AI Is Eliminating Entry-Level Work. Now What?

Early-career employees have always learned by doing the work that more experienced colleagues have moved beyond: building the first draft, reconciling data, troubleshooting code, preparing analyses and sitting in on important decisions. Much of it is repetitive, but it also gives new professionals something essential—exposure to how problems unfold, how mistakes get fixed and how judgment develops.

As AI takes over more of that work, companies face a problem hiding inside a productivity opportunity: If the first rung of the career ladder disappears, what replaces it?

Members of the Senior Executive AI Think Tank—experts in machine learning, generative AI and enterprise AI applications—are seeing this challenge across industries. They point to a new approach: Let AI handle more of the execution while giving early-career employees more responsibility for evaluating work, making decisions, solving problems and learning from experienced leaders.

That means rethinking more than training programs. In the sections that follow, members of the Think Tank explain how leaders can redesign junior roles around judgment and verification, use simulations and AI-generated work as training tools, preserve mentorship and hands-on experience and give emerging talent meaningful accountability earlier in their careers—ensuring that as AI makes work faster, it doesn’t make the path to becoming a capable leader disappear.

“I believe redesigning the junior talent pipeline will be one of the defining leadership challenges of the next decade.”

– Gabriella Goddard, CEO and Director of AI Catalyst Leader™ at Brainsparker Ltd

SHARE IT

Redesign the Junior Role Around Judgment

Gabriella Goddard, CEO and Director of AI Catalyst Leader™ at Brainsparker Ltd, has spent more than 20 years coaching innovators, entrepreneurs and senior leaders across industries. She says the leadership challenge is bigger than protecting individual jobs.

“I believe redesigning the junior talent pipeline will be one of the defining leadership challenges of the next decade,” she says.

She argues that entry-level employees traditionally developed judgment through “repetition: trial, error, feedback and execution.” Employees will now need to move sooner into activities that require interpretation and judgment.

“If AI can generate the baseline output, the role of early-career employees moves up the value chain to questioning it, applying it and improving its impact.”

That means organizations should explicitly define what juniors are expected to learn from AI-assisted work. Critical thinking, collaboration, communication and strategic decision-making should not be treated as skills employees acquire later in their careers.

And managers have to change with them.

“For this to work, leaders’ roles need to evolve as well, shifting from supervising tasks to coaching judgment, critical thinking and managerial skills,” she says.

This is a fundamental redesign of the manager’s role: less quality-control supervisor, more apprenticeship coach.

“Reviewing, stress-testing and verifying automated output becomes the new hands-on training ground for architectural judgment.”

– Vivek Kumkar, Sr. GenAI Leader at Amazon Web Services (AWS)

SHARE IT

Make Verification the New Hands-On Experience

For Vivek Kumkar, Sr. GenAI Leader at Amazon Web Services (AWS), the problem is particularly visible in engineering. He cautions against assuming routine engineering work was merely low-value production.

“Working through routine execution built mental models, debugging instincts and system intuition over time,” he says.

When generative AI handles routine syntax, that repetitious practice can disappear. Kumkar’s answer is to replace coding repetition with more sophisticated forms of hands-on practice.

“Early-career engineers develop stronger judgment by running adversarial testing on model outputs, performing root-cause analysis on distributed edge cases and auditing automated code for security, latency and memory bottlenecks,” he says.

This creates a new apprenticeship model: AI generates, juniors interrogate and senior engineers coach.

“In the multi-disciplinary organizations I lead,” he says, “reviewing, stress-testing and verifying automated output becomes the new hands-on training ground for architectural judgment.”

Don’t Cut the Training Program to Save Money

Rodney Mason, Chief Marketing Officer at Minty, warns leaders about the cost of eliminating entry-level work.

“Reducing entry-level positions, the least expensive in the organization, increases the long-term costs of having trained professionals leading AI and short-term costs of having more expensive senior employees monitor and correct AI mistakes.”

Mason recommends five imperatives, beginning with preserving entry-level training programs while using AI to enhance onboarding and development.

“Design learning deliberately, keeping some AI-doable work human for development sake,” he adds.

That does not mean resisting automation. It means distinguishing between work that creates business value and work that creates employee capability. Some tasks may be worth keeping in human hands temporarily because they teach something the organization will need later.

Mason also recommends a “build and then verify” sequence, arguing that “juniors can’t audit what they’ve never made.” He advises organizations to use simulations and structured decision reviews to accelerate training and track “time to trusted judgment”—a useful metric for determining whether productivity gains are being purchased at the expense of the leadership bench.

