One-person company has been the hype for a while now: use agents to fill in your execution, and your understanding of remote domains.
I admire the people who manage to build these one-person empires. The learning ability, execution, and judgment it takes are all worthy of respect. I ran a small experiment of my own for a while.
But maybe: what OPC gives you is freedom. And the more capable you are, the higher the price you pay for that freedom.
Agents make execution cheap. As we all know, the hard part is judgment.
I believe my judgment at engineering x finance is on par with the best AI models in the world. But building good taste in other domains is not something you can rush. What's a good UX, what's the right marketing strategy: there, I look up to my human peers.
OPC is a very cool thing to do. But once it reaches a certain scale, or maybe right from the start, to go from good to great, you need a small, sharp team. One person can only hold one or two frontier lines. A great venture needs someone standing on every line, each with their own agent team.
Of course, such a team — how to find it, how to attract it — you have to establish yourself first. And OPC happens to be a viable way to prove yourself.