I had a weird moment this week that stuck with me.

I built something in about an hour that would have taken me a month a year ago. Same outcome. Completely legitimate – built to my standards, but in minutes instead of weeks.

It was REALLY something to behold, but it genuinely made me sick to my stomach. Depressed, even. The thing that always made me feel protected, writing code, just stopped feeling like protection, and I just watched it evaporate right in front of my face.

I’ve been processing this a lot, what it means, and how it changes my world, and I’ve realized that I’ve been treating code as the moat. Maybe even as the meaning.

I’ve concluded that if AI can build software instantly, then the thing I sell has to be something software alone can’t confer.

That’s the pivot I see coming for SaaS founders.

If your product is basically a nicer interface on top of a data structure, AI is about to eat your lunch, but if your product is something that can’t be reduced to “a fancy hat on a database,” you likely have more room to breathe.

This year, that’s the line I’m using to decide what’s worth building. I’m not concerning myself with “better dashboards” or “cleaner UX” for the sake of it – I’m concerning on building valuable things that someone can’t replace with a prompt, while casually sipping on a nightcap.