A Case for Hiring Junior Engineers in the Age of AI
How GenAI is reshaping teams and why juniors still matter to ensure long-term success.
If you work for a mid to large-sized tech company, chances are high that most of your open positions (if you have any) are for senior engineers. After the exponential demand for software developers during the covid crisis, the hiring market for developers has reached its lowest point in the past five years, according to Gergely Orosz’ findings in the Pragmatic Engineer.1 While the reasons for this shift have been widely discussed, there is a broader shift in hiring more senior candidates. When talking to people in the industry, three assumptions regularly come up:
Juniors are slowing down senior engineers
Senior engineers are more productive due to GenAI
A lot of the traditional work juniors do is now done by AI
Let’s break these assumptions down.
Juniors are slowing down senior engineers
There’s some truth here. Mentoring takes time. Onboarding, reviewing PRs, and explaining architectural decisions all pull senior engineers away from their own deliverables. In high-pressure environments, this can feel like a drag. However, this view ignores the long-term benefits of growing talent internally.
Juniors bring fresh perspectives. Their questions often surface hidden assumptions. Their eagerness to explore new technologies—sometimes too early—can be annoying but occasionally unlocks new perspectives and potential breakthroughs. Additionally, teaching others has a funny way of exposing your own knowledge gaps and reinforcing your expertise. Pair programming, for instance, is not just about mentorship— it’s a form of active debugging for your thinking. If used correctly, this can have a compounding effect on your overall team’s productivity.
Seniors are more productive with GenAI
Senior engineers can indeed use GenAI tools like LLMs more effectively. They understand how to craft better prompts, set proper context, and—most importantly—review the output critically. According to the “Anthropic Economic Index”, GenAI has increased productivity across many industries, but especially in software engineering, where the benefits compound with experience.2 However, productivity gains don’t translate into needing fewer people; they usually mean more gets built faster. More features, more iterations, more complexity. More maintenance.
The bottleneck shifts—not disappears. We still need experienced engineers to review, debug, and steer the outputs of AI. As the systems we build grow more complex, the demand for judgment, context, and architectural thinking grows with it. In fact, as GenAI becomes more widespread, the need for subject-matter experts increases. Or, as IBM famously stated in the 70s:
A computer can never be held accountable, therefore a computer must never make a management decision.3
The work juniors used to do is now done by AI
There is some truth in this. A lot of boilerplate, CRUD logic, or basic scripting can now be generated by AI. But real-world systems aren’t just code, they’re living ecosystems that require debugging, tuning, context, and care. AI may handle repetitive tasks, but it doesn’t replace the need for humans to understand why something is built the way it is.
The real danger here is not that GenAI replaces junior engineers, it’s that we stop giving people the chance to become senior engineers in the future. If we stop hiring juniors today, who will be our seniors in five years?
Conclusion
The current market favors senior engineers, but not because juniors have become irrelevant. It’s a response to short-term pressures and the allure of quick productivity wins. But, organizations that fail to invest in junior talent risk long-term stagnation. Innovation doesn't just come from expertise, it comes from curiosity, naivety, and ambitious people with something to prove.
GenAI is a tool. A powerful one. But tools don’t build great software — teams do, and they need all levels of experience.
https://blog.pragmaticengineer.com/software-engineer-jobs-five-year-low/
https://www.anthropic.com/news/the-anthropic-economic-index
https://knowyourmeme.com/memes/a-computer-can-never-be-held-accountable