TL;DR: Picture the colleague who learned to work with AI doing your work at three times the speed for the same pay. That threat is more uncomfortable than a machine taking your job, and the doomer framing and the hyper framing both let you off the hook for the one action that matters.
Here’s the line that gets repeated everywhere right now, and gets the threat wrong: “AI is coming for your job.”
It isn’t. AI is software. It doesn’t apply for your role. It doesn’t get hired. It doesn’t show up to work in your place on Monday morning. AI tools will absorb some boring tasks in your current job, but that’s not a job replacement. That’s a job evolving, the way every job has always evolved when new tools arrive.
The accurate version of the threat is more uncomfortable, so nobody repeats it on TV. AI isn’t coming for your job. Other people using AI are.
I think this is the most useful thing anybody can tell you about AI right now, and it’s the least popular.
After 33 years in enterprise IT, I don’t believe a piece of software ever took anybody’s job on its own. Somebody chose to use it, and somebody else chose to wait. Five kinds of competitor show how that plays out, and even the one that looks like a machine has a person behind it.
1. The competitor down the hall
You and a colleague have the same job. You’ve each been doing it for ten years. You’re equally good at it. The difference is that she spent the last six months learning to use AI tools for the parts of the job that drained her time, and you didn’t.
The doomer framing lets you off the hook for the only action that matters.Share on X
She now produces three times as much work in the same hours. Her quality is at least as high as yours, and probably higher, because she has more time for the parts of the job that require her judgment. The boring parts that used to consume her week now take an hour, and the freed time goes into the work that matters.
You’re producing the same amount you produced last year. You’re working the same hours. Your manager, who runs the team, is noticing that one of you is shipping at a pace the other one isn’t.
After 33 years in enterprise IT, I don’t believe a piece of software ever took anybody’s job on its own.
Six months from now, the team needs to cut one position. The decision is between you and your colleague. You’re going to lose, and AI didn’t fire you. Your colleague did, by being better at the job than you were. I don’t like how cold that sounds, and I’m not going to soften it for you. When a manager has to cut a seat, the person shipping less goes first, and nobody stops to ask whether the gap came from a tool.
Watch: AI Won’t Replace You: It’s Here to Liberate You | Darren Radford | TEDxSurrey, TEDx Talks
2. The competitor across town
Same job, different company. The company across town has just realized that one of their people with the same title as yours, using AI tools, can do the work of three people without them. They’ve laid off two of the three, kept the augmented one, and reduced their cost per unit of output by 60 percent.
Your company is now bidding for the same client your competitor across town is bidding for. The competitor can charge less because their cost is lower. They win the contract. Your company loses revenue, then loses headcount, then loses you, because the company that hired AI users put you out of business.
The chain of events that ended with your unemployment had a human at every step.
The human across town who chose to learn AI tools. The human who decided to lay off the two unaugmented colleagues. The human at the client who decided to give the contract to the cheaper bid. The human at your company who had to cut headcount because the revenue was gone. AI made each of these humans’ choices easier. AI made none of the choices.
That chain is why the doomer talk wears on me. Every link in it is somebody’s decision. Blaming the software lets every one of those people off the hook, and it hides the one decision in the chain that’s still yours: whether to spend a few hours learning the tool your competitor already uses.
3. The competitor who doesn’t exist yet
This one is harder to see because it’s not happening to you yet. It’s happening to industries.
Right now, in 2026, people are setting up new companies to do work your company has been doing for twenty years. They run a workforce a third the size, because they’re building the company around AI-augmented humans from day one. They don’t have your company’s institutional inertia, your company’s legacy systems, or your company’s headcount overhead. They have one augmented person doing the work of three.
You won’t see these competitors for two or three years. By the time you see them, they’ll already have the customers who were willing to switch for the cost savings.
Your company will start losing market share. Eventually it’ll respond by trying to do what they did. By that point, the augmented humans who could have done it from inside your company will have left for the new competitor, where they get paid more.
This pattern is older than AI. The Retraining Question walks through the historical version of it. The bank tellers thought their job was safe until banks restructured around ATMs and the people who learned the new systems became the bank managers. The travel agents thought their job was safe until Expedia restructured the industry around travel-self-service and the agents who learned the new tools became the niche specialists who still exist today.
4. The competitor who used to be you
This is the one most people don’t notice because it happens slowly.
Five years from now, the version of you who started learning AI tools this month is going to be a very different professional than the version of you who waited. The augmented future-you is going to be doing more interesting work, faster, with better tools, with deeper expertise in the parts of the job that AI can’t do. The unaugmented future-you is going to be doing the same work, in the same way, while everyone around them gets faster and sharper.
