Three companies. Yesterday. 46,000 jobs gone.
Amazon: 16,000
UPS: 30,000
Pinterest: 700
This isn't a layoff. It's a swap.
Look at the press releases:
Amazon: "Cutting roles to fund AI and data centers."
Pinterest: "Eliminating staff to hire AI-focused talent."
UPS: "Reducing workforce for automation."
They're not cutting and pocketing the savings.
They're trading you for something else.
Pinterest said it plainest: We're firing you to hire someone whose job is to make you unnecessary.
The email.
This is what it looks like when it happens to you.
2:47 AM. Subject line: "Important Information About Your Role at Amazon."
"Hi [name],
I have some important, but difficult, news to share with you. After a thorough review of our organization, our priorities, and what we need to focus on going forward, we've made the hard business decision to eliminate some roles across Amazon. Unfortunately, your role is being eliminated as part of these changes and your employment will end after a notification period.
These decisions weren't made lightly, and we recognize the significant impact they have..."
That's it. That's the email.
Years of work. Reviews. Promotions. Relationships. Stock vesting next month.
Reduced to "your role is being eliminated" at 2:47 in the morning.
Someone at Amazon posted this screenshot yesterday. By now it'll be one of 16,000.
The numbers no one's putting together.
I spent the week digging. Here's the picture:
Where the money's going:
$600 billion into AI infrastructure in 2026. Up 36% from last year.
Amazon: $125 billion capex. Same quarter they cut 30,000 people.
Microsoft: $80 billion into data centers. Same year they cut 15,000.
By 2030? $1.4 trillion. Annually.
Where it's coming from:
775,000+ tech workers laid off since 2022
2025 cuts are 54% higher than 2024
Entry-level tech postings down 67%. In one year.
Middle management postings down 40%+
The swap in real numbers:
AI job postings: up 134% from pre-pandemic
Everything else: down 34%
AI engineer salaries jumped $50K in one year. Average now $206K.
$600 billion going in. 775,000 jobs going out.
That's not a coincidence.
Entry-Level Tech Hiring (Index)
What it actually looks like.
The numbers are abstract until you read what people are posting.
This week. Real people. Real posts.
Someone posted on Reddit today: "I got laid off at Amazon Jan 2026. Not sure what to do next."
That's it. That's the whole post. Years of career distilled into "not sure what to do next."
An Amazon employee — 8.5 years at the company — got pushed into the layoff pipeline despite being called a "Super Hero" by their manager last quarter. Eight and a half years. Didn't matter.
Another one planned to job search in January. Then a family crisis hit. Then layoffs hit. Now they're "freaking out" because the timing is the worst possible.
Someone on Blind mentioned sending out 300 résumés. Zero responses. "I know I'm qualified. I'm losing hope."
A Citi employee in NYC — first layoff ever — wrote: "I'm not okay." Described chest-tightening anxiety, cycling through disbelief and sadness, feeling guilty for not being more "grateful." Each day is something to just survive.
A software engineer lost their job right after winter break. 17% of the company gone. They're a single-income household with three kids. They regret the money they spent on Christmas.
A Meta employee who joined Reality Labs in October has already lived through multiple layoff rounds in three months. Working 9-9-6 and still doesn't feel safe. "There is no stability."
These aren't people who slacked off. These aren't people who didn't see it coming. Some did everything right. Had the reviews. Hit the metrics. Got the "Super Hero" label.
Didn't matter.
The math changed. They were on the wrong side of it.
I found out by accident.
Early January. I needed to reach my AWS account managers in Canada. Standard business stuff.
Emailed the first contact. Bounced.
Emailed the second one. Bounced.
Emailed the third one. Bounced.
Three contacts. Three bounces. No one told me anything.
The layoffs had happened in November. Two months earlier. And no one thought to send a "hey, your account team changed" email.
Just... silence. Bounced emails. Like they never existed.
I finally got through after trying five different channels. The response?
"Oh yes, [name] and [name] have transitioned out of the company."
Transitioned out.
That's the phrase now. Not "laid off." Not "let go." Transitioned out.
Like they chose to leave. Like it was peaceful. Like there wasn't a 2:47 AM email that ended it.
Project Dawn.
That's what Amazon internally called this round of cuts.
This week, an email about "Project Dawn" was accidentally sent to AWS employees who weren't supposed to see it.
The email implied that affected workers in the U.S., Canada, and Costa Rica had already been notified.
