My first results from Sora were garbage. Absolute, unmitigated trash.
I’m not saying that for dramatic effect. I’m Sahin Boydas, and I’ve been in the tech game for a while. I’ve co-founded companies that got acquired, I’ve written over 200 angel checks into companies like Anthropic, OpenAI, Scale AI, and Hugging Face. I’ve seen the AI sausage get made from every angle. I’m not easily impressed, but I’m also not easily defeated.
But Sora? Sora had me beat.
I had seen the demos. The stunning, photorealistic videos that seemed to defy reality. A woman walking through a neon-lit Tokyo street, a wooly mammoth trudging through a snowy field. It felt like the future had arrived overnight. As an investor in OpenAI, I was ecstatic. As a creator, I was itching to get my hands on it.
Then I got access. And the magic died.
My prompts, which I thought were brilliantly crafted, produced nightmare fuel. I asked for a “Silicon Valley founder pitching his startup to a VC.” I got a man with three arms, a melting face, and a laptop that seemed to be breathing. I tried for a “serene drone shot of the Golden Gate Bridge at sunrise.” The bridge was made of what looked like spaghetti, and the sun was a pulsating, malevolent orb.
It was a disaster. For two weeks, I spent every spare moment trying to crack it. I tweaked my prompts endlessly. I tried being more descriptive. I tried being less descriptive. I used negative prompts. I studied the prompt structure of Midjourney and Stable Diffusion, hoping to find some transferable skills. Nothing worked.
I was getting frustrated. Here was this supposedly revolutionary tool, and I couldn’t even generate a coherent 10-second clip. I felt like a fraud. I’m supposed to be a guy who gets this stuff. I started to think the public demos were cherry-picked flukes. I was ready to write off Sora as another overhyped AI toy and move on.
I was one click away from closing the tab and forgetting about it forever. Then, on a Tuesday night, scrolling through a Discord channel dedicated to AI video, I saw a message that changed everything.
It wasn’t some grand pronouncement. It was a simple, almost offhand comment from a user with an anime avatar. They were talking about camera movement.
The Breakthrough: It’s Not a Prompter, It’s a Director’s Chair
The comment was this: “Stop telling Sora what to show you. Start telling it how to show you.”
It hit me like a ton of bricks. I had been treating Sora like a text-to-image generator. I was focused entirely on the subject of the video. The founder, the bridge, the whatever. I was describing the scene, the lighting, the mood. But I was completely neglecting the camera.
I had been writing prompts like this:
“A photorealistic video of a programmer coding in a dark room, green text on the screen, a can of energy drink on the desk.”
And I was getting those weird, distorted results. The programmer’s fingers would merge with the keyboard, the can would float in mid-air.
After reading that comment, I tried a new approach. I started thinking like a film director. I added one simple thing to my prompt: camera instructions.
My new prompt looked like this:
“Slow dolly shot moving forward through a dark room where a programmer is coding. Green text scrolls on his monitor. The camera pushes in slowly on his face, showing his intense focus. A can of energy drink sits on the desk, slightly out of focus in the foreground.”
I hit “generate.” And I waited.
What came back was… perfect. It was the video I had been picturing in my head all along. The movement was smooth. The focus was right. The programmer looked like a real person. The can of energy drink stayed put.
It was my “aha!” moment. The single trick that unlocked everything.
From Frustration to Professional-Grade Results
That one shift in mindset changed my entire relationship with Sora. I stopped being a prompter and started being a director. I began to think in terms of shots, not just scenes.
Instead of: “A dog on a skateboard.”
I wrote: “Low-angle tracking shot following a golden retriever on a skateboard as he cruises down a sunny, tree-lined street.”
Instead of: “A chef cooking in a kitchen.”
I wrote: “Over-the-shoulder shot of a chef’s hands as he expertly chops vegetables. Quick cuts to a sizzling pan and a pot of boiling water.”
Suddenly, I was getting the results I wanted. And not just getting them, but exceeding them. The level of control was incredible. I could specify the lens type, the camera angle, the movement, the focus. I could create dynamic, engaging videos that looked like they were shot by a professional film crew.
I spent the next week recreating some of my failed prompts. That spaghetti Golden Gate Bridge? It was now a majestic, sweeping aerial shot, the morning fog rolling in beneath it. The three-armed founder? He was now a charismatic entrepreneur, captured in a series of tight, expressive close-ups during his pitch.
The Future is Directed, Not Just Generated
My journey with Sora was a humbling one. It reminded me that even with the most powerful tools, there’s a learning curve. It’s not about just typing in a few words and expecting magic. It’s about understanding the logic of the tool and learning to speak its language.
For Sora, that language is the language of cinema. It’s not just an image generator that happens to create videos. It’s a filmmaking tool. And to use it effectively, you have to think like a filmmaker.
I’m convinced that the future of AI video isn’t just about generating more realistic or more fantastical scenes. It’s about giving creators more granular control over the storytelling. It’s about empowering us to direct the AI, to guide it, to collaborate with it to bring our visions to life.
Sora is not a replacement for human creativity. It’s a powerful new instrument. And like any instrument, you have to learn how to play it. It takes practice, it takes patience, and it takes a willingness to look beyond the obvious. But once you find that breakthrough, that one little trick that makes it all click? The possibilities are endless.
I almost quit. I’m so glad I didn’t.
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.
What's the most common pushback you get on this?
People often push back by citing exceptions or edge cases. And they're usually right that exceptions exist. But building a strategy around exceptions rather than patterns is a losing game for most founders.