**
We keep asking how many jobs artificial intelligence will replace. But we rarely ask a far more important question: If AI replaces the work of entry-level employees, where will the senior professionals of the next decade come from?
**
While reflecting on how artificial intelligence is transforming software engineering, cybersecurity, and the translation industry, I came across a contradiction that deserves serious attention.
Today, an experienced engineer equipped with AI tools can accomplish tasks that previously required several junior engineers.
From a business perspective, this represents remarkable technological progress: lower labor costs, faster delivery, and greater short-term efficiency.
But from the perspective of talent development, the implications may be entirely different.
A senior engineer does not become an expert simply by mastering more theoretical knowledge. Expertise is built through years of practical experience, complex challenges, repeated mistakes, and continuous learning.
If AI takes over the very tasks through which junior employees traditionally develop their skills, and companies consequently stop hiring and training newcomers, a serious question emerges:
We may be using today's senior engineers and AI systems to replace the very opportunities needed to develop tomorrow's senior engineers.
This is not merely a question of unemployment.
It is a structural problem that could threaten an industry's long-term ability to develop and replenish its professional workforce.
1. The Value of Entry-Level Jobs Has Never Been Limited to Completing Simple Tasks
Consider a fundamental question:
How does someone become a senior cybersecurity engineer with ten years of experience?
Nobody graduates from university already possessing the skills required to investigate sophisticated security incidents, audit complex systems, or evaluate critical vulnerabilities.
Professional expertise generally begins with relatively simple responsibilities.
At the beginning of their careers, junior engineers may organize security logs, investigate basic alerts, write assessment reports, or examine common vulnerabilities.
After several years, they gradually encounter more complicated security incidents. They learn to distinguish false positives from genuine threats, investigate connections between systems, and evaluate the business consequences of technical weaknesses.
Only after accumulating substantial experience can they independently assume responsibility for complex security decisions.
Entry-level jobs do more than generate economic output. They also produce the next generation of skilled professionals.
This distinction is frequently overlooked.
From the company's perspective, a task is evaluated according to the time and money required to complete it.
From a junior engineer's perspective, however, that same task also represents an opportunity to learn.
An experienced engineer might identify a false security alert in ten minutes, while a beginner could spend several hours investigating it.
Measured purely by immediate productivity, the beginner appears less valuable.
Yet during those hours of investigation, the beginner develops analytical methods, technical knowledge, and professional judgment.
These are precisely the capabilities that eventually form the foundation of expertise.
Therefore, the fact that AI can complete an entry-level task more cheaply does not mean that the corresponding entry-level position has lost its value.
The value of a job should not be measured solely by what it produces today, but also by the professional capabilities it helps create for tomorrow.
2. This Is Not Pure Speculation: Early-Career Employment Is Already Showing Warning Signs
On August 12, 2026, Stanford University's Digital Economy Lab released an updated version of its study, Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence.
Using payroll and employment data from ADP covering millions of American workers, the researchers examined changes in employment following the widespread adoption of generative AI.
They found that:
Among occupations with high exposure to AI, workers aged 22 to 25 experienced an approximately 19% relative employment gap compared with employment growth trends among similarly aged workers in less AI-exposed occupations.
More experienced workers did not experience a comparable employment gap.
The researchers also found that much of the adjustment occurred through reduced hiring of younger workers rather than widespread dismissals of existing employees.[1]
It is important to interpret these findings correctly.
A 19% relative employment gap does not mean AI directly eliminated 19% of jobs.
The researchers also acknowledged that their findings were primarily descriptive and should not automatically be interpreted as proof of causation.
Nevertheless, the results raise an important question:
Why are younger workers in certain AI-exposed occupations experiencing significantly greater employment pressure than their more experienced counterparts?
Research on the software industry provides additional clues.
In June 2026, the Institute of Labor Economics (IZA) published a study examining how generative AI is redefining entry-level software work.
Using extensive online job-posting data from the United States, the researchers found that demand for junior software development positions had declined by approximately 14% to 15% relative to senior positions.
Even more importantly, companies were increasingly raising experience requirements within existing job categories rather than simply changing job titles.[2]
In other words, companies may still advertise positions as entry-level, while simultaneously expecting applicants to possess greater professional experience.
What does this suggest?
A growing number of companies may want employees who already possess experience, without being willing to bear the cost of helping newcomers acquire that experience.
This is the fundamental talent-development problem that concerns me.
3. What AI May Actually Undermine Is the Path from Junior to Senior
Traditionally, professional development follows a relatively recognizable progression.
Junior employees begin with lower-risk tasks and learn through practical experience.
As their knowledge and judgment improve, they take on increasingly complex responsibilities.
Eventually, some become senior professionals capable of independently managing major projects and making critical decisions.
But when AI can perform many basic tasks, companies face a new set of incentives.
Imagine a company that originally employed five senior engineers and ten junior engineers.
As AI coding assistants and automated analysis tools become more capable, those five senior engineers may be able to accomplish a significant portion of the work previously handled by the entire team.
The company consequently reduces its recruitment of junior engineers.
In the short term, its performance indicators may look excellent.
Operating costs decrease. Productivity rises. Projects are completed faster.
But what happens five or ten years later?
Existing senior engineers will eventually retire, change careers, or leave the organization.
Where will their replacements come from?
Some might answer: recruit experienced professionals from other companies.
But if more and more companies adopt the same strategy, the problem spreads from individual organizations to the entire industry.
We can express the underlying logic through a simplified model:
The future supply of senior professionals depends on how many people entered junior positions years earlier, the practical learning opportunities available to them, the quality of their training, and how many remained in the profession.
A reduction in entry-level opportunities does not necessarily create an immediate shortage of senior talent.
Companies still have experienced employees who developed their skills under earlier employment structures.
However, professional development involves significant time delays.
The consequences of reducing junior recruitment today may not become fully visible for several years.
This creates a dangerous possibility:
Short-term productivity gains and long-term declines in the supply of skilled professionals can occur simultaneously.
I believe this is one of the most underestimated structural risks associated with AI-driven changes in employment.
4. The Cybersecurity Industry Deserves Particular Attention
One reason I am especially concerned about this issue is the nature of cybersecurity work.
In cybersecurity, theoretical knowledge and professional judgment are not
Top comments (0)