Table of contents
Open Table of contents
The graveyard of ideas
My Notes app is a cemetery. Hundreds of ideas, half-written, abandoned. Some were genuinely bad ideas — an AI-powered toaster app, a better calendar (as if the world needs another calendar). But some were good. Really good. Ideas that, if I’d spent just a few hours on them, might have become something.
They died for the same reason most side projects die: the gap between “this would be cool” and “let me open my editor and start typing” is filled with psychological friction. Is this worth my time? What if it’s harder than I think? What if I spend a weekend on it and it goes nowhere?
That gap used to take real energy to cross. Now it doesn’t. Because Claude is there. At 2am. Without judgment.
The cost of starting dropped to zero
Here’s what used to happen when I had an idea at 2am: I’d write it down, promise myself I’d look at it on Saturday, and by Saturday the momentum was gone. The idea felt smaller in daylight. Less urgent. Buried under actual work and actual life.
Here’s what happens now: I open Claude and I describe the idea. Not in a polished spec. Just a brain dump. “Hey Claude, I’m thinking of building X that does Y. Is this feasible? What would the architecture look like? Can you sketch the first version?”
And within minutes — sometimes seconds — I have something tangible. A working prototype. An architecture diagram in text. A list of trade-offs I hadn’t considered. The idea goes from abstract to concrete while the energy is still there.
The 2am version might be ugly. It might be wrong. But it exists. And something that exists can be improved. Something that’s still in your Notes app cannot.
The best thing Claude did wasn’t write better code. It was remove the activation energy between having an idea and seeing it work.
The judgment-free zone matters more than you think
There’s something specific about Claude that matters here, and it’s not the model quality. It’s the absence of social friction.
When you pitch an idea to a colleague, you’re managing impressions. Is this smart enough? Will they think it’s obvious? Am I wasting their time? Even the most supportive colleague creates some of this friction — it’s human.
Claude doesn’t care if your idea is stupid. It doesn’t care if you’ve asked the same question three different ways. It doesn’t care that it’s 2am and you’re in pajamas and your spec is three misspelled sentences long. It just engages with the idea.
This sounds trivial. It’s not. Most creative work dies in the gap between private thought and public exposure. Remove the exposure, keep the engagement — and more ideas survive long enough to prove themselves good or bad on their own merits.
The permission to be messy
I used to think I needed a clean spec before I could start coding. Clean requirements. Clear architecture. A proper understanding of the problem.
Claude taught me that messy input is fine. A rambling brain dump with contradictions and half-thoughts works better than a polished spec, because the mess contains the real texture of the problem. The contradictions are where the interesting decisions live. The half-thoughts are the things you’ll need to figure out anyway.
Now my workflow is: dump the mess, let Claude organize it, read the organized version, spot what’s still missing, refine. The first draft of anything doesn’t need to be good. It just needs to exist.
This is a mindset shift that extends beyond AI. Once you get comfortable producing messy first drafts, you produce more first drafts. And volume beats perfection — always has.
What I’ve shipped because of this
Over the last year, I shipped more small side projects than in the previous three combined. Not because I had more time or more skill. Because the barrier to starting collapsed.
A few of them would have died in my Notes app: a natural-language-to-SQL tool for our internal data team, built in one late-night session and used daily now; a dashboard that pulls GitHub notifications across thirty repos into one place. None of these are startups. None will make money. But they solve real problems for real people, and they exist for one reason — the cost of starting got low enough that I actually started.
The caveat
This only works if you still exercise judgment. Claude will help you build a bad idea just as enthusiastically as a good one. The judgment — “is this worth building at all?” — is still yours. The filter between “I could build this” and “I should build this” hasn’t moved.
What has moved is the gap between “I should build this” and “it exists.” That gap went from days or weeks to hours or minutes. And when the gap is that small, you build more. Most of it will be forgettable. Some of it will be great. But you can’t get to the great stuff without wading through the forgettable stuff first.
Claude just makes the wading a lot faster.