Your job is being automated.
Your company knows.
HR has the script ready.
I write about what's actually happening before it happens to you.
Weekly audit ↓
I was on a podcast last week.
The thing I kept coming back to: the entry-level collapse isn't the story. The story is what happens to mid-level in 5 years when there's nobody below them who knows how to do anything.
The pipeline is already broken. The talent shortage is just on a delay.
https://t.co/rGGprZGkTv
This is why I keep pushing back on the idea that displacement concerns are purely theoretical.
You can debate whether AI is the sole cause, whether companies are using it as cover, or whether the numbers are overstated.
But it's clearly become part of how executives explain workforce reductions.
It is incredibly exhausting to look at your calendar, see eight hours of meetings, and realize none of them produced a physical artifact.
A lot of mid level tech workers are realizing their entire career was built on optimizing a corporate machine that no longer needs them to turn the gears.
That realization hurts. It feels like an identity crisis because it is one.
You are not crazy for noticing that the vibe changed completely.
@atmoio The interesting thing is that both sides could be right.
The AI bubble can absolutely pop while AI still permanently changes the labor market.
The dot-com bubble popped too. The internet didn't disappear.
The absolute worst career move right now is trying to become more technical if your core job is strategy.
If you are a product manager, do not go spend three weeks learning python.
The junior engineers who write python all day are the ones currently exposed.
Your leverage is context. You need to become the person who translates messy operational reality into tight specs.
The machine handles the code. You handle the liability.
Just read a thread about how "the best engineers love vibe coding because it removes the boring parts."
The boring parts are often where the judgment lives.
Debugging a weird edge case at 2am is how you learn what the system actually does vs what the docs say it does.
@Kalshi It's becoming hard to reconcile the messaging.
On one side: "AI won't take jobs."
On the other: companies openly discussing smaller teams, hiring freezes, productivity gains, and doing more with less.
People notice the contradiction.
@unusual_whales For decades the formula was simple:
Get a degree. Get a better job.
Now more people are following the formula at exactly the moment technology is starting to compress parts of the white-collar labor market.
The next round of cuts won't be announced as AI cuts.
They'll be announced as "efficiency improvements" in Q3 planning.
The engineers who spent 2026 learning every new tool will look identical on a spreadsheet to the ones who didn't. The metric that saves jobs isn't tool adoption. It's outcome ownership.
The phrase that catches my eye is "every home and every desk."
That's not a niche product vision.
That's a vision where AI becomes as expected as electricity, internet, or a PC.
If that happens, the productivity gains stop being exceptional.
They become the baseline everyone is measured against.
The labor angle is hard to ignore.
If every new PC ships with increasingly capable local agents, the productivity gains don't stay confined to tech companies.
The pressure spreads everywhere.
The question stops being "Do you use AI?"
The question becomes "Why does your role still need the same number of people now that everyone has it?"
@StockSavvyShay We've spent years talking about AI as a cloud service.
Now the industry is trying to make AI a default layer of computing itself.
Once it's built into the device, adoption stops being an active choice and starts becoming part of the environment.
Also worth remembering that "more software engineers are being hired" doesn't automatically solve the labor market problem.
If a displaced accountant, analyst, recruiter, writer, or customer service worker can't realistically transition into those roles, the aggregate numbers don't tell the whole story.
The bottleneck was never just jobs.
It's pathways.
@chamath I think part of the backlash comes from the order of operations.
We're told AI will transform civilization, yet most people encounter it through email drafting, meeting summaries, customer support bots, and corporate headcount reductions.
That's a much less inspiring story.
Three things that are harder to automate than most people think.
Knowing which system will break when you change this one.
Knowing when the estimate is wrong before the sprint starts.
Knowing when to escalate and when to absorb it.
None of these are on the tools list.
Monday again.
The engineers running the math at home this morning aren't the ones who are struggling. They're the ones who built their careers on being technically excellent, and now technically excellent keeps getting redefined every six months.
That's a specific kind of exhausting.