This is a guest post by Sahin Boydas, a serial entrepreneur, and investor in 200+ companies, including OpenAI, Anthropic, and Scale AI.

Why Your Approach to Synthetic Media is Wasting Your Money

Hot take: Your synthetic media strategy is likely flawed. It's time to address the silent killers of creative output and discover what the top 1% are doing differently.

I see it all the time. A founder gets a demo of the latest AI image or video generator—maybe it's Midjourney, maybe it's Sora—and their eyes light up. They see a slick, futuristic interface and immediately start calculating how much they can save on stock photography and marketing videos. They throw a few thousand dollars at a subscription, tell their marketing team to “go play with it,” and expect magic to happen.

Six months later, what do they have to show for it? A folder full of weird, uncanny-valley images where people have six fingers, a slightly lighter bank account, and zero impact on their actual business. The needle hasn’t moved. It’s a complete and utter waste of time and money. And it’s because they're fundamentally misunderstanding what this technology is for.

I learned this lesson the hard way, and it was an expensive one. Back at MovieLaLa, my second startup which was later acquired by Gfycat, we were building a movie discovery app. This was years before the current AI boom, but we were early believers in its potential. We had this grand vision of using AI to generate personalized, dynamic trailers for every single user. Imagine a trailer for a new blockbuster that only featured the actors you loved, highlighted the genres you preferred, and was cut to the rhythm of your favorite music. It sounded revolutionary.

We raised a seed round on this idea and hired a team of brilliant, but very expensive, computer vision PhDs. We spent over a year and a huge chunk of our funding trying to build this system. The result? A complete flop. The technology was nowhere near ready. The "AI" could barely distinguish a car from a person, let alone edit a compelling trailer. But the bigger lesson wasn't about the tech's immaturity. It was that we were so mesmerized by our own cleverness, by the technical challenge, that we forgot to ask the most important question: is this the best way to solve the user's problem? We were trying to build a product that didn't exist, with technology that didn't work, and we burned a lot of cash to learn that lesson.

The Old Way vs. The New Way

Most founders are still stuck in that old way of thinking. They see synthetic media as a black box. You write a prompt, you pull the lever, and you get an image or a video. It’s a slot machine. You’re gambling your time and money, hoping for a jackpot. That is not a strategy.

The new way—the way the top 1% of founders are thinking about this—is to see synthetic media as a powerful, controllable tool. It’s not magic, it’s a manufacturing process for pixels. And like any manufacturing process, you need to understand the inputs, the machinery, and the workflow to get a high-quality output. You need a deep, intuitive feel for the models, the data, and the prompting techniques. You need a clear strategy for how you’re going to use it to create real business value.

Mistake #1: Chasing Vanity Metrics

So many founders get bogged down in the wrong metrics. They get obsessed with image resolution, generation speed, or the number of different artistic styles they can produce. Who cares? Seriously, who cares if your image is 8K if it doesn’t make a potential customer click “buy”?

The only metric that matters is impact. Is your synthetic media strategy actually helping you acquire more customers? Is it increasing your conversion rate? Is it making your product more engaging? Is it reducing churn? If you can't draw a straight line from your Midjourney subscription to one of your core business KPIs, you're doing it wrong.

When I made my angel investment in Scale AI, I didn’t do it because they had the fastest data labeling platform on the planet. I invested because Alex Wang and his team understood a fundamental truth: the quality of any AI system is a direct function of the quality of the data it’s trained on. Garbage in, garbage out. It’s the same with generative media. The quality of your output is directly proportional to the quality of your input. And I’m not just talking about the prompt. I’m talking about the data, the examples, the style references, and the feedback loops you use to refine your process. A good process isn't about one-shot prompting; it's about iterative refinement, control, and consistency. It's about building a system, a workflow, that reliably produces on-brand, high-impact visuals, not just a series of one-off lucky shots.

Mistake #2: The Black Box Mentality

This is a big one, and it drives me crazy. Founders who are otherwise brilliant and hands-on suddenly become passive when it comes to AI. They treat it like some dark art that only engineers with PhDs can understand. They hire a few people, delegate the entire “AI strategy,” and hope for the best. That’s a huge mistake.

