When AI takes over the beginner tasks engineers learned from, do juniors lose the hard struggle that makes them experts?

This explores why handing entry-level work to AI might remove the hands-on difficulty that turns juniors into experts. The clearest evidence comes from interviews with South Korean software engineers Does generative AI prevent juniors from getting entry-level work?. The tasks juniors used to learn on, such as boilerplate, small bug fixes and first-pass implementations, are now done by seniors working with AI. The work still gets done, but it moves into a senior-plus-AI workflow, and the junior is no longer the one doing it. The authors also found that seniors and juniors see the problem differently. Seniors usually see a productivity gain. Juniors see the loss of a ladder they need to climb, and only juniors are affected by it.

The less obvious consequence is that seniors face the same problem in a milder form. In Anthropic's internal survey of 132 people, engineers reported roughly a 50% productivity boost and 67% more merged pull requests, yet most said they could fully delegate only 0–20% of their work Does AI assistance erode the skills needed to oversee it?. Because they can delegate so little, they still have to check what the AI produces, and they worry that leaning on Claude for routine work wears away the hands-on practice they need to catch its mistakes. So formative struggle is not only how juniors learn. It is also how anyone keeps the judgment needed to supervise AI. A junior who never does the routine work may never build the skills that would let them oversee the AI doing it.

One helpful way to see why struggle matters comes from an argument about AI separating the finished form of intellectual work from the thinking that produced it Does AI separate intellectual form from the thinking behind it?. Working code that a junior did not write looks the same as working code they did write, but the understanding behind it is missing. A hiring study finds the same pattern in a different market. When writing became cheap on Freelancer.com, employers lost a signal of effort, and hiring became 19% less merit-based Does cheap writing weaken hiring based on worker ability?. Effort was doing two jobs at once: it built skill, and it showed others who had that skill. Removing it can damage both.

The environment juniors work in also shapes the problem. A study of 10 junior and 10 senior engineers found that company policies, such as required tools, allow-lists and data rules, decide how much control engineers have over agentic AI before personal preference comes into play Does personal preference shape how engineers use AI tools?. Within those limits, novices swung between relying on AI too much and avoiding it entirely, and neither builds skill well. There is also a quieter loss. A field experiment at Procter & Gamble found that individuals using AI performed as well as two-person teams without it Can generative AI replace the benefits of having a human teammate?. In those settings, AI can take the place of the colleague, and for many juniors, pairing with a senior was where much of their struggle was guided.

The corpus has gaps. It documents that formative struggle is being displaced, but it has little learning-science material on why struggle builds expertise or on whether deliberately designed struggle could replace it. One hopeful thread comes from labor economics. When AI affects only a few of the tasks in a job, workers can shift to the tasks it does not affect, and net job losses stay modest Does concentrated AI exposure enable workers to adapt and reallocate?. The open question is whether junior roles can be redesigned so that the tasks left to them still teach them something.