Making Software Is So Easy, I Stopped Doing It

May 31, 2026

making software has never been easier. i mean that literally. with opus 4.8, the cloud, the tooling we have now, the distance between an idea and a working, deployed product is basically zero. i can think of something in the shower and have it live by lunch. and for a while, that's exactly what i did.

every week i was building something new. an app i wished existed. a tool to scratch some itch. a little experiment i got curious about. everything that had ever lived in my head as a "someday maybe" was suddenly a saturday afternoon. it felt incredible. like a superpower. i even wrote about how i hadn't coded in three months and somehow shipped more than ever. this was the logical endpoint of that.

and then i stopped.

the building stopped meaning anything

not because it got hard. the opposite. it got so easy that it stopped meaning anything. when you can build anything, building a thing isn't an achievement anymore. it's just output. i was producing software the way a printer produces pages. fast, endless, and after a while, kind of numbing.

i'd finish an app and feel nothing. open a new file and start the next one. i wasn't choosing what to build. i was just building because i could. execution had become a reflex, and somewhere in there i'd stopped actually thinking.

so i added friction on purpose

i did something that probably sounds backwards. i downgraded my subscriptions. cancelled some tools, dropped to smaller plans. not to save money. to slow myself down. to make execution a little less frictionless so i'd be forced to think before reaching for the keyboard. to observe more than i produce.

because the bottleneck was never the tools. it was me, and what i was pointing them at.

answers are free now

here's something i've been repeating to people for weeks. getting an answer used to be the hard part. you had to know things, dig around, figure it out. now it's basically free. you can get a decent answer to almost anything in a few seconds. so the value quietly moved. it moved to the question. when answers cost nothing, the only thing left that's worth anything is knowing what's worth asking, and that part hasn't gotten any easier.

answers are free now. the question is the only part still worth anything.

what i actually do now

i still make software. i have to, it's my job, and i genuinely love the work. and yeah, every so often i still build something for myself on a weekend, when it actually matters to me. i'm not romanticizing doing nothing.

but the default has flipped. instead of asking "what can i build," i spend more time on "what's worth building at all," and honestly, "what's worth paying attention to." i'd rather sit with where humans are heading right now than add one more app to the pile.

find your spot instead

here's the thing i keep landing on. if making software is this easy for me, it's this easy for everyone. when everyone can build anything, building stops being the point. what you choose to build does.

i have friends right now shipping real software who have never written a line of code in their life. no degree, no background, none of it. and look, i'm not going to pretend it's all clean or that it scales. some of it is held together with tape. but it works, it's theirs, and a couple years ago it just wasn't possible. the building was never the hard part for them. figuring out what was worth building was.

the world is about to be flooded with apps, most of them fine, most of them forgettable. and the uncomfortable truth is that the big ai labs are going to absorb most of that surface area anyway. the generic crud app, the wrapper, the little tool. that's not a moat. it's a feature someone's model ships for free next quarter.

i wrote about the ai hangover a few months ago, and this is the same thread pulled further. the answer to "everyone can build everything" isn't to build more. it's to figure out where you specifically fit. what you see that others don't. what you care about enough to have an actual point of view on.

so i'm trying to place myself in the future instead of racing through the present. less execution, more judgment. less output, more taste. i don't think the people who matter over the next few years will be the ones who built the most. they'll be the ones who understood the moment well enough to build the right thing, or to know when not to build at all.

making software got so easy that i stopped doing it for its own sake. and weirdly, that's the most useful decision i've made in a while. the tool got so good that the tool stopped being the question. what you do with your attention, that's the whole game now.