Thousands of people panicked. Refreshing inboxes. Checking Slack. Waiting to find out if today was their day.
Amazon walked it back as an "error."
But the panic was real. And for 16,000 people, it wasn't an error. It was just early.
The PIVOT trick.
Here's one that's making the rounds.
An engineer — less than a year at Amazon — got tagged as a "performance problem" for things like missing meeting notes once. Once.
They were handed a PIVOT plan. Two weeks to accomplish what their predecessor couldn't do in two years.
The math: If you get hit by a "real" layoff, you get three months severance. If you get PIVOTed out, you get two.
One month of pay. That's what the performance theater saves them.
Commenters are calling it what it is: Amazon gaming severance by disguising layoffs as performance issues.
The line between "you're not good enough" and "we don't have budget for you" disappeared. Now it's just "we're managing you out" — and they get to pretend it was your fault.
The stat that should scare you.
77% of new AI jobs require a master's degree.
Read that again.
You're a 45-year-old engineer. Amazon just let you go. Your options:
A) Spend 3 years and $100K+ on a master's to compete for AI roles B) Fight for shrinking non-AI jobs against everyone else who got cut C) Leave tech entirely
For most people, C wins.
"Just reskill" sounds nice on a podcast. But the reskilling path has a credential gate that locks out most of the people being displaced.
The 28-year-old with a CS degree can pivot. The 45-year-old with a mortgage and two kids cannot.
This isn't a skills gap. It's a life stage trap.
The quote that says everything.
Salesforce CEO Marc Benioff, September 2025:
"I reduced from 9,000 heads to about 5,000. AI handles 50% of tasks."
Not "we made difficult decisions."
Not "market conditions forced our hand."
"AI handles 50% of tasks."
He's not apologizing. He's flexing.
And here's the thing — Wall Street rewarded him for it. Stock went up. Analysts praised "efficiency gains."
That's the game now. The market pays you to replace humans. It punishes you for keeping them.
The part that's almost funny.
More than half of companies regret their AI layoffs.
Forrester found 50%+ end up rehiring. Often at higher cost.
Why? AI can't do institutional knowledge. It doesn't know that weird billing edge case exists because of a deal you made in 2019.
It doesn't know who to actually call when production breaks at 2am. It doesn't know that the CTO hates Slack messages before 9am.
Humans spent years learning this stuff. Companies fired them. Then discovered AI doesn't know any of it.
But here's the joke: they're not rehiring the people they fired. They're hiring externally. Often offshore. At higher rates.
So the regret doesn't help you. You're still gone. They just pay more to learn the same lesson.
The time bomb no one's talking about.
Entry-level hiring collapsed 67%.
Companies figured out: senior engineer + AI can do the junior's work. Why train anyone?
The math made sense on a spreadsheet somewhere.
Here's what that spreadsheet missed:
67% fewer juniors in 2024-2026 = 67% fewer senior engineers in 2031.
Today's junior devs are tomorrow's tech leads, architects, and CTOs. Cut the pipeline by two-thirds and you don't save money. You delete your future leadership.
One VP told me: "The leadership vacuum already arrived. We just haven't felt it yet."
Companies are optimizing for this quarter and destroying next decade. CFOs get bonuses. Future CEOs get nothing to work with.
Classic short-term thinking dressed up as innovation.
The two labor markets.
There used to be one tech job market. Now there are two.
Market A: AI roles
Postings up 134%
Salaries up $50K in one year
Multiple competing offers
Remote, flexible, companies fighting for you
Market B: Everything else
Postings down 34%
5-6 month job searches
200+ applications to get interviews
Companies ghosting you after final rounds
Same industry. Same people. Completely different realities.
If you're in Market A, this is the best time to be alive.
If you're in Market B, you're competing against everyone who just got pushed out of Market A.
The gap is getting wider every quarter.
What actually happened yesterday.
The split became visible.
Two sides now:
Side A: Employed.
Someone else decides if you work tomorrow
Your skills compete against AI + people using AI to replace you
100% exposed to decisions you don't make
One email can end it
Side B: Own something.
You decide how to adapt
AI is a tool, not a replacement
80% exposed, but you control the response
No one can send you that email
20% gap. That's everything.
It's the difference between getting the email and sending it.
The dread before the email.
Here's the part no one talks about.
It's not just the people who got cut. It's everyone waiting to find out if they're next.
TeamBlind threads this week are full of people counting down to January 28. Trading inside info. Guessing which orgs get hit. Refreshing Slack. Doom-scrolling.
