A software engineer said he's "100% concerned, 24/7" about losing his job to AI.

A delivery driver is using Claude to build an e-commerce business on the side.

A lawyer says the AI diverges from her instructions every single time and she's not even sure it's helping.

81,000 people. One survey. Anthropic asking its own users what AI is actually doing to their work.

The obvious finding is in the first chart. People whose jobs Claude is doing a lot of? They worry about losing them. Software developers worry. Web developers worry. Elementary school teachers don't. Clergy don't. Fine.

That's the finding everyone will quote.

The weird finding is buried a few pages later. And honestly, it's the one that’s way more interesting.

They asked everyone how much faster AI makes them.

People said things like "I used to take months to make the website I [made] in 4-5 days." Or "I'm a non-tech guy but now I'm a full stack developer."

The researchers took those answers and put them on a 1-7 scale. 1 is much slower. 4 is no change. 7 is much faster.

Then they plotted that speed against how scared each person was of losing their job.

You'd expect a clean downward line. Slower equals more scared. Faster equals more chill.

That's not what happened.

The people who said AI slows them down are scared. Fine. They're mostly creative workers.. writers and artists.. who find AI rigid and also watch it flood their industry anyway.

But the people who said AI makes them much faster? They're the most scared of anyone in the survey. The "Much faster" bar is the tallest bar on the whole chart.

Just process that for a second.

The workers getting the biggest productivity boost are the ones losing sleep. The guy doing two hours of work in fifteen minutes is the one doing the quiet math at 2am.

And the math isn't irrational. It's the most rational thing in the entire dataset.

If AI barely helps you, your job feels safe. The tech isn't good enough to take it.

If AI makes you 5x faster, the tech IS that good. And you're sitting there figuring out what happens to the other four people who used to do what you now do alone.

The researchers said it more politely. Something about "if the time required to do your tasks is shrinking quickly, there may be more uncertainty about the future viability of the role."

Same sentence. Just dressed up.

There's a quote in the paper from a software dev that sums up the whole thing. He said "when AI arrived, the project managers started giving harder and harder tickets and bugs to solve." That's the U-shape in one line. AI made him faster. His workload expanded to match. The gain didn't go to him.

It went to the company.

Your browser gets you to a page. Norton Neo gets you to the answer. The first safe AI-native browser built by Norton moves with you from idea to action without slowing you down. Magic Box understands your intent before you finish typing. AI that works inside your flow, not beside it. No prompting. No copy-pasting. No switching apps.

Built-in AI, instantly and for free. Privacy handled by Norton. Built-in VPN and ad blocking protect you by default. No configuration. No extra apps. Nothing to think about.

Fast. Safe. Intelligent. That's Neo.

Download Norton Neo

There's another finding I want to flag.

Early-career workers are about twice as scared as senior workers. Around 8% of juniors said their job is at risk. Around 4% of seniors said the same.

And the gap gets worse when you look at who's getting the upside.

80% of senior professionals said the productivity gains were flowing to them personally. Only 60% of early-career workers said the same.

So the people most exposed to displacement are also the least likely to be capturing the gains.

This lines up with something Anthropic mentioned in earlier research. Tentative signs of a hiring slowdown for recent graduates in the US. If you're a junior engineer right now, your manager just got handed a tool that does junior engineer work pretty well.

That math is not fun.

And it complicates the standard advice, right? The standard advice when a new tech shows up is learn it, ride it, be the one using AI and not the one being used by it.

The 81,000 voices sort of broke that advice too. Because the people who DID learn it, the ones using it most aggressively.. they're at the tall end of the U. They're more exposed. Not less.

Not great news for the "just learn AI bro" Twitter crowd.

If you've been reading me for a while, you know I don't do the doomer thing. I think AI is generational and most people will end up better off because of it.

But reading this report.. I think the AI discourse has a specific problem.

We keep treating fear of displacement as the position of people who haven't tried the tools. The older cousin who's scared of ChatGPT. The uncle who calls it "the algorithm."

The data just said the opposite.

The people most worried about AI are the people using AI the most. That's someone who paid for Claude, loved it, and then felt their stomach drop a little.

One in five respondents voiced concern about displacement. That's a fifth of the most engaged AI users on the planet. They're not scared of what they don't understand. They're scared of what they understand very well.

I keep thinking about that software engineer. 100% concerned. 24/7. He's not some guy on Twitter screaming about AI doom. He's a working engineer who pays for Claude, uses it every day, and lives with it.

And honestly? I don't have a clean answer for what to do with that information.

But I think the first step is being honest that the fear isn't coming from the outside anymore.

It's coming from inside the speedup.

What about you? Are you at the tall end of the U? Faster than ever and quietly more nervous than you'd admit out loud?

Let me know. I read everything.

If you made it this far, you're not a casual reader. You actually think about this stuff.

So here's my ask. If this article made you think, even a little, share it with one person. Just one. Someone who's in the AI space. Someone who reads. Someone who would actually sit with these ideas instead of scrolling past them.

That's how this newsletter grows. Not through ads or algorithms. Through you sending it to someone and saying "read this."

Anthropic Report: What 81,000 people told us about the economics of AI