Tech has convinced itself that everyone is about to start building their own software.
The argument goes: AI makes building so cheap and fast that everyone will have custom tools tailored exactly to their needs and that programmers will be obsolete. The three-person accounting firm. The 40-truck logistics company. The store down the road. They can now build software cheaply themselves.
I doubt if they will.
Not because the technology isn't real. It is. Not because custom tools aren't sometimes worth having. They are. But the people most excited about this revolution are the people who would be doing the building.
Most people don't love software. They tolerate it. They use it because it solves a problem they have, not because they want a closer relationship with it.
The accounting firm doesn't want a new system – they want the paperwork gone. The logistics company doesn't want a software project – they want the routes optimised. The goals are not the same, and conflating them is a mistake that people who love building software make constantly.
A powerful excavator doesn't turn a homeowner into a contractor. The homeowner wants the hole. The contractor wants to dig holes. The excavator makes digging easier, but it doesn't change what either one actually wants – and the homeowner never wanted to operate an excavator.
AI is the excavator.
The people who were already building or interested in building will build more, and faster, and with less friction – and some of what they build will be genuinely good. The people who wanted the hole dug will still want the hole dug. They'll use AI tools. They'll automate things. And when it breaks, they'll still call someone.
This matters if you're thinking about where software value goes. The businesses that win aren't the ones talking clients into bespoke builds. They're the ones who already made something useful, packaged it so it runs without them, and made it easy for someone else to own. Utility that transfers is worth more than utility that needs a developer babysitting it.
If building got cheap, buying got easier too. Most people will do what they've always done – pay for the SaaS, grab the off-the-shelf tool, subscribe and forget about it. A few will go further and acquire the thing outright, because owning it ends the effort entirely. Either way the instinct is the same: make the problem someone else's. The gap between "I have a problem" and "someone already solved this and I can buy it" keeps shrinking. That's the opportunity – not the building, the handover.
Most people don't want to build software. They want software that works. Increasingly, they can just go and buy it.
← Back to Notes