Think back to the first real job you had in your field.
It was the plain, repeating work that nobody brags about. You built the slides that someone else would present. You checked the report. You sat in a meeting you barely understood. You took notes. You watched how the leaders handled the moment when everything seemed to go wrong. You did not know it then, but that simple work was how you learned. That is where you built good judgment. That is where you learned how things really work. You picked it up one small task at a time.
Here is the question I keep thinking about. AI is now doing all of that work. It is doing the first jobs where people learn by doing. So where will young workers learn what you learned?
We are about to find out. And I do not think we are ready for the answer.
Entry-Level Work Is Disappearing
This is already happening. The numbers make it clear.
In November 2025, Erik Brynjolfsson and his team at the Stanford Digital Economy Lab shared a study. It was named Canaries in the Coal Mine. They used pay records from ADP, the biggest payroll company in the country. They found a 13% relative drop in jobs for workers ages 22 to 25 in the fields most affected by AI. Think of software developers. Think of customer service. These are the first jobs where future leaders used to learn. At the same time, jobs for older workers with more skill and practice stayed the same or grew.
Read that again. The work is going away for people at the very start of their careers.
You can argue about the exact numbers. But the trend is clear. Entry-level job postings are down about a third since the start of 2023. More than four in ten companies told surveyors they plan to replace jobs with AI. Entry-level jobs are the first to go. These jobs are being cut quietly. Most companies are happy about the money they save. But they are not asking the next question.
If nobody starts in entry-level jobs, who will be ready for senior roles in ten years?
The Middle Is Thinning Too
Entry-level jobs are going away. And the middle of the company is shrinking too.
Between May 2022 and May 2025, the number of managers at public companies fell by 6.1%. Korn Ferry's 2025 research found that 41% of employees say their company has cut management layers. Google cut about a third of its small-team managers. Meta, Intel, Amazon, and others have also cut manager jobs in public. Gartner says that through 2026, one in five companies will use AI to cut more than half of their middle managers.
People even have a name for it now. They call it the Great Flattening.
Here is why this is more than a money story. Middle managers did more than keep things organized. They also helped people grow. A strong worker would get two people to lead, then five, then a small team. That is where they learned the hardest lesson in leadership. They learned how to get work done through other talented people instead of doing it all alone.
"When you flatten the middle, you also remove the place where managers become leaders."
What about the managers who keep their jobs? They now have more and more people to manage. The average team size for each manager has gone up. And the simple work that AI was supposed to take is landing on their desks instead. So the people who could best coach young workers have less time to coach than ever. We are asking fewer leaders to grow more people in less time. And we call it being efficient.
Why This Should Worry You
For many years, the best research on how leaders grow has pointed to the same place. Leaders grow through everyday work, not the classroom.
The Center for Creative Leadership built on the work of researchers like Morgan McCall. They gave us the 70-20-10 model. About 70% of leadership growth comes from hard, real jobs on the job. About 20% comes from good relationships, like mentors and managers who let you see how they think. Only 10% comes from classes and training.
Look closely at those numbers.
Ninety percent of how a leader grows happens during real, daily work and real relationships. It is the hard task that was a little past what you could do. It is the manager who walked you through why they made a certain choice. It is the mistake you made on something small. It did not hurt the company, but it taught you a lesson you never forgot.
That is the process AI is now breaking. When the tool writes the first draft, the young worker never fights through the hard middle part where judgment is built. When the tool handles the planning, the future manager never learns how to plan. Deloitte said this plainly in its 2025 Human Capital Trends research. It warns of a growing "experience gap." That gap keeps getting bigger unless companies work on purpose to rebuild the ways people used to grow.
The hardest part is the timing. The savings come right away. They show up on this quarter's numbers, and someone gets a bonus. The cost comes later, and you cannot see it. It shows up three to five years from now. That is when too few leaders are ready for promotion. The plan for who takes over stalls. And a room full of senior leaders asks why no one is ready to step up.
By then, the person who approved the cuts may be gone.
The Honest Counterargument
I owe you the other side, because it makes some good points.
Some people are hopeful. They say AI makes early-career learning better instead of erasing it. A new worker is free from the boring work. So they can do thinking work much sooner than we did. They will not spend two years making slides and fixing reports. Instead, they will spend that time checking what the AI makes. They will catch its mistakes. They will decide which hard cases need a person. And they will learn how the business works faster than we ever did. The World Economic Forum has shown this exact change in entry-level work. The work is moving away from doing tasks and toward checking and judging.
