Every senior I ever hired was somebody's junior first. So was I.
My own junior years started in 2000 at 24/7 Media, an international company competing with DoubleClick in the years before Google bought DoubleClick. It was my first job after studying business intelligence, and I learned mostly by doing things wrong at a survivable scale. Four employers later, in 2009, I co-founded This Page Amsterdam, and across twelve years there we hired juniors constantly. Designers, developers, producers. Some of them left as seniors. A few of them became better than the people who trained them, which is how it is supposed to go.
Looking back at those hires, I can see something I never once said out loud in a hiring conversation: the junior role was doing two jobs, and we were only paying attention to one of them.
The job we paid for, and the job we got for free
The job we paid for was cheap production. A junior billed out at a rate above their salary, took the work the seniors should not be spending time on, and made the pyramid work. Slicing designs, fixing bugs, resizing the campaign into forty formats, building the page that was almost the same as the last page. Every agency P&L in the industry rested on that arithmetic, whether the owner ever drew the pyramid or not.
A second job came bundled in, invisible and unbudgeted. All those boring hours were an apprenticeship. The junior fixing bugs was learning what breaks. The junior resizing banners was learning why the design system existed. The junior sitting near a client call was absorbing how a scope conversation actually sounds, which is not something a course can teach. Judgment was being built out of repetition, proximity, and small failures that cost the business almost nothing.
Nobody invoiced for the second job. It looked free. It was actually the industry's entire pipeline for producing seniors, running quietly underneath the production work that justified the desk.
The arithmetic just broke
AI is eating the first job, and it is eating it fast. The production work a junior needed a week for, a senior with a capable model finishes before lunch, at a higher standard, without the review round. I know this from the inside, because it is how I work now. Most of the code I ship is written by AI. The tasks I would once have handed to a junior, I hand to a model, and the model does not need the work explained twice.
Once the production arithmetic breaks, the junior desk stops paying for itself, and businesses respond the way businesses respond. Junior hiring is the easiest line to freeze, because nothing visibly breaks when you freeze it. The seniors are still there. The work still ships. The margin actually improves for a while, which makes the freeze look like a good decision.
Nothing visibly breaks, that is, for about five years.
Where the 2032 seniors come from
Nobody in the industry seems to be asking this seriously: if the junior desk disappears, where does the next generation of seniors come from?
Judgment does not arrive by download. Mine was built on thousands of hours of exactly the work I would now delegate to a model. I learned what good looks like by producing a large volume of not-good and having it corrected by people three seats away. I learned to smell a scope problem in a brief because I had lived through the projects where nobody smelled it. That kind of knowing has only ever been grown one way: real work, real stakes, cheap mistakes, someone senior close enough to catch you.
Courses do not replace it. A course is the map. The apprenticeship was the territory, and the territory was funded, accidentally, by the production work AI just absorbed. Remove the funding and the industry has not made a decision about training. It has made the decision without noticing there was one.
The part I find most uncomfortable is my own seat in this. I run a solo practice where AI holds the roles a junior once held. I climbed the ladder, benefited from every rung, and my current way of working removes rungs for whoever comes next. I do not have a tidy answer for that. I notice it.
People, and therefore margins
It would be easy to file this under culture, or ethics, or the general category of things that are sad but not urgent. I think that filing is wrong, and I think the reason it is wrong is financial.
Seniors are about to become the scarcest asset in the industry. The generation trained the old way, on boring reps that no longer exist, will spend the 2030s as the last people who learned judgment at production scale. Scarcity does one thing to price. The agencies that froze junior hiring to protect this year's margin are setting up a decade where senior salaries, senior retention, and senior poaching become the margin problem, with no internal pipeline to relieve the pressure.
Somewhere in that gap, a few businesses will figure out what apprenticeship looks like when production no longer funds it. Deliberate, paid for on purpose, built around the judgment instead of the output. Those businesses will spend money in years when spending it looks unnecessary, and they will own the seniors everyone else is bidding for later.
The junior problem is not really about juniors. It is about whether the industry understood what it was buying all those years, back when the second job came free. The trophy on the P&L was cheap production. The thing of value was the pipeline. We priced one, we got both, and we are about to find out what the unpriced one was worth by declining to pay for it.
Originally published on LinkedIn.