Here's something nobody tells you about being a software engineer: the better your tools get, the faster you burn out.
Sounds backwards right? We have AI that writes code, IDEs that autocomplete entire functions, tools that let us manage 10x the context we could a year ago. We should be thriving. Instead, a lot of us are just moving faster toward the same exhaustion.
The problem isn't the tools. It's that we forgot how to be slow.
The Speed Trap
Let me paint you a picture of modern software engineering, at least from what I've experienced:
You're solving 10 problems in the time it used to take to solve one. Context switching between user flows and backend functionality like you're flipping browser tabs. Every bug fixed reveals three more. The AI suggests a solution before you finish typing the question.
It feels productive. It feels like leverage.
But here's what I've noticed actually happens:
Your mental clutter compounds hourly, brain fog accumulates.
You race down the wrong path because backtracking feels like failure
You overestimate your abilities and thus keep pushing forward but don’t backtrack to fundamentals (there's research on this for AI-assisted work)
You end up at a point of no return, staring at last week's commit, wondering where it all went wrong
This is burnout, before it fully hits. And most of us don't see it coming because we're moving too fast to notice.
What Running Taught Me About Code
Since running, I can’t help but notice that coding and running are complete opposites, yet complementary. Like Yin and Yang:
Coding is chaos with infinite possibilities:
Static body, racing mind
Endless decisions every session
Creative complexity with unclear outcomes
Progress that can vanish instantly
Tools that give you 100x leverage
Running is boring simplicity:
Moving body, calm mind
One decision: keep going
Repetitive action with predictable results
Progress that compounds reliably
No shortcuts to a 2-hour marathon (you need a lifetime of training)
The boredom is the point. That's what I missed.
The Two Things Running Gives You
1. A longer horizon for expectations
Running teaches you that real improvement takes time. You can't hack your way to a marathon PR. You can't automate the work (an unfortunate curse on technical people). You just show up, consistently, and trust the process.
This perspective bleeds into how you view code. Suddenly you're less attached to yesterday's work. Restarting doesn't feel like failure. You think in years, not sprints.
2. A shorter view on actionable items
During a run, every step brings you closer to the finish line. It’s like a stritly positive monotonic function. One foot in front of the other.
In code, you start thinking the same: one bug fix at a time. One function. One test. Breaking down the massive system into digestible pieces instead of trying to hold the entire architecture in your head at once.
The Long-Term Engineer
There's a type of engineer I’m started betting on. They have a long horizon view on their code's value without getting lost in dreamy scenarios. They focus on:
Leveraging AI tools intelligently
Building and maintaining software repositories
Learning new skills continuously
Creating an environment that favors deep work
Taking breaks to recover and reflect
Letting their mind wander
I'd bet on these people over the ones grinding for months without sleep. I find it hard to believe you don't get led astray at some point if you never step back.
(Yes, there are exceptions. Short bursts of deep flow, like what I'm in right now writing this, or even weeks when the vision is crystal clear and you have a roadmap or someone steering the ship. But that vision must be clear, or you're just grinding for grinding's sake.)
The AI Paradox
Here's my controversial take: AI tools might be making us worse engineers if we don't change how we work.
Not because the tools are bad. They're incredible. You can task switch between contexts that would've taken hours to load into your brain. Your fundamental understanding of individual functions might decrease, but your ability to understand broader systems and overall functionality heightens dramatically.
The issue, at least from what I've seen, is we're using this newfound power to go faster in the same way we always worked. We're sprinting when we should be pacing.
What actually happens with AI tools:
You can manage more context (10x the files, tokens, complexity)
You elevate to higher-level thinking (architecture over syntax)
You adopt new perspectives on a larger scale
You become a "10x coder" not by working 10x harder, but by thinking 10x bigger
But only if you pair this with recovery. With movement. With boredom.
Why Software Engineers Will Become Fitter (Or Fail)
My prediction: the best software engineers will become noticeably fitter over the next decade. Not as a nice-to-have, but as a competitive requirement.
This may be complete bias, as an engineer who loves to exercise, but hear me out:
The math is simple:
AI gives you time back
You can either fill it with more coding (the burnout path)
Or fill it with complementary activities that make your coding better (the sustainable path)
Looking back at past projects that didn't work, I put so much into them that when faced with an obstacle, I faced significant resistance to overcome it. Now, I put slightly less effort every day but get more results and feel better overall. When resistance arises, I feel less attached to my past work. Restarting is easier.
The best engineers will:
Move more and go outside regularly
Occupy themselves with tasks opposite to coding's nature: art, music, etc.
Lead nicer lives and paradoxically work less
Design their lives intentionally (high salaries make this possible)
This field used to be for nerds and esoterics. It's now opening up for freedom-chasers and people who want accountability over their lives. Digital nomads were just the beginning. I'm building proof of this.
Lacing Up
Even if my bet isn’t right, the fact that many of the fundamental AI models are designed for coding and adapted to an engineer’s work, must imply an improved productivity of these engineers. This will remain a valuable skill and therefore contribute to higher salaries and time back. Following this train of thought, if you're a software engineer who spends hours at a desk, as I occasionally do, start doing this:
Start today:
Take breaks (not productive ones, real ones)
Work standing sometimes
Optimize your environment for deep work, not constant output. Check out brain.fm, get earplugs or invest
Start this week:
Begin running or any boring, repetitive movement
Embrace the simplicity
Notice how your thinking changes
Remember:
Exercise isn't time off work, it's a complement that aids your work
Physical health impacts mental health, which impacts code quality
The mental clutter from coding never ends unless you deliberately end it
The beautiful harmony:
Both coding and running share something: long-term expectations paired with short-term actionables. The consistent base of your longer vision combined with the rhythmic melody of daily actions and ups and downs.
You need both. The vision and the steps. The sprint and the marathon. The code and the run.
If you're a software engineer or CTO, I'd love to hear your take.
"Slow is steady. Steady is smooth. Smooth is fast. Fast is deadly.”