Let me tell you about a project that wasn't impressive.
Back in university, I built a small .NET application. I don't even remember all of what it did — some forms, some logic, a database behind it, nothing you'd put in a portfolio. By any technical measure it was tiny. A beginner's project.
But I remember how it felt to finish it.
I remember getting stuck — properly stuck, the kind where you stare at an error message that makes no sense and you have no idea what you did wrong. I remember googling, reading half-relevant forum answers from 2011, trying things that didn't work, and slowly, painfully, figuring it out. I remember the specific moment it finally ran — the thing I'd been fighting for hours just quietly worked — and I sat back in my chair and felt something I can only describe as pride. Real pride. I made this. I actually did something.
It wasn't about the code. It was about me. I had walked into something I didn't understand and come out the other side having understood it. That feeling was the whole reason I wanted to do this for a living.
I've been thinking about that feeling a lot lately, because I've realized I haven't felt it in a long time.
And now
Here's the strange part.
These days I build things that would have stunned that student. Bigger, more complex, more real than that little .NET project by every measure. I ship more in a week now than I probably shipped in a semester back then. By any metric you'd care to name, I am a dramatically more productive, more capable developer than I was.
And I feel almost nothing.
I finish things — real things, things that work, things people use — and there's no moment. No sitting back in the chair. No quiet "I did that." I close the laptop and open it the next day and it all kind of blurs together into one long, flat, competent stretch of output. The days don't feel like anything anymore. They just feel like days.
For a while I assumed it was burnout, or getting older, or the shine wearing off the way it does with any job. But it's not that. I've been burned out before and this isn't it. This is something more specific, and it took me an embarrassingly long time to see what it actually was.
The thing I finally figured out
The joy was never in the finished project. It was in the struggle to get there.
That's the whole thing. That's what took me so long to understand. I thought the good feeling came from having built the thing, so I assumed building bigger, better things would bring more of it. But the good feeling never came from the output. It came from the fight — from being stuck and getting unstuck, from not understanding something and then understanding it, from the friction of my own mind working against a hard problem until something gave.
The struggle wasn't the obstacle standing between me and the joy. The struggle was the joy. I just couldn't see it, because who thinks the hard part is the good part?
And here's what changed, quietly, without me noticing when: the struggle is mostly gone now. I describe what I want, and it appears. I hit something I don't know, and I don't sit with the not-knowing anymore — I ask, and the answer is just there, complete, before the confusion ever has time to become the thing I push against. The wall I used to climb has been demolished, helpfully, in advance.
Which sounds like a gift. It ships faster. The code is often better than what I'd have written stuck at 2am. But the wall was where the feeling lived. Take away the wall and you take away the climb, and it turns out the climb was the part I loved. I optimized away the friction, and the friction was the point.
What I think we're actually losing
I don't think this is just me. I think it's happening quietly to a lot of us, and we're not really talking about it, because it feels ungrateful to complain that your job got easier.
But something real is going. The small, private satisfactions that made this feel like a craft instead of an assembly line. The bug you chased for three hours and finally cornered. The elegant little solution you found yourself and were quietly proud of all day. The thing you understood so deeply it started to feel like it was part of you. Those moments are getting rarer, because the machine is very good at producing the output of those moments and completely unable to give you the moment itself — and the moment was always the reason.
In tech, we're slowly losing the things that were ours. The hard-won, personal, slightly private sources of joy that had nothing to do with productivity and everything to do with why we got into this. And because what replaces them is faster and shinier and more impressive, we're calling it progress, and mostly we're not admitting that something we loved is leaving in the trade.
I don't have a fix, and I'm not going to pretend I do
This is the part where a post like this is supposed to turn around and offer you five ways to fall back in love with coding. Do a passion project. Code without AI on weekends. Rediscover your spark.
I'm not going to do that, because I don't believe it, and I think you'd know I was lying.
The truth is you can't un-know that AI exists. You can't really opt out — the pace of the work assumes it now, and "just struggle for fun on your own time" doesn't survive very long against a job that measures you on output. And even when I do deliberately do something the hard way, it's not quite the same, because I know the easy way is sitting right there and I'm choosing the friction on purpose. Manufactured struggle isn't the same as the real thing. The innocence of having to figure it out — that's gone, and you can't fake your way back into it.
So I'm not going to tell you how to get the feeling back. I don't know how. I'm not sure it comes back.
So
Here's where I actually land, which is not anywhere comfortable.
Something I loved is fading, and I think a lot of us feel it and haven't said it out loud, because it sounds soft, or ungrateful, or like we're just bad at adapting. But I don't think naming it fixes it, and I'd rather be honest than useful here. A specific joy — the one that made a tiny .NET project feel like the biggest thing in the world because I built it, stuck and unstuck and stuck again until it worked — that joy is quietly leaving the work, and I miss it, and I don't have anywhere hopeful to put that.
Maybe that's okay. Maybe some things are allowed to just be a loss. Maybe the most honest thing we can do right now isn't to fix the feeling or spin it into a lesson, but to sit here together for a second and admit that once, finishing something small with our own two hands and our own stubborn mind felt like everything — and that lately, it doesn't.
