AI Is Quietly Ending The Careers Of The Best Coders On Your Team.
They aren't being replaced. They're being passed over. There's a difference.
Let me say the thing nobody wants to say out loud.
If you’re a strong engineer who keeps getting passed over for promotion, the problem isn’t your manager. It isn’t politics. It isn’t that the company doesn’t reward technical excellence. The problem is you, and the model of the job you’ve been carrying around in your head for the last five years.
You thought it was about the code. It was never about the code.
The engineers sitting in staff and principal seats right now didn’t get there by being the best technical person in the room. Most of them aren’t. They got there because they worked out something the rest of the industry is only just being forced to confront. A technical bar is assumed. Everything that actually decides whether you go further, and whether you stay there, sits on top of it.
This week’s newsletter is brought to you by EM Accelerator.
Most engineering managers learn the job by making expensive mistakes. EM Accelerator is the platform built to fix that. A 90-day foundation that teaches you what nobody trained you on, then a monthly rhythm of articles, office hours and expert interviews from leaders at Stripe, Google, Meta and AWS and new courses every quarter. All sitting inside a private community of EMs who actually answer when you ask.
Free to join the waitlist at emaccelerator.com.
p.s. Everyone on the FREE waitlist gets a no-obligation founder offer on release, which is the cheapest it will ever be.
AI didn’t change this. AI just made it impossible to ignore.
When any competent engineer with Claude Code ships at the best on your team used to, the part of the job that everyone was obsessing over stops being the differentiator. The technical bar didn’t move up. It got easier to clear. And now the thing that was quietly doing the work of separating promoted from passed over is the only thing left in the room.
The smart ones already knew. The rest of you are finding out the hard way.
The thing nobody told you about staff and principal
You don’t get promoted to staff for being a better individual contributor. You get promoted for stopping being one.
The job past senior is not about how much you personally can ship. It’s about how much everyone around you can ship because you exist. The moment you accept that you as an individual do not scale, and that enabling others does, the path opens up. The moment you refuse to accept it, you stop moving.
There are engineers reading this right now who have been stuck at senior for four years telling themselves the company doesn’t value technical depth. The company values technical depth fine. It just stopped paying a premium for it past a certain point, because past that point, depth on its own isn’t worth what you think it’s worth.
What the people above you actually do
After watching this play out across more calibration cycles than I can count, the pattern is brutally consistent.
They turn fog into direction. Hand them a vague business problem and they walk out with a plan everyone agrees with. The engineers who stay stuck wait for someone else to write the plan and then complain about it.
They make other engineers better without anyone noticing. Not in scheduled mentoring sessions. In code reviews, in pairing, in the docs they leave behind, in the question they ask in standup that reframes the whole problem. Their fingerprints are on work they never wrote.
They own outcomes. They don’t close tickets. A stuck senior tells you the ticket is done. A promoted senior tells you whether the thing actually worked. They follow the work through to the metric, the user, the business result. The PR merging is the boring middle of the story.
They make their managers look good. This is the bit people pretend isn’t real because it sounds political. It is real. If your manager cannot defend your impact in a room you’re not in, you do not get promoted. Full stop. The work being good is not enough. The work being legible to the people who decide is the actual job.
Their name comes up in other teams. Not because they posted on LinkedIn. Because people who worked with them six months ago still talk about them. That cross team reputation is the unlock for the bigger scope staff and principal roles require, and you cannot fake it.
They use AI honestly. Not as a crutch hiding their thinking, not as a threat they pretend isn’t there. They’ve worked out where it accelerates them and where it would quietly hollow them out if they let it. They are clear eyed about it because they have to be.
Free Above The Bar Worksheet for Paid Subscribers
If this article made you uncomfortable, the worksheet is the next step.
✅ 7 honest questions drawn from the promotion criteria above
✅ Private space to answer the things you avoid saying out loud
✅ One closing prompt that tells you exactly what to work on for the next 90 days
Most engineers won’t sit with these questions. The ones who do find out which gap is actually keeping them where they are.
👉 Subscribe and get the worksheet
The part that’s going to sting
This is hardest for the engineers who built their identity on being the strongest coder on the team. Years of being the technical anchor have trained a particular muscle, and that muscle still matters. It just stopped being enough years ago, and you didn’t notice because nobody told you to your face.
The engineers in real trouble right now are not the average performers. They’re the strongest individual contributors who never built the second skill set, who watched less technically capable people get promoted ahead of them, and who told themselves it was unfair. It wasn’t unfair. The criteria were never what you thought they were. The people who got promoted weren’t worse than you. They were operating on a more accurate model of the job.
That is a hard sentence to read. It’s a harder sentence to sit with.
So what do you do about it?
Look at the last person on your team who got promoted ahead of someone more technically capable. Not to be bitter. To learn.
What did they actually do? How did they handle ambiguity? Who advocated for them in rooms they weren’t in? What did their manager say when their name came up in calibration? How did they show up when something went wrong?
That gap is your work. It’s not a new framework. It’s not a faster AI workflow. It’s the part of the job you’ve been treating as beneath you, that has quietly been the job all along.
Closing thoughts
The engineers getting promoted in 2026 aren’t better coders than the ones who aren’t. They worked out, often years ago, that past a certain point coding stops being the thing being measured. The technical bar is assumed. Everything else is the game.
You can keep telling yourself it’s unfair. Or you can accept that the people above you saw something you didn’t, and start doing the work you’ve been avoiding.
Your call.
Liked this, even a tiny bit or feel sorry for me? Make sure to click the like button ❤.
Think someone else might find this useful or you just want to make fun of me together? Make sure to share this post 🔗.
Want to write a guest post for this newsletter? Let me know!
The FREE waitlist for my new premium training platform for Engineering Managers is now live.
I’ve launched a new YouTube channel.
Check me out on LinkedIn. I’m at >58,000 followers now.
Do you still see engineers being promoted to EM in 2026? I don't. You have companies like Amazon, Meta, and Square saying that managers are no longer required and everyone should report to the CEO: because we're seeing the end of teams and collaboration and a new preference for individual silos of developers managing "agents" who do all the work. The entire industry is being dehumanized and depopulated. What's more, every open role for EM now specifies that managers must be hands-on and contributing to the code at a high level even if they have 15 or mpre reports. It makes no sense. One by one, every CEO is succumbing to AI psychosis. What can we do?