I want this to be true. In some of the best companies, it already is.
But there is a quiet guess hidden inside all that hope. It guesses that someone is teaching that judgment to the new worker on purpose. It guesses that a manager is sitting next to the 23-year-old. It guesses the manager is explaining how the choice gets made, why this AI answer is wrong, and what the machine is missing. That coaching does not just happen. It used to come built into the simple act of checking a young worker's work. Now it has to be planned on purpose. And it has to come from the same managers we just watched get cut and stretched too thin.
A team at IMD said it well. The hard part now is building judgment while AI does the work. The chance is real. It just does not happen on its own. And in a busy company that is trying to save money, anything that does not happen on its own is the first thing to get skipped.
What Leaders Worth Following Do Now
I will not name a problem that has no answer. This is a design flaw, and a design flaw can be fixed. Here is where to start.
Put development on the books. If 70% of growth comes from hard experience, then stretch jobs are too important to leave to luck in a flattened company. Name them. Assign them. Track them as carefully as you track money. For every junior role, ask this: where exactly is this person being pushed past what they can do now? If you cannot answer, you have found your first gap.
Pick one early-career person on your team. Write down the one hardest, most growth-building thing they will do in the next 90 days. If the honest answer is "nothing that AI is not already doing," then you have just found the weak spot in your future leadership pipeline.
Coach the thinking behind the task. When AI writes the first draft, the old way of coaching by marking up someone's work goes away. Put something better in its place. Say your choices out loud. Tell your people what you decided. Then tell them how and why you got there. Let them watch you look at an AI answer and say no to it. People learn judgment mostly by watching someone use it. So they need to see you do it.
Protect the managers who mentor the future leaders. You cannot flatten the company, give a manager more people to lead, hand them the extra junior work, and still expect them to mentor anyone. If growing people matters, give your remaining leaders the room, the time, and the clear job to do it. Then reward them for the leaders they grow, along with the targets they hit.
Measure your leadership pipeline alongside the budget. The savings from cutting entry-level jobs are easy to measure. The harm to your leadership pipeline is easy to ignore, right up until it becomes a crisis. Put both in front of the people who decide, on the same page. Make the later cost easy to see before anyone signs off on the quick savings.
The Choice In Front of Us
Here is what I believe, and I believe it strongly.
Neglect is what breaks the leadership pipeline. AI only removed the kind of growth we built by accident. That was the early-career learning that came built into simple entry-level work. That automatic growth is gone now. So whatever takes its place will have to be built on purpose. It will take leaders who see the flaw in this new design.
Whatever we allow today becomes the culture that shapes tomorrows leaders.
So now you can see the warning signs. Do not let your company ignore the danger and save its way into a leadership crisis before it is too late to fix. Look at the people in your entry-level roles. They are the ones AI could most easily replace. Ask a better question than the one your spreadsheet is asking. Do not just ask can we automate this? Ask how are we going to grow the person who used to do that work?
The next group of leaders is sitting in your company right now. They are doing work a machine could do. They are waiting to find out if anyone is going to grow them into the plan for who takes over.
That somebody is you.
It's time to step up and lead.
Sources
- Brynjolfsson, Chandar & Chen, Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence, Stanford Digital Economy Lab / SIEPR, November 2025. siepr.stanford.edu
- Yale Insights, The Real Job Destruction from AI Is Hitting Before Careers Can Start. insights.som.yale.edu
- CNBC, AI is not just ending entry-level jobs. It's the end of the career ladder as we know it, Sept 2025. cnbc.com
- HR Dive, Over one-third of companies plan to replace entry roles with AI, survey says. hrdive.com
- Axios, Middle managers in decline as "flattening" spreads, AI advances, July 2025. axios.com
- Korn Ferry / Lepaya, The Great Flattening: Middle Management Cuts in the AI Era. lepaya.com
- Center for Creative Leadership, The 70-20-10 Rule for Leadership Development. ccl.org
- Deloitte, 2025 Global Human Capital Trends (the experience gap), via RewardsDNA analysis. rewardsdna.com
- eMarketer, AI helped cut costs in 2025 but may hurt future leadership pipelines and increase workloads. emarketer.com
- World Economic Forum, How AI is changing the nature of entry level work, 2026. weforum.org
- IMD, How to build judgment when AI does the work. imd.org
- Gartner middle-management prediction, via People Managing People, The Great Flattening. peoplemanagingpeople.com