You, the founder, need to be involved. You need to get your hands dirty. I’m not saying you need to learn how to code a transformer model from scratch. But you absolutely need to know enough to be dangerous. You should be spending hours every week using the tools yourself. You should understand what a LoRA is. You should know the difference between diffusion models and GANs. You should be able to have an intelligent conversation with your team about model architecture, training data, and fine-tuning.

My investment in Hugging Face was a bet on the democratization of machine learning. Their platform makes it easier for everyone to access powerful models. But easier access doesn’t mean you can afford to be ignorant. In fact, it means the opposite. When everyone has the same tools, the only competitive advantage is your unique insight and your ability to apply those tools in a novel way. You can’t have that insight if you’re afraid to look inside the black box. The best founders I know can talk shop with their engineers and then, in the next meeting, explain the business implications to their board. You need that level of fluency.

Mistake #3: Having No Real Strategy

This is the most common and most fatal mistake of all. Founders get so mesmerized by the what (we can make cool pictures!) that they completely forget to define the why. Why are you doing this? What specific business problem are you trying to solve? What does success actually look like, and how will you measure it?

Without a clear strategy, you’re just throwing spaghetti at the wall. You’re generating a bunch of random assets, hoping one of them goes viral or magically improves your brand image. That’s not a strategy. That’s a recipe for burning through your seed round with nothing to show for it.

Your synthetic media strategy must be tightly integrated with your core business strategy. It needs to support your product roadmap, your marketing campaigns, and your sales goals. For example, at RemoteTeam (which was acquired by Gusto), our entire strategy was about building a product that made remote work simpler. Every feature, every blog post, every marketing decision was aligned with that single goal. If we had used generative AI, it wouldn’t have been to create generic images of people working from laptops. It would have been to build a tool that, say, automatically generated visual onboarding guides for new remote hires, directly solving a core customer pain point. That’s strategy. It’s specific, it’s measurable, and it’s directly tied to user value.

The 1% Approach: From Cost Center to Product-Led Growth

So what are the top 1% of companies doing differently? They’re avoiding these mistakes, for starters. They have a deep understanding of the technology. They are obsessed with data quality and workflow. They have a crystal-clear, metric-driven strategy.

But they’re also doing something more profound. They’re not thinking about synthetic media as a way to save money or a tool for marketing. They’re thinking about it as a core component of their product.

They’re not just using AI to create assets for the business; they’re using it to be the business. They are building products that were simply not possible before. They are creating entirely new kinds of user experiences.

Look at what my friends at OpenAI are doing with Sora, or what the team at Anthropic is building. These are foundational model companies, and my investments in them were bets on this very idea: that AI is a new creative medium, a new primitive for building software. The goal isn’t just to build a better video generator. The goal is to build a new tool for thought, a new engine for creativity. A tool that will enable a whole new generation of artists, filmmakers, developers, and storytellers to create things we can’t even imagine yet.

The 1% approach is to ask: "How can we weave this technology into the fabric of our product to create a 10x better experience?" Not "How can we use this to make cheaper blog post images?" Think about a design tool that doesn't just give you templates but generates a unique, on-brand design system from a simple description. Think about a game where the world and its characters are generated in real-time, responding to the player's actions. That's the level you need to be thinking at.

The Uncomfortable Truth

Let’s be honest. Your current approach to synthetic media is probably a waste of money. You’re likely focused on the wrong things, you don’t understand the technology deeply enough, and you don’t have a clear strategy that connects to real business impact.

It’s time to change that. Stop playing with the slot machine. Stop treating this powerful technology like a toy. Start thinking like the top 1%. Go deep. Get your hands dirty. Develop a real, product-led strategy. The future of your business might just depend on it. Don’t be the founder with a folder full of six-fingered aliens. Be the one who builds the next indispensable product. The choice is yours.

Frequently Asked Questions

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.

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.