I don't know what to do with that. I just didn't want to pretend I couldn't feel it.
Tell me about the project that made you feel it — the small one, the unimpressive one, the one that wouldn't mean anything to anyone else but felt huge because you built it yourself and it was yours. I want to remember with you what this used to feel like. Mine was a little .NET app that did almost nothing. What was yours?
Top comments (35)
This is so real, but as you said. I dont think that there is a fix for this. so just let it happen. there would be a time when coding would be totally dead, and no one would even notice this layer of technology.
exactly and soon coding would become somewhat ancient to everyone's ears.... prompting would be the new coding for sure
maybe! we don't know what will happen
Ok i will read you article !🥲
this wasn't my intention
I know brother, chill!
Maybe — though I hope you're wrong about the "no one noticing" part; even if the coding itself fades into a layer nobody sees, I'd like to think someone still remembers it used to feel like magic.
I hope so
😊
My first website was a photography website that i created with basic html and css, and little to none JS i even documented it on the #100daysofcode challenge (I don't even know if that's a thing anymore because of ai) [(x.com/maame_codes/status/155543594...]
I spent about three months in doing that project, it wasn't anything that would wow anybody, but i will never forget the feeling of accomplishment i felt when i did that website, the dopamine is all fading away these days because of the swift change in the way we code now... that dopamine rush you feel after actually spending hours to figure out why your code isn't working maybe?
I started feeling like that as well when ai started evolving... because I thought to myself, 'what do you mean i don't have to personally check if i missed a semi colon somewhere anymore'... and also people 'shipping' apps in weeks, some even in days. I think the ai era has really changed our perspective on things. And now if your code doesn't work, all you need is a good prompt to 'figure' it out for you. There isn't any 'aha!' moments anymore when coding because all you need is a good LLM to solve it for you. I always say its such an interesting wave in tech that we would experience in the next decade to come. This was a good read James! Brought back some good memories
A photography site in plain HTML and CSS, three months, documented day by day on #100DaysOfCode — that's exactly the thing, and the fact that you'll never forget the feeling even though it wouldn't wow anyone is the whole point. "What do you mean I don't have to check if I missed a semicolon anymore" is the perfect small example of it: the tiny frustrations were where the aha lived. The dopamine came from the not knowing, then knowing. I don't have an answer either, but I think the people who felt it — like you clearly did — will find some way to keep reaching for it, even if the shape changes. Thank you for sharing the memory; that's the realest part of all this.
I can relate to this. AI definitely makes coding faster, but sometimes the process of struggling with a problem and finally solving it is what makes coding satisfying. The challenge may take longer, but it also helps build deeper understanding and confidence.
Satisfaction came through the process of building, the outcome however is what we already expect since the beginning we start coding.
Satisfaction is lost when our contribution is nearly zero in building something.
Exactly — the struggle takes longer but it's where both the satisfaction and the understanding come from, and speeding past it quietly costs you both.
For me the stuck part just moved somewhere else. Now it is working out that a thing was the wrong thing to build, which takes longer and never ends with anything turning green. I miss knowing when I was finished.
That's a sharper loss than the one I wrote about — the struggle didn't vanish, it moved from "how do I build this" (which ended in green) to "is this even the right thing" (which never resolves), and losing that clean moment of done is its own quiet grief.
Quiet grief is about right. What I did not expect is that I now have to make the done moment up on purpose.
I write down the thing I am going to call finished before I start, even when it is small and a bit arbitrary. Otherwise the day just stops instead of ending.
it is a thinner feeling than green tests. still better than nothing.
Manufacturing the "done" moment on purpose is a quietly sad thing to have to do — but naming the finish line before you start, even arbitrarily, is a real adaptation. Thinner than green tests, like you said, but a chosen ending beats a day that just stops. There's something almost defiant in it: if the work won't hand you the moment anymore, you make one anyway.
I think we are confusing efficiency with understanding. The joy wasn't just in solving the bug; it was in the process of elimination—the painful, manual narrowing down of possibilities until only one path remained. I remember spending 12 hours tracking down a race condition caused by an outdated library dependency that nobody had documented. That deep dive taught me more about concurrency than any textbook ever could.
That's what felt as a real achievement , which is fading now!
The line that hit me: "The struggle wasn't the obstacle standing between me and the joy. The struggle was the joy."
I'm on the opposite side of this. I started learning Python about two weeks ago — and I've felt more of that pride in the last 14 days than I have in years at anything else.
Not because I built anything impressive. I built a calculator that adds two numbers. I made a script that prints my name. I figured out why range(1, 6) stops at 5 and spent 20 minutes genuinely stuck on it, frustrated, convinced I was too dumb for this — and then the moment it clicked, I literally said "OHHHH" out loud in my room.
Reading your post, I realized that feeling you're describing — the one that made the .NET app feel huge — is the exact thing that's keeping me going right now. And it's also the thing I know I'll lose eventually.
What terrifies me a little is this: AI can teach me faster. But if I let it skip the struggle, I lose the pride. So I've been trying to do this weird dance — use AI for syntax, use my own head for the hard problems.
I don't know if that balance holds as I get better. But your post made me want to protect it.
