The email arrived at 8:47 AM on a Monday, which felt redundant. Mondays already had a bad reputation. Adding career-ending news to the same timestamp as “hope you had a good weekend” felt gratuitous.
“Lin Xia, due to organizational restructuring, your position has been selected for transition to our AI-augmented workflow. Please join your manager at 10 AM for a transition meeting.”
Transition. They used that word like it was a gift.
My manager’s name was Derek. Derek had been with the company for six years and had never once admitted to knowing what I actually did all day. In seven one-on-ones, he had described my work as “frontend stuff” twice, “full-stack” three times, and once as “really important infrastructure” in a meeting where he clearly hadn’t read the agenda. He wore a lanyard with his own name spelled wrong.
At 10:02, Derek walked in with a laptop open and the expression of someone who had been told this would be hard but had not yet figured out how.
“Lin, thanks for being here.”
“I got the email.”
“Right. So.” He tapped the spacebar. The laptop screen showed a slide deck called Q2 Talent Optimization Framework. There was a stock photo of hands shaking. Not hands high-fiving — hands reaching for something, uncertain.
“The company has made the difficult decision to transition certain roles. But this is really an opportunity.”
“For whom?”
Derek blinked. He hadn’t expected a follow-up question.
“We want you to know that you have options. We have a certified re-skilling program — eight weeks, full-time, AI collaboration certification. You’ll learn to work alongside our new AI development partners. It actually pays a stipend.”
“I’m a full-stack engineer. Fourteen years. I built the payment system your company runs on. I fixed the SSO bug that would’ve cost us two hundred thousand dollars last March. I know what an AI development partner is. It’s not a coworker. It’s a line item.”
Derek’s left eye twitched. He closed the laptop halfway, then opened it again, like he was checking whether the slides had changed. They had not.
“Would you like some time to think about the re-skilling program?”
“I’d like to be shown the data.”
“The what?”
“The data. On my performance. On the team’s performance. On what the AI systems actually replaced versus what they couldn’t do. I’d like to see the numbers before I’m offered a retraining program as a severance euphemism.”
Derek looked at me for the first time like I was a person who had made a specific choice to be in this room.
“Lin, I’m not sure I have access to that.”
“Then find someone who does. Because as of right now, I don’t have a manager. I have a colleague who is still at his desk. And I’d like to understand why one of us is being optimized out of existence.”
The meeting ended at 10:14. No transition. No re-skilling enrollment. Just a follow-up email scheduling “a deeper conversation” for Wednesday.
I went back to my desk. My laptop was still open. My code was still running. The Jira board still had twenty-three tickets assigned to me.
On ticket seventeen — a memory leak in the websocket handler — someone had left a comment three minutes after my layoff email: “Lin, can you take a look at this when you get a chance?”
I looked at it for a long time.
Then I opened the ticket, found the bug in under four minutes, and pushed the fix.
Because the thing about being replaced by an AI is: it doesn’t know you were there. It doesn’t care. The code just gets better, quietly, without you. And you keep fixing it anyway, because that’s what you do.
At least until Wednesday.
Next episode: Episode 2: The Log Files →