The augmented future-you is the same person who made a different choice this month. Five years from now, the two versions are competing for the same opportunities, and one of them is going to win.
This is what I tell every client who tells me they’re too old, too busy, or too far behind to learn AI. The choice you’re making this week is which version of yourself is going to exist in 2031. The other version isn’t going to exist, because you chose not to be that person. The 71-Year-Old Memoirist Who Uses AI Better Than You Do is the proof of concept for what the augmented version looks like at any age.
5. The competitor who isn’t a human at all
The doomer framing gets one thing partly right. There are jobs that will, in fact, be done by AI alone, with no human in the loop. They’re the minority of jobs that were already on the road to automation, the ones where the work is routine, bounded, low-stakes, and easy for a machine to do without human oversight.
AI is increasingly doing the data entry job that a junior assistant used to do. A chatbot is increasingly deflecting the basic customer service inquiry that an entry-level agent used to handle. An AI-powered tool is increasingly generating the simple legal document drafting that a paralegal used to do, with a senior attorney reviewing the output.
Those are tasks going to AI, and the job around each one reshapes itself.
In a well-run company, the junior assistant whose data entry got automated moves to more interesting work for the same boss, the service agent moves into complex cases, and the paralegal reviews generated drafts and carries more files. I wouldn’t count on every company running that well. Some will just cut the seat, and when they do, the person who learned the tool is the one who keeps it.
If you’re doing one of these jobs, the part of it that gets automated is going to free your time for the part that doesn’t. Your job changes.
Your employer expects you to do more interesting work, and pays you for it, if you’re augmented. If you aren’t, your employer hires somebody who is, and you’re out of work because of the augmented competitor. The pattern Augmented Beats Replaced walked through is the workforce version of the same math.
Why is the AI-takes-your-job framing so popular?
Because “AI is coming for your job” lets two groups of people off the hook.
It lets the doomers feel like they’ve identified the threat without having to do anything about it, because the threat is the technology itself, which they can complain about while waiting for somebody else to solve.
Reading another article about AI replacing humans is easier than spending an hour learning a new tool. The narrative is satisfying. The action it implies is to be afraid, and that’s what people do right before their colleagues outcompete them.
It also lets the hypers feel like they’re on the right side of history without doing anything specific. If AI is going to replace all human work, then anyone who’s enthusiastic about AI is on the winning team by default. They don’t have to learn the tools or apply them to their own work. They’re inevitabilists, riding the wave.
Both stories are bullshit that feels good. The doomers get to feel smart and the hypers get to feel ahead, and neither group has to open a tool. The people who pay for it are the ones who believed them, and some of those people are good at their jobs and one quiet decision away from falling behind.
Both framings produce inaction. Inaction is what your augmented colleague needs you to do, because every week you stay inactive is a week she gets further ahead.
The wrong framing of the threat is useful to everyone except you. The right framing is uncomfortable because it implies you have to do something this week, not next year, and not after the company runs the training.
My take
Fear the colleague who learned the tools, because she’s doing your work three times faster for the same pay. The doomer story and the hype story both let you off the hook, and both are bullshit that feels good. Neither one asks you to open a tool. Spend an hour this week learning one. In a well run company the augmented person gets more interesting work. In a badly run one the augmented person keeps the seat.
Who is really competing for your job?
From the podcast
Trevor Sumner, a twenty-five year veteran of the tech industry, puts the same point in a single line on Leaders and Their Stories:
AI isn’t going to replace us. People who use AI better than you will replace you.He expects the question of how you use it to become standard in job interviews within the year.
AI isn’t coming for your job. Other people using AI are.
This is what the threat looks like. The competitor down the hall, the competitor across town, the competitor who doesn’t exist yet, the competitor who used to be you. Five different people, all of them augmented humans, all of them about to outcompete the unaugmented version of you for the same finite set of professional opportunities.
The fix is to be one of them. Pick one tool this week. Spend an hour learning it. Spend two weeks using it. Notice the time you got back. Use the time to be better at the part of your job that only you can do. Repeat next month.
That’s how you become the augmented version of yourself, and the augmented version of yourself is the only version that’s going to exist in 2031. The longer version of what that looks like is in The Birth of the Augmented Human. The shorter version is the one above. Either way, the choice is the same, and the choice is this week.
Capable people are losing ground because a headline told them the threat was a machine they couldn’t fight. That’s a loss for everyone. Pick a tool, point it at the dullest hour of your week, and see what comes back. The person down the hall who spent that hour last month is the one who can take your job, and I’d much sooner that person be you.
Be the colleague who learned to work with AI. The Birth of the Augmented Human is my book on using it to raise your output and your decisions.
