Developer friction is rising while leaders double down on AI and metrics. The only way to fix what’s actually broken is to ask the people doing the work.
09.12.2025 12:00
👍 0
🔁 0
💬 0
📌 0
Developer friction is rising while leaders double down on AI and metrics. The only way to fix what’s actually broken is to ask the people doing the work.
Success isn’t about pushing developers harder, it’s about shaping the system around them. When doing the right thing feels effortless, excellence becomes inevitable.
Why developers are slower with AI tools despite believing they're faster and what actually works
The three-layer system, exact scripts, and implementation playbook for protecting your focus when everyone's convinced their request is the exception.
Max Kanat-Alexander breaks down why cycle time, focus, and cognitive load determine whether your engineering team thrives or burns out
Every team thinks one more feature will make the launch better. In reality, it’s the fastest way to delay, over-engineer, and burn out before you ever ship.
Quick sale on annual subscriptions – 50% off.
You'll get a full year of weekly posts about the messy reality of developer experience: the organizational dysfunction, the culture issues, the stuff that actually matters beyond just tooling.
blog.pragmaticdx.com/e884d54f
Lock it in while you can!
Every team thinks one more feature will make the launch better. In reality, it’s the fastest way to delay, over-engineer, and burn out before you ever ship.
Success isn’t about pushing developers harder, it’s about shaping the system around them. When doing the right thing feels effortless, excellence becomes inevitable.
Why developers are slower with AI tools despite believing they're faster and what actually works
The three-layer system, exact scripts, and implementation playbook for protecting your focus when everyone's convinced their request is the exception.
Max Kanat-Alexander breaks down why cycle time, focus, and cognitive load determine whether your engineering team thrives or burns out
Success isn’t about pushing developers harder, it’s about shaping the system around them. When doing the right thing feels effortless, excellence becomes inevitable.
Every team thinks one more feature will make the launch better. In reality, it’s the fastest way to delay, over-engineer, and burn out before you ever ship.
Max Kanat-Alexander breaks down why cycle time, focus, and cognitive load determine whether your engineering team thrives or burns out
The three-layer system, exact scripts, and implementation playbook for protecting your focus when everyone's convinced their request is the exception.
What's the most ridiculous workaround you've built because fixing the actual problem would require approvals from 6 different teams?
What's your team's unspoken rule for when documentation is 'good enough' vs when it needs to be perfect? Or does everything just stay in draft mode forever?
What's the programming language you love that nobody else seems to care about anymore?
How do you explain to non-technical stakeholders that 'just add a button' is actually a 2-week project touching 8 different services?
Your junior developer asks why we don't just rewrite everything in [hot new framework]. What's your response that doesn't crush their enthusiasm but also acknowledges reality?
When you discover a critical security vulnerability in production, what's your incident response? Fix first, tell people later, or panic immediately?
How many layers of abstraction does your codebase have? At what point does 'clean architecture' become 'I need a map to find where anything actually happens'?
Your PR has been approved but you spot a bug right before merging. Do you fix it in a new commit, start over, or just merge and fix it later?
When documentation says 'deprecated' but provides no alternative or migration path, is that helpful or just mean?
What's your strategy when you're 90% done with a feature and realize your entire approach was wrong? Start over or hack it together and refactor later (aka never)?
How do you keep learning new technologies when you barely have time to master what you're already using?
You inherit a 5000-line function that nobody understands but somehow runs the entire business. Do you refactor it or treat it like an archaeological artifact that must never be disturbed?
When you finally solve a problem that's been haunting you for days, do you feel accomplished or just annoyed it took that long?
When your test suite takes 2 hours to run, do you actually wait for it or just push and pray? What's your acceptable CI/CD wait time before you start cutting corners?