The one-person company

The idea of building a substantial company alone used to sound like a slogan more than a plan. Increasingly, it’s just a description of how a specific kind of company gets built now — not every company, but more of them than used to be possible.

What actually changed

It’s not that any single tool got dramatically better overnight. It’s that the whole stack of things a solo founder used to have to either learn slowly or hire out — writing code, drafting a pricing page, sketching a UI, structuring a go-to-market plan — now has a capable AI collaborator sitting behind each one. None of them replace judgment. What they remove is the tax of being slow at something you’re not naturally good at, which used to be the thing that forced founders to either hire early or stall.

The parts that don’t change

Taste still matters — knowing which feature to cut, which customer complaint to ignore, which one to treat as a five-alarm fire. Follow-through still matters, arguably more than ever, since the barrier that used to filter out people who wouldn’t finish (needing a team, needing funding) is gone. What’s left standing between an idea and a real business is mostly discipline, not access.

Starting from evidence, not a blank page

This is exactly why building from a dead product tends to work better than building from a blank page, for a solo founder especially. It removes the most expensive unknown — does anyone actually want this — and replaces it with a narrower, more answerable question: what would make this work where the last attempt didn’t? That’s a question one focused person, with the right tools, can genuinely answer.