future Is Not More Realistic, It’s More Imaginative
There’s a lot of talk about the “holy grail” of generative video: photorealistic, feature-length films created entirely by AI. I think that’s a red herring. The real potential of this technology is not to replicate reality, but to create entirely new ones. To take us to places we’ve never been, to show us things we’ve never seen. The future of generative video is not more realistic, it’s more imaginative.
I’m not interested in creating a perfect replica of the world we already know. I’m interested in creating worlds that are strange, and beautiful, and thought-provoking. I’m interested in using AI to explore the outer limits of my own imagination. That’s what gets me excited. And I think that’s what will ultimately resonate with audiences. We don’t need another Hollywood blockbuster. We need new myths, new dreams, new ways of seeing the world. And AI is the tool that’s going to help us create them.
7. This Will Change You as a Creator, So Let It
I’m not the same creator I was two years ago. Not even close. This journey has fundamentally changed the way I think about art, and technology, and the creative process itself. I’m more open to experimentation, more comfortable with uncertainty, and more willing to embrace the unexpected. I’ve learned to let go of my preconceived notions and to trust the process. I’ve learned to collaborate with the machine, to treat it not as a tool, but as a partner.
It’s been a humbling, and at times, frustrating experience. But it’s also been incredibly rewarding. I’ve created things I never thought possible. I’ve pushed the boundaries of my own creativity. And I’ve had a hell of a lot of fun doing it. My advice to anyone who is just starting out on this journey is to embrace the chaos. Don’t be afraid to make mistakes. Don’t be afraid to look foolish. And most importantly, don’t be afraid to let this technology change you. It will, whether you want it to or not. So you might as well lean into it.
The Road Ahead
So, what’s next? For me, I’m going to keep pushing the limits of this technology. I’m going to keep experimenting, keep learning, and keep creating. I’m not sure where this road will lead, but I’m excited to find out. One thing I do know is that we are at the very beginning of a new era of storytelling. And I, for one, can’t wait to see what we create together.
Frequently Asked Questions
Which item on this list has the highest impact?
It depends on your stage and context, but in my experience, the items near the top of the list tend to have the broadest applicability. That said, sometimes the less obvious items create the biggest breakthroughs for specific situations.
How do I know which items apply to my situation?
Start by honestly assessing where your biggest bottleneck is right now. The items that address that specific constraint will give you the highest return on your time and energy.
Can I implement all of these at once?
I'd strongly recommend against it. Pick the 2-3 items that resonate most with your current situation and focus there. Trying to do everything simultaneously is a recipe for doing nothing well.