Should I build SaaS without AI? I honestly don't know

For a while now, I've been thinking about this. Mostly because my flagship SaaS is about as boring as it gets. It's a docs editor. Nothing fancy. The biggest advantage will come once I can ship an iPad version, because my whole product hinges on the split-view feature and writing on iPad.

I don't have ambitions to compete with the big web docs editors. Let's first see if I can survive building a native Mac app, then the iPad app.

Read also: How do you vibe code a Mac app? Fuck I don’t know.

Anyway, that's not really the point here. This week I've been buried in recording for Story Valley, recalibrating my podcast workflow, and recording for Story Brew on top of that.

But the biggest chunk of my time has gone into polishing the MVP. Fixing every UI/UX bug. Closing every hole, because the app was inconsistent in design, flow, all of it. Worse than I thought, honestly.

Read also: Docs editor SaaS refinements: it's more work than I thought

All that, for a simple docs editor. No AI involved anywhere. And today, somewhere in the middle of fixing shit, a question landed on me:

"Should I just keep building a micro-SaaS portfolio under OnPoint VC without any AI integration? No "empowered by AI" tagline, nothing."

Should I bet on non-AI builds first?

I really don't know. On one hand, with my current vibe coding experience, I'm too inexperienced to calculate real AI usage costs or factor in what it would cost me long term. Maybe that changes after a few more builds, once I'm further into this.

Read also: My honeymoon phase with vibe coding is over

But everything I currently build or run is non-AI. I don't mean AI assistance for myself, because without it I couldn't code a single fucking line, and I've been building like a maniac with it.

I mean the products themselves. Study Brew has no AI. TWO Docs has no AI. My podcasts are a hundred percent my voice. Every other build I have currently shelved is also non-AI.

Maybe that's me being afraid to implement it. Or maybe it's my subconscious telling me AI fatigue is real, and that I should bet on a large portfolio of non-AI apps first, then expand into AI territory later.

I don't fucking know yet.

What I do know is that AI is turning into a commodity, which makes it so much harder to build any real moat around it alone. The boring stuff runs a business. Not everything needs to be AI-powered or AI-assisted.

Notion proved my point

Nearly every build I see on Product Hunt or Reddit has gone AI. That's part of why I stepped away from Notion a while back. The more aggressively they pushed AI, the more diluted the actual product felt, buried under constant automation and workflows.

Great for enterprises maybe, but are individuals and solo operators always that keen on using AI everywhere when the bills and subscription costs just keep climbing?

A few weeks ago, Notion proved my point without even trying. They killed Notion Mail, their own AI-branded email client, and admitted flat out that more than half their users had stopped opening the inbox altogether, letting AI agents run the whole thing instead.

I'm not saying we shouldn't use AI. There's a real difference between quietly running it in the background and making it the flagship feature every landing page leads with.

But I can't tell yet. I need to educate myself more.

I did some surface-level digging. Thousands of niches still run on good old goddamn spreadsheets.

And I've watched companies swing from firing people because AI was supposedly going to replace them to quietly rehiring the same talent. After all, AI isn't replacing human labor just yet.

So...the question?

This always leads me to the same question. Should I build more without AI and carve out a niche on the boring business side? Or lean into the crowd on Reddit's /degoogle, the people who won't share their data or their privacy no matter what?

I really don't know.

I've had some thoughts about adding AI to both my SaaS builds eventually, but it would run in the background, very subtle, doing its job. Not the parade of the fucking century, not the flagship feature.

I'm honestly not sure what to do. This question just popped up, and I don't have an answer yet.

All I know is that right now, I'm extremely glad I took the non-AI route with my docs editor. I prefer simple products, and I'm probably not the only one.

It's not keeping me up at night, but I do need to think about my long-term bets, sustainability, and how I build different products going forward.

If I ever add AI to my docs editor, it would be optional, never forced. That feels like the best middle ground for the future. AI affects subscription tiers and pricing, and not everyone can afford to stack multiple expensive subscriptions.

What am I going to do?

Based on all this rambling, it's clear I don't know the answer yet. But when everyone chases the same AI idea, competition gets crowded fast, customer acquisition costs climb, and expectations turn unrealistic before you've even shipped anything.

It can look great on paper and fall apart in execution. The moment your API costs spike, the moment your AI credits get burned through by a wave of trial users who never convert.

Building genuinely excellent, non-AI software has quietly become the overlooked opportunity, I think.

I need to sit with this longer. Maybe my idea bench will spit out something that settles it, AI or not. Or maybe I keep building boring things and let the answer show up on its own.

Who the fuck knows.