I’ve been in Silicon Valley for a long time. I’ve seen hype cycles come and go. I’ve built and sold two companies, RemoteTeam to Gusto and MovieLaLa to Gfycat. I’ve written over 200 angel checks into companies you’ve probably heard of, like Anthropic, OpenAI, and Scale AI. So when a new technology comes along, I’ve trained myself to ignore the noise and look for the real signal.
And the noise around DALL-E 3 is deafening. Everyone’s talking about it. Your Twitter feed is full of photorealistic images of things that don’t exist. It’s a fun toy, a cool demo. But the public conversation is completely missing the point. The real story, the one that matters for builders and founders, isn't the pretty pictures. It's the API.
I see it in my portfolio companies. I hear it in pitch meetings. The most innovative founders aren't just generating cool art; they're building entire product categories on top of this technology. And while the public is debating whether DALL-E 3 is more creative than Midjourney, developers are quietly building the future. This is the story nobody is telling.
The API is the Actual Product
Let’s be brutally honest. As a standalone web app, DALL-E 3 is a curiosity. It’s a showcase for what OpenAI’s tech can do. But the real product, the one that will generate billions in value, is the API. The ability to programmatically generate images at scale is what separates a novelty from a platform.
Think about it. When I was building MovieLaLa, a social network for movie lovers, we spent an insane amount of time and money on visual assets. We needed posters, promotional images, social media content. We had a small design team working around the clock. If I had access to an API that could generate unique, high-quality images based on a simple text prompt, we could have moved faster, experimented more, and saved a fortune. We could have generated personalized movie posters for our users. The possibilities would have been endless.
This isn't a hypothetical. I have a portfolio company right now in the e-commerce space. They sell custom-printed apparel. Their old workflow involved photographers, models, and graphic designers to create mockups for every new design. It was slow and expensive. Now, they use the DALL-E 3 API. A new design comes in, and within seconds, they have a dozen photorealistic mockups of that design on different shirt styles, colors, and even on different models. Their time-to-market for a new product went from two weeks to three minutes. That’s not an incremental improvement; it’s a fundamental business transformation.
This is what the media misses. They show a picture of a surrealist painting and say, "Wow, AI is creative." I see a company's entire operating model being reinvented and say, "Wow, that's a 100x return."
Beyond Marketing: System-Level Integration
The most common use case people think of for image generation is marketing. And yes, it’s great for that. Generating ad creatives, social media content, and blog post illustrations is a huge time-saver. But the most successful companies I see are pushing it much further. They are integrating DALL-E 3 at a system level.
What does that mean? It means the image generation isn't a separate, manual step. It's a core part of the product's workflow. I’m an investor in a company in the real estate tech space. They help agents create virtual tours. They used to rely on photographers to go to each property. Now, they use a combination of 360-degree cameras and the DALL-E 3 API. They can take a standard photo of a room and use the API to generate variations. "Show me this living room with hardwood floors." "Show it to me with modern furniture." "Show it at sunset." They are giving their users superpowers, and it's all automated.
This is the difference between using a tool and building a system. The former saves you time. The latter creates a moat. How can a competitor that is still manually staging homes compete with a company that can generate a thousand virtual stagings in an hour for a fraction of the cost? They can't.
Another example: a startup in the gaming space I advise. They are building a platform for dynamic, procedurally generated worlds. They use DALL-E 3 to create textures, character assets, and environmental objects on the fly. The game world is literally being created as the player explores it. This isn't just about saving money on artists; it's about creating a new kind of gaming experience that was impossible before. The game is infinite. The content is never-ending. That’s a system-level integration.
The Unfair Advantage for Startups
For the first time in a long time, startups can compete with incumbents on a key dimension: visual content. For decades, high-quality visual branding was the exclusive domain of large corporations with massive budgets. You needed to hire expensive agencies, photographers, and designers. It was a huge barrier to entry.
That barrier is now gone. A two-person startup in a garage can now produce visual content that is just as good, if not better, than a Fortune 500 company. I’ve seen it happen. One of my portfolio companies, a small team of three, is building a direct-to-consumer brand. Their product photography, all generated via the API, looks like it was shot by a world-class photographer. Their website is stunning. Their social media is a masterclass in branding. They look like a 100-person company. But it’s just three engineers and an API key.
This is a massive, almost unfair, advantage. While big companies are stuck in endless meetings debating brand guidelines and legal reviews for a single image, startups are shipping. They are testing a dozen different visual concepts in a single afternoon. They are personalizing their marketing to individual users. They are moving at a speed that incumbents simply cannot match.
