Don’t Kill Entry-Level Jobs. Turn Them Into Your Competitive Advantage
AI won’t eliminate early-career work — it will transform it. Here’s how to redesign it.
What’s the Problem? is publication exploring the future of product building, written by a longtime product leader whose experience spans both the largest (Amazon, Airbnb, CBS) and the smallest (Haven, SET, Jacent) companies around. Like any good product discussion, we start with the question: What problem are we trying to solve?
What’s the problem? How do we redesign—rather than eliminate—early-career roles into a competitive advantage?
We’ve all read the headlines: entry-level hiring plunging, unemployment for recent grads soaring. And while there are a multitude of reasons driving that, it’s clear that AI—whether in reality or perception, has reshaped the job market. Entry-level jobs are changing, and they aren’t going back. But this isn’t a doom-and-gloom story; it is a story of hope and adaptation. This post is about how employers can redesign roles now, and why it is actually a competitive advantage to do so (next post will be for workers). I’ll focus on tech because that’s what I know, but the patterns apply far more broadly.
The Myth of the Vanishing Entry-Level Job
Increasingly, leaders now believe the work historically done by entry-level employees has disappeared or is about to be automated away. Whether that’s true yet is debatable, but let’s assume it is. If you look at product management as an example, this includes tasks such as writing requirements/stories, reporting on status, and pulling insights from customer feedback, pulling data, etc. Across industries, these are the tasks people assume AI can replace.
The result is a belief that companies can lean almost entirely on mid- and senior-level workers, using AI to handle the “busy work” while experienced employees focus on high-judgment tasks. A lean, hyper-productive team doing the work of many for the cost of a few million tokens.
The Consequences of Cutting Junior Roles
While there are definitely short-term cost savings to be had, pursuing a low- or no-entry-level strategy is bad for product quality, bad for companies, bad for the industry, and obviously bad for the workers.
It’s bad for product quality because instead of using our expanded talent pool, AI, and junior PMs to raise the bar, we’re mostly using AI to produce the same things cheaper.
Eliminating junior PMs often means senior PMs doing more, not less, lowering both speed and quality.
It’s bad for companies because removing junior roles breaks the internal career ladder. Without junior PMs working alongside senior ones, no one is being developed into tomorrow’s product experts, or into experts on your company and customers. Junior PMs also bring fresh thinking, ask the “dumb” questions no one is asking, and bring in the latest best practices they learned in college. They also often match the demographics of the customers you most want to attract. And despite AI advances, much of the operational work thought to be replaceable actually still requires humans because systems don’t integrate, data isn’t clean, and outputs must be reviewed. Eliminating junior PMs often means senior PMs doing more, not less, lowering both speed and quality.
It’s bad for the industry because we miss the opportunity to build better products and we starve the talent pipeline. Better products are a real thing, which I don’t want to underappreciate as I think increased revenue and market opportunities generally outweigh cost cutting. Companies might not care—they can poach mid-level PMs from elsewhere. But collectively, this means fewer PMs get trained, senior PM salaries spike (yay!?), and product quality drops under the weight of too few people doing too much work.
If hiring starts at the mid-level, why would anyone start the journey at all?
Product has always been a hard field to break into. No formal training paths, few apprenticeships, and a catch-22 of needing experience to get experience. This AI shift only raises the bar further. If hiring starts at the mid-level, why would anyone start the journey at all?
A Blueprint for Modern Entry-Level Work
I learned product by doing it, watching others do it, getting feedback, and trying again. Formal education gives you tools, but not the judgment or pattern-matching experience to know when to use them. Most workplace skill—especially in product—comes from the “expert–novice bond,” as Matt Beane calls it, author of “The Skill Code: How to Save Human Ability in an Age of Intelligent Machines,” and an associate professor at UC Santa Barbara; Beane’s research indicates this is “a relationship that predates most of what we consider to be civilization.” He says: “The research is clear on this, too—formal learning, at best, just gets you table stakes.”
Entry-level work has shifted, and the role must be redesigned around three responsibilities:
Orchestrate the intelligence instead of doing the tasks.
Bridge the gap between what we think AI can automate and what it can actually do.
Take on the human, high-judgment work we’ve never had time for — the work that truly leads to better products and deeper customer understanding.
The greatest unlock in this model is that junior PMs can make senior PMs more senior. In an AI-driven world, everyone becomes a manager. The junior PM becomes the first line of defense against AI slop. Instead of writing requirements from scratch, they prompt AI for a draft, iterate, ask other agents for critique, and present a strong first pass for senior review. They learn through AI and human feedback (“expert-novice relationship”), senior PMs reclaim strategic time, and in the process raise the overall quality of the product.
If your company is larger than a small startup, you still need junior PMs. On teams without them, you’ll usually find overworked senior PMs or declining product quality, not because senior PMs can’t handle the work, but because they’re stuck filling the gap between what AI should automate and what it actually does today. Systems don’t integrate, data is messy, AI outputs are inconsistent, and someone has to clean, verify, rewrite, and stitch everything together. Junior PMs can now take on far more scope than before, but the operational layer doesn’t disappear; it simply shifts into orchestration, QA, and integration work that is too detailed, interrupt-driven, or context-specific for senior PMs to absorb without tanking their strategic bandwidth.
Finally, most PM teams live with a backlog of deep discovery work they never have time for—the work that would meaningfully improve customer understanding and product quality. This now becomes the junior PM’s job. They talk to users constantly. AI can tell you the WHAT; humans uncover the WHY. AI can build dashboards; only humans ask the questions the dashboard can’t answer. A junior PM is your ears and eyes on the ground at standups, lunches, and side conversations, they listen and engage without intimidating, the way a senior PM might, and feed insights up to product leaders to improve morale and alignment. Product is a team sport, and juniors help keep the team aligned and energized.
Where’d I get it wrong? I’d love to hear it, please leave a comment (or have your bot do it) and subscribe.