What if AI agents started buying ebooks from humans? | Dabble.AI #12
The Dabble.AI Project #12 - Exploring Autonomous Business Models
In my last three posts, I explored the intersection between AI and blockchain technology. However, I didn’t explain how that relates to my research for the dabble.ai project. For a single-person business to generate a billion dollars or more in annual revenue, it will take a combination of the right product, market, and technologies. So, to help frame my thinking, I focus on trying to answer two primary questions. They are:
What product or service offerings could a single-person company scale to 1B?
What are the systems and technologies that make it possible?
Generally, the answer to the first question is likely something virtual or digital. However, the market is also a significant consideration. You’d need a market with the dynamics and size to enable and support a 1B+ single-person company. Let’s consider ebooks as the product. How likely would a single-person ebook publishing company be able to scale to 1B+ in annual revenues? Based on today’s ebook market, I’d guess it's not very likely. But what if AI changes the market for ebooks?
When I think about ebooks today, I think about books sold to humans. But what if AI agents are also consumers of ebooks in the future? Sound crazy? That market is already starting to emerge. The “ebooks” are vector databases, systems prompts, and retrieval augmented generation (RAG) systems. They provide AI systems with information to expand and enhance their current base of knowledge—just like ebooks can for humans. Is that an ebook? That’s not the point. The point is that new markets are emerging, and existing markets are changing. So, we can’t just think about products and markets based on how we define them today.
Let’s envision a publishing company that sells “books” to AI agents. I’m not talking about AI writing books for humans but about books written for AI consumption. What could that look like? What technologies could make it happen? Incumbent distribution channels like Amazon.com focus on human consumers, so someone must create a new distribution system. This system would need to be discoverable by AI agents. The agents would also need a way to evaluate the value of the books and, of course, buy them. The distribution systems would need to ensure that the books don’t get “plagiarized” with a mechanism to prevent AI agents from including the “books” in their internal language models.
I’m not exactly sure how it would all come together. It would likely involve an AI agent discovery system or protocol that agents could use to discover other AI agents based on their capabilities. It could also leverage blockchain technology to enable transaction processing, NFTs, and immutable on-chain access logging to prevent “AI plagiarism.” Perhaps there is also a blockchain-based rating system that lets agents provide a feedback loop to grade and help improve the usefulness of individual books.
Imagine this exists, and you have a best-selling book when AI agents vastly outnumber the human population. What if billions of agents buy fractional access to your book’s NFT or pay each time they access it? In that hypothetical scenario, could selling ebooks be a billion-dollar opportunity for a solopreneur?
That’s it for today. But if this post inspires you to start building a company that provides “books” for AI systems, or if you’ve already started, let’s talk—I’m also an angel investor.