One person at Meta said they're working 9-9-6 and still don't feel safe. Three months at the company, multiple layoff rounds already. They're questioning whether staying even makes sense given the chronic stress.
Amazon employees are sharing WARN notices — 2,303 people across 56 facilities in Washington alone. Separations running January through May.
This is what "job security" looks like now. Not security. Anticipatory trauma. Living in fear between announcements.
People are saying "2026 is picking up right where 2025 left off."
They're not wrong.
The reactions that won't save you.
Wrong move #1: "I'll learn AI tools."
Pinterest isn't hiring people who can prompt ChatGPT.
They're hiring people who can architect systems that delete entire departments.
"I use AI" vs "I replace people with AI" — that's the gap between keeping your job and losing it.
Knowing how to use a tool doesn't make you safe. Knowing how to use a tool to eliminate the need for other humans? That's what they're paying $206K for.
Wrong move #2: "I'll update my resume."
For where?
Amazon spent $125 billion on AI. Microsoft spent $80 billion. That's not their strategy. That's everyone's strategy.
Your resume is a bet that someone will choose to pay a human over a machine. That bet gets worse every quarter.
Every quarter there's more AI. Every quarter there are fewer jobs. Every quarter your resume competes against more people for less opportunity.
That's not a job search. That's a losing trade.
What survives.
I keep coming back to five things AI can't touch:
Identity — AI can't be you. Your specific story. Your scars. Your weird opinions. Your taste. Anonymous is vulnerable. Generic is dead. You is defensible.
Relationships — AI can't text your old coworker and get a real favor. Can't make an intro that actually gets answered. Can't build trust through years of showing up. Your network isn't a vanity metric. It's a moat.
Stakes — AI advises. It doesn't bet. No skin in the game. When you put your money, reputation, or future on the line, you're operating where AI can't follow.
Selection — AI analyzes everything. Can't decide what matters. Your ability to say no — to clients, to projects, to opportunities — is worth more than your ability to say yes.
Accountability — When it breaks, AI doesn't care. Doesn't show up at 2am to fix it. Doesn't have its name attached to the outcome. You do. That's not a burden. That's a feature. Charge for it.
Work in those layers? You don't get swapped.
Work outside them? You're in line. Just don't know your number yet.
What I did.
December 2025. I looked at my business and asked: "What would AI eat in 18 months?"
The answer was uncomfortable.
I killed:
My SaaS tools (anyone builds these now)
My intro content (AI teaches better than I do)
My playbooks (AI generates these instantly)
My consulting model (AI advises for free)
Everything that didn't require me specifically
Hurt to do it. Some of that stuff took years to build.
But I'd rather kill it myself than watch it die slowly.
What I kept:
Selection (I reject 70% of people who want to work with me — that filter IS the value)
Accountability (my name on it, my reputation at stake)
Relationships (took years to build, can't copy-paste)
Stakes (equity over fees, skin in the game)
I didn't "learn AI."
I moved to where AI can't follow.
The uncomfortable math.
Meta is printing billions in profit. Cutting thousands anyway.
Amazon had its best quarter ever. 30,000 people gone.
Microsoft stock is up. 15,000 jobs down.
Performance doesn't protect you anymore. Profit doesn't protect you. Tenure doesn't protect you. Being called a "Super Hero" by your manager doesn't protect you.
The only thing that protects you is being in a layer that can't be swapped.
Or owning the thing yourself.
The question.
Look at your job.
If your company did the Amazon math — "what can we spend on AI instead of humans?" — do you survive?
Not "is my company planning layoffs."
Every company is planning layoffs.
The question: Are you in a layer that gets swapped, or one that doesn't?
Are you building something you own, or renting your time to someone who's calculating your replacement cost right now?
Are you moving toward stakes, selection, relationships, accountability — or are you optimizing skills that depreciate faster every quarter?
46,000 people got their answer yesterday.
The guy at Amazon with 8.5 years and a "Super Hero" rating got his answer.
The engineer with three kids who just spent money on Christmas got his answer.
The person who posted "I got laid off at Amazon today — not sure what to do next" got their answer.
My three AWS contacts who "transitioned out" got their answer. I found out from bounced emails.
The Meta employee working 9-9-6 who still doesn't feel safe? She's waiting for hers.
What's yours?
George
P.S. — If you're reading this and feeling that pit in your stomach, good. That's the correct response to reality. The people who should be worried are the ones who aren't.