OpenAI GPT Builders: creating millionaires from technology that isn't theirs

How you earn millions with someone else's technology.

OpenAI launched GPT Builders: a platform where you can create your own AI bots and sell them. Such a bot isn't really your creation. You take GPT-4, give it context, tell it "this is now your personality," and you sell it. The technology is OpenAI's. The parameters are OpenAI's. You're the intermediary.

Yet this will generate billions for some companies.

This is the most elegant example of how tech power works. OpenAI built a machine. That machine costs billions in research, energy, data. But they open it, give it away with some guardrails, and say: make money from it.

What you're really doing is adding an interface. Your wrapper. You say "this is my GPT for X" and users pay you monthly. You get a cut of revenue. OpenAI gets the rest.

In economic terms this is called "rent." You earn money from something you didn't build, didn't maintain, don't risk. You sell access to someone else's system.

This isn't new. AWS sellers do the same. App store developers. Shopify merchants. But it feels different with AI. Because it feels like creation. It feels like you're building something. But you're not. You're promoting something.

The millionaires that emerge from this will say: we found it. We saw a gap. We filled it. Entrepreneurship! But what they really did is: they made a thesaurus for OpenAI's machine.

This says much about where we are. We're in a phase of tech where the tools have become so powerful that simple tweaks are worth billions. That's not progress. That's concentration of power.

In previous waves this was different. The internet emerged, and thousands of companies emerged. Website builders. Email. Search. Each built on something new, not standing on the shoulders of a single giant. Now they all build on GPT. They build on Anthropic's Claude. They build on open models.

But they build on something that isn't theirs.

This touches iRecord exactly. We built identity verification. Not on Google's Identity Platform. Not on Samsung's biometrics. Ourselves. Own technology. Own architecture. It costs more. It risks more. But it means we don't reside in someone else's house.

The question for GPT Builders is: you earn millions. But who are you really working for?

Sources: OpenAI GPT Store documentation (2024); Y Combinator research on generative AI startups; McKinsey 'The Economics of Artificial Intelligence' (2024)