I always tell the founders I invest in to look for these kinds of asymmetric opportunities. Look for the places where a small team can have an outsized impact. The DALL-E 3 API is one of the biggest asymmetric opportunities I’ve seen in my career. It’s a force multiplier for lean, fast-moving teams.
What This Means for You, the Developer
So, you're a developer. What do you do with this? First, stop thinking of DALL-E 3 as an image generator. Start thinking of it as a visual data plane. It's a new primitive, a new building block for creating software. Just like we have databases for storing structured data and APIs for communication, we now have a service for generating visual data.
What can you build with this new primitive?
- Personalization Engines: Imagine an e-commerce site where every product image is tailored to the user's preferences. Or a news site where the article illustrations are generated based on the reader's location and interests.
- Content Creation Tools: Build a platform that helps creators make comics, storyboards, or children's books. I’ve seen pitches for all three in the last month.
- Design and Prototyping: Create tools that help designers and developers quickly visualize and iterate on user interfaces. Instead of wireframes, you could generate interactive, photorealistic mockups from a simple description.
- Synthetic Data Generation: This is a huge one. Training other AI models requires massive datasets. What if you could generate that data synthetically? Need a million images of a specific type of car for a computer vision model? You can generate them.
These are just a few ideas off the top of my head. The real breakthroughs will come from developers like you who are close to a specific problem in a specific industry. The opportunity is not to build a better DALL-E 3. The opportunity is to use DALL-E 3 to build something that was previously impossible.
I’ll leave you with this. The internet wasn't built by the people who invented the TCP/IP protocol. It was built by the developers who used that protocol to create things like Google, Amazon, and Facebook. The same will be true for generative AI. The long-term value will be captured not by the model builders, but by the application builders.
So, go build. The tools are here. The opportunity is massive. And while everyone else is busy making pretty pictures, you can be busy building the future. I’m betting on the builders. I always have.
Where Developers Go Wrong
I see a lot of smart developers get tripped up when they first start playing with this technology. They make a few common, unforced errors. The first is treating it like a magic box. They throw a vague prompt at the API and are disappointed with the generic results. "A picture of a car." What do you expect? You get a generic car. The developers who are getting incredible results are masters of prompt engineering. It's a new skill. It requires precision, creativity, and a lot of iteration. They aren't just describing what they want to see; they are specifying the camera lens, the lighting, the time of day, the artistic style, even the mood. They are treating the AI as a collaborator, not a vending machine.
The second mistake is a failure of imagination. They use the API to do something they are already doing, just a little bit faster. They replace a stock photo with a generated one. That's fine, but it's a local optimization. It's like buying a Ferrari and only driving it in school zones. The real wins come from rethinking the entire workflow. Don't just replace the stock photo. Ask yourself: "Now that I can generate infinite images, what new product can I build? What new service can I offer?" Don't automate the old process; invent a new one.
I saw a team that was building a simple presentation tool. They were going to use DALL-E 3 to generate background images for slides. A nice feature, but not a company. I pushed them. What if the presentation itself was generated? What if a user could just type in a topic, and the AI would create a 10-slide deck with text, images, and a coherent narrative? That's a company. They were thinking about features; I pushed them to think about systems.
The Future is Built, Not Generated
Look, the hype will fade. The novelty will wear off. There will be a new, shinier model next year. But the fundamental shift that has happened is permanent. The cost of creating high-quality, custom visual content has dropped to near zero. This is a tectonic shift in the landscape of software and business. It will create new winners and losers. The winners will be the ones who understand that the model is not the product. The product is what you build with the model.
As an investor, I’m actively looking for founders who get this. I’m looking for the teams that are using these tools to build things that were previously unimaginable. The ones who are not just following the trend but are setting the agenda. The ones who are building systems, not just features.
So my advice to you is simple. Don't be distracted by the pretty pictures. Don't get lost in the Twitter debates. Go deeper. Understand the API. Master the craft of prompt engineering. And then, think bigger. Think about the systems you can build, the industries you can disrupt, and the new experiences you can create. The future isn't going to be generated by an AI. It's going to be built by developers like you. Now, what are you waiting for?
Frequently Asked Questions
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.
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.
How has this view evolved over time?
My thinking on most topics has changed significantly over the years. Early in my career, I held many conventional views that experience proved wrong. I try to update my beliefs when the evidence changes.
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.