I have a folder on my machine called projects. It has 34 directories in it.
Maybe six of them are alive.
The rest are quiet. Some have a README with a big vision. Some have a working prototype and zero users. Some just have a main.ts and a dream. I stopped counting the dead ones a while ago because the number stopped being funny.
why I kept starting things
Some were built for fun. Some to learn a new stack. Some because I had an idea at 11pm that felt urgent.
That part is fine. Starting things is how I learn. I do not regret the graveyard itself.
What changed was AI.
With AI, I can move faster than I can think. I can start a project in a domain I barely understand, get past the "I have no idea what I am doing" wall, and have something running in a weekend. That used to take weeks. Now it takes a Saturday.
The problem is that the cost of starting dropped to almost nothing, but the cost of finishing stayed exactly the same.
AI made the cycle faster, not better
Before AI, I would abandon a project because it got hard. The implementation was too slow, or I hit a wall I did not know how to climb.
After AI, I abandon projects for a different reason. I exhaust myself.
I read more. I architect more. I over-engineer features for users that do not exist yet. I build the settings panel before I have a single person who needs a setting changed. The implementation is fast. The imagination is the problem. I run out of energy before I run out of ideas, and then I stop.
The graveyard did not shrink. It just filled up faster.
so I built a tool to fix it
At some point I got self-aware enough to notice the pattern. I kept abandoning things not because I lost interest, but because coming back to a cold project felt heavy. The context was gone. The momentum was gone. Starting over felt easier than continuing.
So I built KeepGoing. A tool that lives in your menu bar and terminal, watches your project momentum, and nudges you back with one small next step when something goes cold. No notes to write. No setup. Just a quiet companion that makes coming back feel light.
It was the most personal thing I had ever shipped. I built it because I needed it.
then I abandoned it too
Months in, I had over-engineered it. Good bones, real users, solid core. But I had spent 80% of my energy on the product and maybe 2 to 5% on getting it in front of people. I told myself I would do marketing after the next feature. Then after the one after that.
I never got there.
Eventually I had a new idea. A new target. And KeepGoing quietly joined the graveyard it was supposed to prevent.
I still feel guilty about it. Not because it failed, but because I knew exactly what I was doing and did it anyway. I diagnosed the disease and still caught it.
the thing I actually learned
I am not going to tell you to never quit. Sometimes quitting is right. Sometimes a project was just practice, and practice is fine.
But if you care about something, the last 20% is the only part that matters to anyone else. The architecture, the feature set, the clean code, none of that is visible from the outside. What is visible is whether you shipped it, whether you told people it exists, whether you showed up for it after the exciting part was over.
Get it out the door. Then tell one person. Then another.
I did not do that with KeepGoing. Now I keep one post a day going for it, no new features, just a small signal that it is still alive. It is not enough to grow it. But it is enough to not let it die quietly.
That is the rule I have now: if you build something you care about, give it at least that much before you move on.
