I talk to worried professionals every single week. Managers who have been in their industry for fifteen years and suddenly feel like the floor is moving. People who are good at their jobs, have built real careers, and are now genuinely unsure whether any of that matters anymore.
The headlines are not helping. Every few days there is another breathless article about AI wiping out entire job categories. Executives make nervous announcements. LinkedIn fills up with think pieces. And the people in the middle of their careers are left trying to figure out what is actually true.
So let me tell you what I know. The economy, at the macro level, is probably going to be fine. History is clear on that. But that does not mean your specific career will automatically be fine. Those are two very different things, and confusing them is costing people time they do not have.
01
Why the “AI Will End Work” Story Falls Apart
The fear that AI will eliminate work is not new. Economists have a name for the thinking behind it: the lump-of-labor fallacy. The idea is that there is only so much work available, and if a machine takes some of it, humans are left with less. It sounds logical. It just is not how economies actually work.
Human ambition does not sit still. When one layer of work gets automated, people find the next thing to do. John Maynard Keynes predicted in the 1930s that automation would have us all working fifteen-hour weeks by now. He was wrong because he underestimated what people do when they have more capacity: they want more things, build more things, and invent entirely new problems to solve.
The research backs this up. A 2026 Atlanta Fed study found that more than 90 percent of firms reported no headcount impact from AI. Yale’s Budget Lab, also in 2026, found no statistically significant relationship between AI adoption and unemployment. Multiple NBER and Census Bureau papers found that where AI is having an effect, job creation and job destruction are running nearly equal.
The historical pattern is consistent. Farm mechanization collapsed agricultural employment from about a third of the U.S. workforce to roughly two percent. But those workers did not stay unemployed. They moved into factories, offices, hospitals, and eventually into industries that did not exist before. Farm output tripled. The economy expanded.
When spreadsheets showed up, people predicted the end of bookkeeping. What actually happened: roughly a million bookkeeper roles shifted, and about 1.5 million financial analyst positions appeared in their place. When Expedia and Kayak decimated travel agency employment, the agents who remained saw their wages outpace the broader private sector by 2025. The industry shrank. The survivors did better.
The macro case for stability is solid. That is not the problem. The problem is that people hear “the labor market will be fine” and assume that applies to them personally. It does not automatically.
02
The Macro Story and Your Story Are Not the Same
There is a saying I use with clients: averages do not live your life. Saying the labor market will be fine in the aggregate is a bit like saying the average person is perfectly healthy. That tells you nothing about the specific person sitting across from you.
The same research showing neutral aggregate effects also confirms what a lot of mid-career professionals are already sensing. Entry-level roles with high AI exposure are getting harder to find. Routine administrative, clerical, and transcription work is contracting. The pay difference between roles where AI makes you more effective versus roles where AI can replace you is getting wider, not narrower.
Entry-level, high-AI-exposure roles are harder to land right now. Routine clerical and administrative work is contracting. The wage gap between augmented and substituted roles is widening steadily. None of this is apocalypse. All of it is worth paying attention to.
Structural changes like these show up in people’s lives before they show up in government data. The person who cannot get callbacks on applications, or who watches their department slowly consolidate, is experiencing the shift before any economist writes a paper about it.
If your role sits in the substitution zone, waiting for clarity is a strategy that works against you. The professionals who fare best in disrupted markets are the ones who move while they still have options, not after those options narrow.
03
The People Pulling Ahead Are Not Necessarily the Most Talented
Here is something that does not get said enough: the workers benefiting most from AI right now are not the ones with the most credentials or the most years of experience. They are the ones who decided to get comfortable with the tools.
Software engineering is a good example. For years the doomer narrative pointed at developers as the most obvious casualties. What is actually happening: demand for software engineers has been increasing since early 2025. Product manager job postings are at their highest level since 2022. AI did not replace those roles. It made the people in them more productive, which created more work, which created more demand for more people.
On corporate earnings calls, companies mention AI as an augmentation tool over a replacement tool at an eight-to-one ratio. That tells you something about where management attention is actually going versus where the headlines would have you look.
The workers pulling ahead are the ones who decided to engage. Not the youngest. Not the most credentialed. The ones who chose to learn. That is a decision available to anyone reading this right now.
Wages are growing faster in AI-augmented roles, particularly in systems design, data strategy, and technical leadership. The professionals in those roles are not necessarily smarter than you. They just got moving sooner.
04
What I Tell My Clients to Actually Do
I have spent years helping professionals navigate major career transitions. What follows is the honest, practical framework I use with clients, built around three principles I call the ATG method: Adapt, Transfer, Grow.
Get an Honest Read on Your Exposure
Not all risk looks the same. A customer service rep answering the same fifteen questions on rotation faces a very different situation than a senior operations manager who synthesizes judgment across multiple stakeholders and competing priorities. You need an accurate picture of where you actually sit, because most people get this wrong in both directions. Some overestimate their risk and panic unnecessarily. Others underestimate it and wait too long.
Take Stock of What You Already Bring
The skills that made you good at your job do not disappear when a role shrinks or disappears. Regulatory knowledge, stakeholder relationships, domain expertise, the ability to read a room and make a judgment call under pressure. None of that gets automated. It gets repositioned. The question is whether you know how to present what you have in a way that speaks to roles that are actually growing, and that takes some deliberate work.
Move Before You Have to
The clients I see struggling most are the ones who waited for more information before acting. The labor market does not pause while you think it over. The best time to explore your options and build new ones is when you are still employed, still current in your field, and still in a position to be selective about what you move toward. Urgency and panic are not the same thing. You can be deliberate and still move quickly.
Learn Enough AI to Stop Being Afraid of It
You do not need to become a data scientist or learn to write code. You need to understand the tools well enough to use them in your specific field and to speak about them credibly in an interview or performance review. That bar is lower than most people think, and the confidence that comes from crossing it changes how you carry yourself professionally. I have seen it happen dozens of times.
05
The Part of This Story That Is Actually Good News
Most of the jobs that will exist in 2040 do not exist today. That sounds unsettling until you realize it has always been true. “Cloud migration consultant” was not a job anyone could have described in 2001. The entire ecosystem of roles around social media management, UX research, and data privacy compliance did not exist a generation ago.
Right now, new categories are forming in AI governance, human-AI workflow design, robotics operations, and AI ethics oversight. New business formation is surging, with a measurable correlation to AI adoption rates. New apps are hitting app stores at a sixty-percent year-over-year pace. The economy is not running out of work. It is generating new kinds of work faster than most people realize.
New business formation is already surging alongside AI adoption. New apps are hitting stores at 60% year-over-year. Robotics, a field once limited by computational constraints, is now expanding rapidly with AI at its core. The economy is not a museum of existing roles. It is a machine that keeps generating new ones, and it is running faster than usual right now.
The professionals who will be sitting in those new roles in five years are not waiting for a perfect moment. They are making small moves now: taking a course, reframing their resume, having conversations, building adjacent skills. The accumulation of those moves is what separates the people who land well from the ones who get caught flat-footed.
You already have more going for you than you probably think. The goal is to figure out what that is and point it in the right direction before someone else gets there first.
The leverage is there. You just have to use it.
