A hiring manager reads the job description for a junior role they’ve run for five years and realizes most of the tasks on it are things a model does now in a fraction of the time. The question they’re sitting with is whether the role is still worth running, and what it would have to look like if it is.
The junior role has always been the entry point where a person gets paid to learn. The work given to a junior is work that needs doing, and at the same time it’s the work a senior would have to teach them to do anyway. The company gets a rung of capacity that’s cheap to run, and the junior gets a year or two of practice before they’re pushed up to harder work. Both sides get something out of the exchange, and the arrangement has been the shape of the talent pipeline for as long as there have been companies.
The model does the work the junior used to do. The first pass of a document, the first draft of code, the summary of a meeting, the small chart that takes a junior an hour and a model a minute. The work that was given to a junior was work that took a lot of time and a small amount of judgment, and a model does exactly that mix better than a junior did. In some kinds of work, a senior with a model now produces what used to require a senior plus junior support.
Some companies respond by shrinking the role, or removing it for a cycle to see whether the work still gets done. The numbers look clean in the short run. Headcount comes down without output falling, and the cost per output drops. But the company that does this loses the pipeline by which it used to produce its senior people, and the loss shows up as a shortage two or three years later, when the current seniors retire or leave and there is nobody in the company who has done enough of the work to step into their role.
Others redesign the role. Instead of giving the junior the work the model now does, the junior is given the work a senior used to do, one level up, while the model does the layer below them. The junior’s first year looks different in shape. They produce fewer first drafts and review more of the model’s first drafts, deciding what to keep and what to send back. The work is harder than the old junior role was, because it starts where the senior work used to start, but it’s more faithful to what the job becomes later, which was always the point of the junior year.
Both options cost something. Shrinking the role costs the pipeline, invisible for two years and then expensive. Redesigning the role costs the easy onboarding, because a junior who used to take six months to contribute now takes a year, since the work they’re being asked to do is work that used to sit one level up.
Most companies haven’t chosen yet. They’re running the old job description and watching it attract fewer good candidates, and they’re hiring fewer juniors than they used to without having decided to, because the hiring bar has quietly risen and the candidates the old role would have attracted aren’t applying anymore. The choice gets made either way. The company that makes it on purpose gets to keep the pipeline in a shape that works. The company that defaults into the shrink finds itself two years from now with no one ready to step up, and no way to fix it short of hiring from outside at the rank they need, at what that costs.