🤔 AI made building easier. It also made saying "no" much harder.
One thing I've noticed over the last year is that AI has changed the cost of experimentation.
A feature that might have taken me a week to prototype can now be built in an afternoon.
At first, that felt like a huge advantage.
But after a while, I ran into a different problem.
Because everything became easier to build, I started building too many things.
Every new idea felt worth trying because the cost was so low.
Ironically, the real challenge stopped being "Can I build this?" and became "Should I build this?"
Now, before opening my editor, I try to answer a few questions:
Does this solve a real problem?
Is this based on actual user feedback?
Am I building this because users asked for it, or because AI made it easy?
I'm still figuring out the right balance between moving fast and staying focused.
I'm curious how other makers have adapted.
Has AI changed the way you decide what not to build?
Or do you think rapid experimentation is always the better approach?
I'd love to hear how others are navigating this trade-off while building products.
Replies
disagree a bit — some of my best features came from stuff nobody asked for. people don't know what they want until it exists
@charlotte_reed1 That's a fair point. Some breakthrough features definitely weren't requested directly. I'm curious—how do you decide which intuition is worth betting on versus which ideas to ignore?
four half-built features on my end right now, zero shipped, because i kept starting new ones instead of finishing
@evelyn_clark1Â I can relate. AI made it so easy to start new features that finishing became the real challenge. Have you found anything that helps you stay focused on shipping?
before ai, slow builds filtered out the bad ideas for you. now nothing filters itself, you have to do it manually.
@harper_cole That's exactly what I've been feeling. The friction of development used to force prioritization. Now the constraint seems to be our own decision-making rather than engineering time.
the "am i building this because users asked or because ai made it easy" line is going on a sticky note.
@aaron_cooper8Â I'm glad that line resonated. It became a question I started asking myself before opening my editor because I realized "easy to build" isn't the same as "worth building."
hits hardest in agency work — clients ask for stuff just because they saw it's "easy now" and you end up scope creeping yourself
@hazel_brooks That's an interesting angle. I hadn't thought about agency work, but I can see how AI changes client expectations around scope. Have you changed how you handle feature requests because of that?
curious what triggered this for you specifically, one bad build or more of a slow realization over time.
@chloe_madison1Â It was more of a gradual realization. I noticed I had several prototypes that were technically impressive but weren't solving problems anyone had actually prioritized. That made me rethink how I was choosing what to build.