Posts

Ask someone what it costs to build a product solo and they’ll name the visible line items: a coding-tool subscription, an API bill, a domain, maybe a designer. Those are the cheap parts, and they total to less than you think. The expensive parts never show up on an invoice. They’re time, attention, and the quiet load of being the only decision-maker. This is the honest accounting no one gives you before you start — and it matters, because the thing that actually sinks solo projects isn’t running out of money. It’s misreading what the real cost is. ...

The promise of AI building tools keeps inflating: “describe anything, it builds anything.” Every few months someone ships a demo that looks like the software writes itself. Here’s the honest counterweight, from actually shipping with these tools: agents are superhuman at volume and untrustworthy at judgment. They can generate more code, faster, than any team you’ve seen. They cannot decide what’s worth building, what’s actually correct, or what’s good. That last stretch — call it the last 10% — is where the product is won or lost, and it’s still entirely yours. ...

Every solo builder knows the rhythm. Week one is electric: the idea is alive, the first build session flies by, you can already see the finished thing. Week two is momentum. Then week three arrives, and with it a specific, quiet death. The project doesn’t fail at the hard part. It fails in week three — when the novelty has worn off, the ugly middle has begun, and the end is suddenly very far away. This is the pattern, and it’s more predictable than anyone wants to admit. Here’s why it happens, and the system that gets you through it. ...

Every beginner guide in the AI space starts the same way: an overwhelming list of tools, a wall of “learn Python first,” and a tutorial that assumes you already know what an API is. Most people read two paragraphs, close the tab, and go back to watching other people build things. This is the opposite of that. I build AI products for a living — from the inside of a big, slow, regulated company — and I’ve watched dozens of smart people stall on the same step. The step isn’t technical. It’s that nobody tells you what to build first, and why that first thing matters more than any tool you’ll ever pick. ...

If you build things — software, hardware, or projects for your own reasons — you already know the hardest part isn’t the first week. It’s year three, when the novelty is gone and the thing still isn’t done. Builder Decade is about that version of building. What “the AI era” actually changed For most of the history of making things, the bottleneck was capability. You couldn’t build what you imagined without a team, a budget, or a decade of practicing the craft first. ...

For two decades, having an idea and wanting to ship it meant a familiar list of reasons to wait: find a co-founder, hire an engineer, raise money, build a network. “Solo founder” was shorthand for couldn’t find a co-founder. One person with an idea was treated as a freelancer with delusions of grandeur, not a real company. That wall is gone. In 2026, the honest reason not to ship has been quietly dismantled — not by a hype cycle, but by a measurable collapse in the cost of building software. The tooling changed underneath us, and the data says so. This post is the case for why, if you have an idea now, nothing should stop you — with the receipts, and the honest limits. ...