Burnout is a system problem.

You can't out-resilience a broken system.

Maybe you heard me at a conference, caught a podcast, or saw a post. This page is the longer version: how I think about burnout in tech, what AI is changing, and what leaders can actually do about it. Plus talks, podcasts and tools in one place.

Start here

Four Dimensions of Burnout: an anti-burnout framework for the AI era

LDX3 by LeadDev, London, June 2026

The whole idea in one talk: what burnout actually looks like, where it comes from in the way work is set up, and three steps to do something about it. Map, locate, apply.

Slides and materials from the talk
Where it comes from

It lives in the system, not in the person.

Mindfulness, coaching, wellbeing perks. They are all good things and they help individuals. But they don't redesign the place people go back to every day. You rest, you recover, and then you come back to the same conditions.

In most cases the problem is not how people work. It's how much, and how chaotically, they are asked to work.

The anti-burnout framework has three steps:

1Step 1: Map

Burnout is more than being tired.

It usually shows up as a combination of a few patterns. The first three come from how research describes burnout. The fourth is one I added, because I keep hearing it in my work, especially now.

Exhaustion

When your energy stops recovering fast enough for the life you're living.

“I'm no longer solving problems creatively. I'm just trying to survive this week.”

Emotional distance

When you start feeling further and further away from things you used to care about.

“It's working. That's enough.”

Reduced efficacy

When you stop feeling effective, even when you're working all the time.

“I used to be better at this.”

Identity gap

The distance between the professional you wanted to become and the person you find yourself being at work.

Sometimes the hardest part isn't exhaustion. It's not recognizing yourself in how you work anymore.

“I imagined engineering as building elegant systems. Now some days it feels like juggling Slack and reviewing AI output.”

2Step 2: Locate

Demands and resources model

At its core, the Job Demands-Resources model is simple. Every job has demands: things that require sustained effort, like workload, time pressure, constant re-prioritization or context switching. And resources: things that help you perform, recover and grow, like clarity, autonomy, support or time to recover.

Burnout grows when demands rise faster than people can adapt and resources don't keep up. So the question to ask is not “what's wrong with this person”, but “is the way we work producing burnout?”

Resources don't cancel out demands. They change how costly those demands are. Up to a point.

Same workload, same people, different resources: different performance, and a different energy bill.

What AI changes

The system doesn't get simpler. It gets faster.

In theory, AI should make things easier. In practice, efficiency gains quickly turn into expectation gains. The baseline of what's “normal” moves up, and the work around the work grows: reviewing generated output, switching between tools and models, keeping up.

Demands grow and resources stay even, at best. People are burning out even faster.

3Step 3: Apply

Apply two levers: Reduce demands. Build resources

If you want to change the system, you need to reduce demands and build resources.

Reduce demands

For example:

  • If people spend half their week in status meetings, don't teach them time management. Remove the meetings.
  • If everything is a priority, you don't need better planning tools. You need fewer priorities.
  • If your roadmap assumes everything will go perfectly, the problem is not execution. It's the plan.

Build resources

Resources have to match the actual friction in the system. For example:

  • Push decisions to the team level. Agree which decisions don't need approval, and stick to it.
  • Block 2-3 deep-work windows per week. No meetings, no Slack, non-negotiable.
  • Give teams designated time to explore and test new tools, and treat it as part of the work, not extra.

Make the work more clear, easier to decide, and possible to finish.

Where to start

Start with your own system

Go through the three steps with the worksheet. The first run takes no more than 15 minutes. Then do the same with your team.

Download the worksheet
Why this matters now
46%

of engineering leaders and developers report moderate or critical burnout

LeadDev Engineering Leadership Report 2025, n=617
45%

say they work more hours than a year ago, up from 35% in 2024

LeadDev Engineering Leadership Report 2026, n=600
#1

thing developers say gets in the way at work: unclear priorities and requirements

Stack Overflow Developer Survey 2026, n=30,903
Go deeper

Listen and watch

More on the ideas from Four Dimensions of Burnout, from different angles.

Refactoring with Luca Rossi, August 2026

Burnout is a System Issue

An hour on the difference between being tired and being exhausted, the two levers a leader actually has, and how AI quietly raises the baseline of what gets expected.

Also on refactoring.fm
LDX3 London, June 2025

Don't get toasted: a leader's guide to mental well-being

The prequel, about the one system you fully control: yourself. Energy, curiosity, priorities, and small practices that survive a bad week.

Talk materials
All talks and writing

Want to work on this with your team?

I run this framework as a workshop for engineering teams and leaders, and work with CTOs and managers on the changes that follow. Online or in person.

Explore the workshops
Book a call

Stay in touch

I share thoughts on leadership, team dynamics, and the real challenges of managing people - the kind of insights I wish I'd had when I was figuring it all out. Think of it as coffee conversations about what actually works (and what doesn't) in leadership.

Unsubscribe anytime. I respect your inbox.