Saturday morning at the grocery store, which is the worst possible time to go.

I'm at the self-checkout, and my apples won't scan. I punch in the code from memory. The attendant walks over, sees me do it, and says: soon you'll be taking my job. Soon nobody will need me here.

I told her I doubted it.

She came back three more times in the next ten minutes. Once for my bag of grapes, which had a code on it that somehow wasn't the code in their database. Once because the bagging area disagreed with my bag. And once for a reason neither of us ever figured out.

This issue is my take on where all of this is going, built from what I've been reading about AI and the future of work. Nobody knows how this ends. Everyone is talking anyway, so here's my two cents.

The loudest voices on this question do not agree with each other, and that alone tells you something.

Elon Musk - Nobody works. Everybody gets paid.

Musk goes the furthest. He says that the best-case scenario is that probably none of us will have a job. Not fewer jobs, no jobs. Musk believes we would live off of what he calls universal high income. Basic income is a modest cheque covering essentials. Musk is describing something much larger, funded by robots making everything so cheaply that scarcity mostly ends. He has never explained who pays for it or how. He also has a habit of missing his own deadlines. His humanoid robot was supposed to be in factories by the thousands three years ago, and the real number has been in the hundreds.

Bill Gates - Expertise gets free.The week gets shorter.

Gates is warmer and more grounded. His idea is free intelligence: a great doctor, a great teacher, becoming as available as a search result within about a decade. He thinks that solves shortages we've never been able to fix. He keeps floating a two or three day work week, and he names a few jobs he thinks stay human for a long time, including coders and biologists. He also calls all of it just predictions, because at this point that’s all it can be.

Fei-Fei Li - Work changes. Your job doesn’t vanish.

You've probably never heard her name. She built the dataset that taught machines how to see, she is Co-founder and CEO of World Labs, and the field calls her the godmother of AI. She calls herself a pragmatic optimist, and she says she doesn't believe either extreme. Yes, AI changes work. No, your job is not simply disappearing. The real story is the messy middle, and she adds the part the other two leave out: more productivity does not automatically mean more prosperity for the people doing the work.

The self-checkout didn't delete the cashier. It took the easy part of her job, the scanning, and handed her a new one: fixing everything the machine gets wrong. She isn't out of work. She's doing harder work now, for the same pay, while the machine gets credit for the simple part.

That's the pattern. Tasks get automated. Jobs get rearranged. And someone is still standing next to the machine, handling everything it can't.

Forget the predictions for a second. Here's how to get a real answer about your own work.

I pulled a job posting off Girlboss Jobs to test this, and it sounds like a fun one: Director of Global Instructor Development at Peloton. It's the person who finds and coaches the instructors you follow on Peloton! The posting never mentions AI once in the actual work.

So I dissected it. Most of the list is spotting talent in an audition room, coaching someone through how they show up on camera, and getting a room of executives to agree on a direction. None of that is going anywhere. But a role like this also runs on evaluation rubrics, summaries, and decks presenting candidates to leadership.

Then, at the very bottom of the posting, Peloton says it uses AI internally to move faster, and that candidates are not allowed to use AI tools anywhere in the application process. They want to hear you, not a machine.

That tension is the whole story of this issue, sitting in one job ad.

So if you're wondering which parts of your work are exposed, paste this prompt along with your own job description, or one for a role you want next. Better to know now than to be blindsided later.

Do it with your own description first. Then do it with a job you'd love to have in three years. The gap between those two lists is your actual to-do list.

I wrote up the full walkthrough with screenshots, plus what to do when the answer stings a little.

🧠 The godmother of AI says there will be two kinds of workers. Fei-Fei Li sat down for a long interview about what's actually coming, and the title alone is worth the click. She isn't selling doom or utopia. Her whole argument is that this technology carries whatever values we put into it, which means the people who show up and use it get a say in what it becomes. If you only watch one thing this month, make it this..→ Watch it here

🇨🇦 🇺🇸 The jobs apocalypse didn't show up, but Canada is having a different month. South of the border the numbers came in strong. 162,000 jobs added in August, unemployment at 4.1%, and young workers, the ones everyone said would be hit first, are holding up. The Economist went further and estimated AI has created around a million American jobs so far, against roughly 200,000 layoffs you can pin on it.→ Read it here

Canada had the opposite August. We lost 42,000 jobs, unemployment held at 6.4%, youth unemployment sits at 12.9%, better than last year but still well above where it was before the pandemic. → Read it here

Almost none of that is AI. Economists are pointing at tariffs and a slowing trade picture. So if the job market feels harder here than the American headlines suggest, you're not imagining it. It's just that the thing making it harder right now is not the robots.

📱 Workers are ahead of their own companies Microsoft surveyed 20,000 people who use AI at work. Most said they're producing work they couldn't have produced a year ago. Then the uncomfortable part: only 13% said their company actually rewards them for reinventing how the work gets done. A lot of people have real skill and nowhere to put it. If you've ever figured out something clever with AI and had nobody at work care, that's the number you were feeling. → Read it here

So you run the exercise and the answer isn't the one you wanted. A lot of what you do all day turns out to be the kind of thing a machine can already draft, sort, or summarize.

First: that's most people. Including me, on plenty of days.

Second, and this is the part I believe: finding out early is the entire advantage. If you know today that parts of your job are exposed, you still have every option open.

Pick the one task you'd hand off tomorrow if you could, and hand it off this week. Then keep the time you win and put it somewhere that compounds. The relationships. The projects nobody else wants to own. The thing you're the only person in the room who can do.

Be the one who brings this to work instead of the one it happens to. The person who says "I tried this and here's what it does and here's where it falls apart" is the most useful person on any team right now. That role is wide open almost everywhere.

And say something. Fei-Fei Li's whole point is that this technology carries human values because humans put them there. Which means the terms are still being written, and quiet people don't get written into them.

The future of your work is not being decided in a keynote. It's being decided by what you're willing to try this week.

Now hit reply and tell me one thing: what part of your job would you never hand to a machine?

Cinz.AI

Your unfair advantage, one week at a time.

Got a question, a topic request, or just want to say hi? Just reply to this email. I read everyone.