The .NET app with the forms and the database — I bet you still remember exactly what broke first.
Mine was a to-do list that couldn't delete items. It took me two days to figure out why. I've never been prouder of anything.
The "OHHHH" out loud in your room — that's the whole thing, and reading you feel it right now, two weeks in, over a calculator and a range() that stops at 5, is honestly the most hopeful thing anyone's said in this thread. You've already found the answer I couldn't: use AI for syntax, keep the hard problems for your own head — that dance is the protection, and the fact that you're consciously guarding it means you might hold onto it far longer than I did. And a to-do list that couldn't delete items, two days to crack, proudest thing you've built — that's not small. That's exactly it. Protect that feeling fiercely; it's the realest part of all this.
Until some years ago, I felt the same - or, at least similar. At some point I figured out, that I barely create something new. I implemented so many things over the last 20 years that there is rarely something new coming.
The further I went, to more focus I spent on the outcome and the product. What I enjoy now - similar to what you describe - is, steering AI in a way that it is able to fulfill exactly what I want. If the one-shot result is sufficient, that gives me the same vibes. So actually, not much of a change :-)
That's a genuinely hopeful angle on it — after 20 years the novelty of the code itself fades anyway, so the joy naturally migrates to the outcome, and steering AI to land exactly what you envisioned becomes its own version of the craft. Maybe the feeling doesn't die, it just moves — from "I solved it" to "I got it to build precisely what I saw in my head." Thanks for this; it's a kinder read than mine.
I read every single word of your post. This morning I logged in here again after years. I came here today thinking "how boring is to build today. Before it was like to be a magician. Now, everything can be done". And here's your post for me.
I started coding when I was teenager. Then university, then real work. I spent hours, nights, trying to get a solution. Trying to get a clean, smart solution. To be a programmer it was like to have superpowers.
Now I just ask, and it's there. Ready to be use, working, pretty well, and tested. Done, end of programming.
I don't have the solution either. On my side, I think the enjoyable part could be to put something of us into that project. Nowadays wont be the code, but maybe the design, or the goal of the project.
Or maybe this. What I felt reading your post. Maybe the enjoyable part is just this. Sharing with humans, talking with real people. Saying "hey I loved your post, I read it, I didn't ask an AI to summary". So... hey! I ready your post and I love it, I didn't use an AI to understand it. Thanks for sharing! ☺️
This comment is the piece's whole point, living — you logged in after years feeling exactly this, and you wrote to a human about it instead of asking an AI to summarize. That last line got me. Maybe you're right: when the code stops being ours, the connection is what's left — putting ourselves into the design, the goal, or just this, two people saying "I felt that too." Thank you for reading it as a person. It means more than you know. ☺️
I think there’s an interesting difference between removing friction and removing ownership. AI can eliminate the frustrating parts of debugging, but those same moments were often where developers built a mental model of the system that stayed with them afterward. The harder part wasn’t simply getting the code to work; it was having enough time and uncertainty to understand why it worked. That makes me wonder whether the real loss isn’t “coding without struggle,” but fewer opportunities to form those personal connections between problem, reasoning, and solution. The output can still be better and faster while the developer’s understanding of how they got there becomes thinner. That feels like a much more subtle trade-off than productivity alone captures.
"Removing friction vs. removing ownership" is a sharper cut than mine — the struggle wasn't just where the joy lived, it was where you built the mental model that stayed with you. So the real loss might be quieter than "no more pride": it's fewer chances to form that personal link between problem, reasoning, and solution, leaving your understanding thinner even as the output gets better. That trade-off doesn't show up on any productivity chart, and you've named it better than I did.
Although not coding, I remember having that joy when I rewrote a partition table I had accidentally deleted (which recovered the layout of a 300gb bitlocked ntfs partition that was full of data) using a few linux CLI, At that time I knew very few thing about linux & nothing about the CLI tool I used to recover the partition table, it needed a lot of search to know how to do it, & I learned much about filesystems & partitioning because of that (took me about a week, nowadays I think It can be done in 30 minutes with webAI, & less with Agentic AI), I just think we are moving to a higher abstraction layer instead of eliminating software engineering, & I think the wall to struggle through can still be there but nowadays, it must be something much bigger with much more unknowns to struggle with, I think if something new enough to you that it makes you move abit from the area of agenting engineering to middle way between agentic engineering & vibe coding, you would struggle abit in a useful manner that makes you feel the same conventional feeling of growth of "I did it" that you felt back then, maybe I'm wrong, but AI helped me reach such feeling way more than I did before AI, but that is just my hoppyist POV.
Recovering a 300GB partition table over a week, knowing nothing about the CLI going in — that's the feeling exactly, and it stayed with you because the struggle taught you filesystems as a side effect. Your reframe is a hopeful one I keep coming back to: maybe the wall didn't vanish, it moved up — you now have to reach for something bigger and stranger to hit real unknowns. And that "higher abstraction, not eliminated" read feels true. If AI has pushed you toward harder, newer problems and you're feeling the "I did it" more, that's not a hobbyist's small point — that might be the actual answer.
Some comments may only be visible to logged-in visitors. Sign in to view all comments.