The narrative says AI will kill SaaS. Everyone will vibe-code their own tools. No need for Notion, Slack, or Figma. Just prompt your way to custom software.
This ignores 40 years of evidence.
Developers have spent decades making software easier to build.
Languages got simpler. Abstractions got more powerful. You could do more with less code.
Then came the low-code movement. Webflow. Bubble. Airtable. Tools designed for non-developers to ship functional software.
The no-code wave followed. Zapier. Make. Notion databases. Even less friction.
And yet.
People still buy software.
Every time we make building easier, we predict the same outcome: everyone becomes a builder.
It never happens.
Low-code didn't kill SaaS. No-code didn't kill SaaS. AI won't either.
The reason is simple. Most people don't want to build their own software.
If you live on X or in tech circles, you're surrounded by builders. People who tinker. People who automate. People who script their way out of problems.
This creates a warped sample. You assume everyone thinks like this.
They don't.
Most people want to open an app that works. They want someone else to handle edge cases, updates, support, and design. They want to pay $10/month and forget about it.
Building requires:
Most people fail at least one. Usually several.
AI will expand the builder population. That's real.
The person who could never code can now ship basic tools. The developer who took two weeks can ship in two days.
More builders doesn't mean everyone builds. It means the same 90% now have better tools they still won't use.
If 2% of people build today, maybe 5% build tomorrow. Maybe 10% in a decade.
That still leaves 90% who just want software that works.
Consumers outnumber builders. They always have. They always will.
You can't build a mass-market company serving only builders. The math doesn't work.
SaaS companies serve the 90%. The people who don't want to think about infrastructure, schemas, or deployment. Who don't want to debug their own tools.
AI makes the builder toolkit more powerful. It doesn't change human nature.
The idea that everyone will build personal software is a pipe dream.
It assumes everyone wants to be a builder. They don't.
It assumes building is the hard part. Maintenance, edge cases, and ongoing updates are harder.
It assumes people optimize for customization over convenience. They optimize for convenience.
AI will create more builders. Good. The world needs more people who can shape software to their needs.
But it won't eliminate the fundamental split: some people build, most people buy.
SaaS isn't going anywhere.