Burnout, AI Anxiety, and Staying Employable as a Developer
Honest notes on the fear quietly running through every 2026 developer Slack — and what's actually worked for me.
Every developer Slack I'm in has the same low hum running underneath every conversation: will I still have this job in two years?
Nobody says it directly. It comes out as sarcasm about AI demos, or as someone over-explaining why their team can't use Cursor, or as the fifth "I'm taking a break from Twitter" post this month.
This is what actually helped me, written down in case it helps you.
What 2026 actually feels like
"Engineers are humans carrying rent pressure, burnout, ambition, insecurity, and a growing fear of being left behind — and the biggest articles of 2026 understood that perfectly." — Sanajit Jana, Medium
Three things stacked on each other this year:
- AI got good at the parts of the job that used to feel like proof of skill. The "I can wire up a CRUD app in a weekend" thing isn't a flex anymore.
- The job market normalised. The 2021 hiring frenzy is gone. Layoffs are routine. The narrative isn't "everyone is hiring" — it's "nobody is sure what they'll need in six months."
- The advice churn accelerated. Every week there's a new take on what to learn, what's dying, what's the new moat. The signal-to-noise ratio is brutal.
The combined effect is a kind of low-grade panic that lives in the background of every workday.
Three things that didn't help me
- Doom-scrolling AI Twitter. Every demo looked like another nail in my career coffin. Nothing was learned.
- Buying courses. I bought four AI courses in 2024. I finished zero. The act of buying felt like progress; it wasn't.
- Avoiding the fear by working harder. Working 60-hour weeks on the same skills you already had doesn't address the underlying worry. It just buries it under tiredness.
Three things that did
1. Get specific about what you're afraid of
"AI will take my job" is too abstract to act on. Break it down:
Vague fear → Specific question
───────────── ─────────────────────────────
"AI is taking jobs" → "Which 30% of my current
tasks could a 2027 model
do without my supervision?"
Once it's specific, you can do something. The answer for me was: scaffolding, boilerplate, syntax-level fixes, basic test writing. So I stopped optimising my workflow around those things and started optimising around the parts AI is bad at — choosing what to build, talking to users, owning operational quality.
2. Pick one thing that compounds
Not five. One. For me: writing in public. One post a month, indefinitely.
The math: in three years that's 36 posts. Even if half are mediocre, 18 good posts is more than 95% of working developers will ever publish. That's a real moat — and it gets more valuable as AI floods the internet with generic content, not less.
3. Lower the stakes of any single day
The single most useful mindset shift: my career is a 30-year arc, not a quarter. A bad week, a missed promotion, a project that flops — none of it matters at the scale of decades. Almost every mistake is recoverable.
This sounds obvious. It is. It's still the thing that helped most.
The honesty filter
Run this on yourself once a quarter:
- Am I learning anything I couldn't learn last year?
- Am I shipping something a stranger could see?
- Do I know what I want my next job to look like?
- Have I talked to anyone outside my company about my work in the last 60 days?
If three of four are no, that's the warning. Not "you're getting fired" — "you're stagnating, and stagnation is the actual risk in 2026."
What I tell juniors who ask
Don't try to compete with AI on speed. Compete on judgment, communication, and ownership — the things AI is systemically bad at and will be for a long time. Then write about it. Loudly. Publicly. Consistently.
The developers who thrive in 2026 aren't the fastest typers. They're the ones with the clearest signal — and signal is built one piece of public work at a time.
