The Impact On Workers And The Economy

When AI replaces entry-level jobs, we lose the Apprenticeship Dividend

getty

AI is beginning to squeeze the entry point into the workforce. New Stanford research using ADP payroll data shows that AI replacing entry-level jobs is already showing up in the numbers, with the steepest drops in roles like software development and customer support. Emerging information is unusually consistent on this point, suggesting a clear shift is coming into focus. In June, I wrote about what happens when AI does more of the work but companies fail to train the next generation of leaders. This new data supports that concern, and it is no longer a hypothetical.

Now that there is evidence that the entry point is narrowing, it’s imperative that we explore how this will affect job mobility, wages, innovation, and the overall vitality of the economy.

AI Replacing Entry-Level Jobs? Consider the Apprenticeship Dividend

Every healthy company needs a talent pipeline — a phrase so familiar it’s easy to dismiss. What it really represents is an Apprenticeship Dividend: the long-term return businesses earn when new employees learn by doing, grow into greater responsibility, and then teach others what they’ve mastered. When entry-level jobs disappear, that cycle breaks. The most obvious consequences are slower job mobility and weaker wage growth.

The less obvious consequences are just as serious. Companies lose the learning loop that turns new talent into deep-seated ability. When more experienced employees are required to teach and train those at the start of their careers, they clarify and solidify what they know while transferring it to others. They are also asked questions they may not have considered before (or in a long time), which expands their awareness of the work. All of this multiplies learning for individuals, while spreading capability throughout the team.

At the market level, something else happens. Economists describe the “flow of talent” — the way people move from less productive firms to more productive firms over time. That flow is how skills spread, innovation travels, and productivity increases across an economy. If you narrow the entry point, that flow slows down: talent stagnates before it can be developed, and the whole system loses momentum.

These are not just human resource issues. They are competitiveness concerns. If individual businesses treat the AI-and-jobs impact the way communities treat unwanted development — NIMBY, or Not In My Backyard — we end up with the same problem on a different scale. Let’s call it NIMCO, or Not In My Company. In a NIMCO scenario, each firm captures the short-term savings of eliminating entry roles while assuming someone else will carry the long-term burden of training and development. But just as with NIMBY, the costs don’t vanish; they accumulate and spread. The pain shows up in weaker innovation, anemic talent pipelines, and an economy that cannot sustain future growth.

Leaders in law, accounting, banking, and tech are already raising alarms that automation combined with hybrid work is eroding apprenticeship and on-the-job learning. When you remove repetitive tasks and proximity, much of the tacit learning disappears.

What Are the Economic Effects of AI Replacing Entry-Level Jobs?

Dynamic economies depend on movement. People enter at the bottom, build skills, and then move between companies and roles as opportunities grow. That fluidity is not just about individual careers; it’s how new ideas spread, best practices circulate, and productivity gains are compounded. When that movement slows, skill development weakens, wages rise more slowly, and the pace of overall innovation declines. In other words, the knock-on effects of losing entry-level jobs extend well beyond any one firm. They affect the vitality of the economy itself.

Looking ahead, forecasts for the next five years tell a complicated story. The World Economic Forum’s Future of Jobs Report 2025 projects that while technology will create an estimated 69 million new jobs globally, it will also displace 83 million, with clerical, secretarial, and other cognitive roles among the most at risk. The report emphasizes that “net job creation is positive in aggregate,” but the transitions will be disruptive, particularly for workers without access to retraining and reskilling. This is why entry-level opportunities matter so much. They are not only the first step onto the economic ladder for young workers; they are also the safety net for displaced workers trying to re-enter the market. Without them, we create choke points that limit career mobility, career transitions, and overall economic growth.

The Innovation Loss Companies Don’t See Coming

What we need to be concerned about is the innovation that will fail to happen when junior talent disappears. Breakthrough ideas are not generally born in the C-suite. Instead, big ideas tend to emerge when fresh eyes collide with experience; when someone still close to the customer or the process notices something everyone else has learned to overlook. Take away those entry roles, and you take away much of the raw material that fuels innovation.

History offers ample examples of this: 3M’s Post-it Notes, Google’s Gmail, Sony’s PlayStation, Starbucks’ Frappuccino, and Apple’s iPod click wheel. Each breakthrough began, not in the boardroom, but with employees experimenting, noticing, or challenging assumptions. Innovation thrives where questions are asked, assumptions are tested, and unconventional ideas are allowed to emerge. Those conditions often come from people early in their careers. If you optimize them out of the workforce, you also optimize out a significant share of your future differentiation.

The Homogeneity Risk Is Real

Another danger of leaning too hard on AI is loss of differentiation. When every company uses the same large models, trained on the same public data and prompted in similar ways, the results all start to look the same. A customer service script that sounds like everyone else’s may not seem catastrophic, but over time this sameness seeps into brand voice, process refinements, product ideas, and strategic choices.

Researchers call this “model collapse.” Model collapse happens when AI starts learning from its own recycled answers, and over time the results get narrower and less creative. In practice, businesses risk a related kind of collapse: a slide into blandness and homogeneity. The brochures sound alike, the code repeats the same flaws, the strategies cluster around safe bets. What disappears is differentiation, and differentiation is what keeps companies visible, competitive, evolving, and growing.

Assumed Knowledge Versus Applied Knowledge

It is more important than ever to understand the difference between assumed knowledge and applied knowledge, because we are entering a danger zone where assumed knowledge may dominate (and in many fields, it already does). Assumed knowledge is what we pick up from reading, listening, or being told — facts and information that can be repeated but not necessarily understood deeply. Applied knowledge comes from using information in real situations: solving problems, dealing with unpredictable customers, and facing the consequences of decisions. The first can make us feel informed; the second is what makes us capable.

AI delivers assumed knowledge in abundance. It can provide an answer instantly, and in that moment it feels like knowing. But unless there is applied experience to test, question, and challenge that answer, it’s impossible to tell whether it is correct or simply convincing. Even when pushed for more detail, AI is designed to satisfy the user, not guarantee accuracy. The result is a growing risk that people feel more confident without actually becoming more capable.

The risk extends beyond juniors who never get the chance to build skills. Experienced employees who rely too heavily on AI also learn less over time. The mistakes that result may not always be catastrophic, but hundreds of small errors — in code, contracts, compliance, or customer interactions — can weave together into a fragile organization. A company built on micro-errors will be less able to compete.

Taken together, these risks point to more than a workforce issue. They point to a weakening of the very systems that make economies dynamic: the flow of talent, the spark of innovation, the development and transfer of applied knowledge, and the differentiation that keeps businesses competitive. If entry-level opportunities disappear, each of those systems will be diminished, and once they start to erode, the result will be a downward spiral that is hard to stop.

This is why AI replacing entry-level jobs cannot be treated as a narrow employment story. It is a competitiveness story, a growth story, and even a resilience story. The economic effects are clear; what remains is the harder question of how businesses themselves should respond. That is the focus of the next article in this series: what AI replacing entry-level jobs means inside individual companies, and how leaders can manage today’s cost pressures without trading away tomorrow’s strength.

This is the first of two articles. Here the focus has been on the broad economic and societal consequences of AI replacing entry-level jobs. The second article looks at what this means inside individual companies — and how leaders can weigh today’s cost savings against tomorrow’s competitiveness.