AI Fatigue Is Real - Especially for Developers
March 11, 2026
AI has changed how I work as a developer. I'm still building. I'm still shipping. In many ways, I'm getting more done. But I've also noticed something unexpected. I'm more tired than I used to be. Not burned out. Not anti-AI. Just... fatigued. This is a reflection on how AI shifts developers from writing code to managing it - and how that subtle change affects energy, confidence, and flow. Slow Builds: code, money, and life - without pretending we've arrived.
Transcript
Opening (00:00)
Hey, welcome to Slow Builds.
So I've been coding for a long time and most of my career has been spent deep in code, building, debugging, refactoring.
The flow state where you deep dive and disappear into a problem and when you finally resurface a few hours later, sometimes a few days later, you got something working and you're proud of.
And lately I've been more tired than usual.
Not burnt out.
Not overwhelmed.
I'm just very... I feel fatigued. I feel tired.
And if I'm being honest, AI is part of it.
And maybe even the cause of it actually.
I'm pretty sure.
Now I do feel that it's the major part and piece to why I'm feeling so fatigued all the time now.
When AI Feels Like Leverage (01:10)
At my job I work on a code base and I know it very well.
The parts that I interact with and build on I'm very familiar with.
And in that environment AI is very amazing.
If I let it build something inside a system I've lived in for years, I can review it quickly.
I know the patterns.
I know the edge cases.
I know what normal looks like in that code base.
And I can quickly see if something doesn't fit.
So at work and in that code base I can let it help me with things like:
- building documentation
- logging issues
- updating comments
- generating test cases
- creating new methods
- helpers
- small things
Sometimes big.
I can give it a full page and say refactor this or I need to add a new method that does this.
And I feel comfortable because I can quickly look at what it did, compare the differences, and I can point out if it went off the rails a little bit.
And to me that's full leverage.
That's taking a new tool and making my life better and more efficient.
When AI Changes the Role (02:40)
But when I'm building something new...
When I need to put a new integration into work code...
When I need to build a brand new tool inside of our system...
Or integrate something completely new...
Or the biggest part - personal projects.
When I'm starting from scratch it's a whole different ballgame.
AI is not just writing the code.
It's making decisions.
Sure, I give it an idea.
I give it some guidelines.
I open the door.
But I don't always know the mental map yet.
I don't always exactly know what the end result is going to be or what it should look like.
But by letting AI start working on it, it kind of assumes what the end is without me actually telling it.
So instead of just building, I'm reviewing a lot of assumptions.
I'm double checking a lot of edge cases.
Retesting everything over and over.
Comparing different approaches.
Comparing different branches of code just to see what differences there are.
Making sure nothing slips in.
No side effects.
No legacy code affected.
Everything backwards compatible.
So I'm less in a flow.
I'm more in supervision mode.
I've become a manager.
I'm not a coder.
The Verification Loop (04:15)
When I write something myself, I trust the reasoning behind it because I've walked through it in my head.
But with AI, even if it looks right, there's always this little nagging voice:
Did it miss something?
Did it break something?
So I review.
Then I test.
Then I'm rereading.
Then I'm comparing it to another approach.
So it becomes:
Generate -> doubt -> verify -> adjust -> verify again.
Rinse and repeat.
And that loop is tiring.
AI speeds up output.
Code comes quickly.
But pushing the code - deploying the code - comes slower because there's a lot of review in the middle.
And that review process is tiring.
Now I will say it makes the code better.
And it makes me better at spotting things.
But it's still a change in mindset.
And that change is where the fatigue is coming from.
The Constant Shifting (05:50)
There's another layer to this fatigue.
Every time I feel like I'm using AI right...
I feel like I finally have a workflow...
I read something online.
Or I talk to a colleague.
Or I watch a YouTube video.
And suddenly someone is using a completely different workflow.
A new tool drops.
A new model replaces something.
A better prompt strategy exists.
People are shipping systems faster.
Making money faster.
Getting traction faster.
And suddenly it makes you feel like everything you've been doing is wrong.
In my case I start considering scrapping everything.
Relearning everything.
Re-architecting everything.
Going deep into research mode.
Trying to figure out how to switch my process.
And this literally just happened to me this week.
I lost two days.
A lot of late nights.
And I scrapped a bunch of work.
All because another approach looked better.
And with AI right now it feels like the ground is constantly shifting.
It creates this constant low-level feeling of being left behind.
Even when I'm objectively getting more done.
Doing Something Is Better Than Nothing (08:10)
One motto I've always had in life is:
Doing something is better than doing nothing.
If you're not moving, you're dying.
Whether that's health.
Learning.
Or building.
And with AI sometimes the movement itself keeps changing direction.
But even if you're not doing it the best way...
You're still doing it.
You're still learning.
You're still building.
You're still part of this new shift.
And that matters.
The Energy Shift (09:20)
What's strange is...
More is getting done.
But I feel more drained doing it.
I used to get tired from building.
That was like a physical tired.
Like working out.
Now I'm tired from managing.
Reviewing.
Double checking.
Supervising.
It's a mental drain.
Sometimes it even feels like nothing got done.
Because I'm not the one writing the code.
It feels like there's this black box.
And I'm just poking holes into it to inspect things.
I'm less immersed in the code.
And much more in review mode.
And reviewing is exhausting.
The Pressure of Infinite Work (10:50)
I used to open my editor excited to knock out a bug or build a feature.
Now sometimes I hit a wall of tired before I even start.
And there's this background pressure:
"I should get back to it."
Even when I'm watching a basketball game.
Or going for a walk.
It feels like I should be working.
Because the work feels infinite.
There's always another refinement.
Another improvement.
Another optimization.
Another better way.
And it never ends.
Where I'm Landing (12:10)
I'm definitely not anti-AI.
If anything I'm a massive cheerleader for it.
I want to use it in every part of my life.
I use it every day.
Research.
Prototyping.
Organizing thoughts.
Summaries.
Analysis.
It genuinely helps me.
But I think I'm still rebuilding my energy system around it.
My role is changing.
From developer...
to something closer to a project manager.
And that's not what I trained to be.
Finding a Stable Path (13:20)
I don't need the most optimized workflow.
I need a stable one.
I don't need every new tool.
I don't need the flashy new way of doing everything.
I just need to pick a lane...
Stay with it...
And build confidence.
Because AI fatigue isn't about the tool.
It's about the role changing.
The environment changing.
The speed of everything changing.
Closing Thought (14:30)
So I'm still here.
Still adjusting.
Trying to take it one tool at a time.
Trying to stay involved in the code.
Trying to work with AI instead of just managing it.
And honestly...
I'm still figuring it out.
So if you're feeling AI fatigue too...
That might actually be a good sign.
It probably means you're in it.
You're learning it.
You're experimenting with it.
And that's better than running away from it.
Thanks for watching.