Turn AI Output Into a Training Exercise

Mohan Krishna Mannava, Data Analytics Leader at Texas Health, sees a similar problem in analytics. He says AI can remove much of the foundational data work that once helped analysts develop an instinct for what good analysis looks like.

“When entry-level staff skip foundational work, they miss out on developing the intuitive sense for accurate data and practical business decisions,” he says.

His recommendation is to treat AI output as the beginning of the learning process, not the end.

“Require junior team members to fact-check, identify hidden assumptions and find blind spots in AI recommendations before presenting them,” he says.

Mannava also recommends using historical company data to create decision simulations. Rising talent can use AI to revisit old problems, make recommendations and then compare their reasoning with what actually happened.

“Teaching junior staff to question, audit and contextualize machine outputs builds a new training ground that accelerates strategic judgment,” he says.

The principle is broadly applicable: Give junior employees access to AI, but require them to demonstrate the reasoning that makes the output trustworthy.

Move Young Talent Closer to Business Outcomes

Gaurav Rastogi, Senior Director of Enterprise Data Analytics, Data Science and Strategic Insights at Hertz, believes organizations are asking the wrong question when they frame AI as eliminating the need for junior talent.

“One of the biggest misconceptions is that AI will eliminate the need for junior talent. In reality, it eliminates routine tasks, not the need for judgment,” he says.

However, this means leaders will need to redesign the learning path.

“Instead of training people on repetitive execution, we should expose them earlier to problem-solving, cross-functional collaboration, exception management and business outcomes.”

This approach makes early-career roles broader rather than smaller.

“In my experience, the goal is not to protect manual work, but to accelerate human development,” Rastogi says. “The next generation of leaders should spend less time doing tasks and more time understanding why decisions matter and how systems create value.”

“People grow when critical thinking is paired with guidance from experienced leaders.”

– Fredrick Redd, CEO of Metrocap Advisors™

SHARE IT

Make Mentorship the Destination for Freed-Up Time

Fredrick Redd, CEO of Metrocap Advisors™, challenges the assumption that repetitive work automatically produces leaders.

“Repetitive entry-level work was never, by itself, a reliable training ground for future leaders.”

What actually creates leadership capability, he argues, is the interaction between experience and experienced people.

“People grow when critical thinking is paired with guidance from experienced leaders.”

That has major implications for what organizations do with the time AI creates. If AI saves a manager 10 hours a week, the organization can either absorb those hours as additional productivity or reinvest some of them in coaching. Redd recommends the latter.

Future leaders still need opportunities to think on their feet, communicate, resolve conflict, present ideas and learn from mistakes, he says. Those experiences cannot simply be replaced with online training.

“As AI absorbs routine work, organizations should reinvest the time saved in coaching and mentoring,” he says. “Education and entry-level work provide a foundation, but leadership is built through feedback, human interaction, accountability and continuous growth.”

Build a Supervised Judgment Layer

Geetha Kumari Kommepalli, Founder, CTO and Chief AI Officer at Thriven Advisory, believes the central issue is not the disappearance of entry-level work. It is the disappearance of the learning loop.

“AI should remove low-value repetition, not the learning loop,” she says. “The risk is not fewer entry-level tasks; it is fewer chances to see patterns, make decisions, manage exceptions and understand the consequences of being wrong.”

That means redesigning junior roles around supervised judgment.

“Let AI handle the first draft, but require people to validate outputs, investigate exceptions, explain trade-offs and own recommendations.”

Kommepalli also recommends rotations across business processes and deliberate mentorship. Emerging employees should encounter real operational problems rather than spend their first years inside increasingly narrow task silos.

Her broader advice is particularly relevant to executives: “Treat development as an intentional operating capability, not an accidental by-product of manual work.”

That means putting development into operating plans, funding it and measuring whether employees are gaining responsibility.

The result is a different definition of readiness. A junior employee doesn’t have to be “fully ready” before receiving meaningful accountability; the accountability itself becomes part of how readiness develops.

Preserve Accountability, Not Busywork

David Obasiolu, AI Security, Governance and Systems Consultant at Vliso AI, starts with a distinction that should shape every AI workforce strategy.

“AI should remove repetitive work, not remove learning,” he says. “The risk is that junior employees lose the exposure that builds judgment, context and accountability.”

His solution is to build development directly into AI-enabled workflows through supervised decision-making, simulations, rotations and reviews of AI-generated work.

The goal isn’t merely to keep humans in the loop for compliance. It is to make the human review process itself a developmental experience.

“Leaders should measure whether employees are gaining decision quality and domain judgment, not just producing work faster,” he says.

This points toward human-in-the-loop AI. As AI systems take on more responsibility, organizations need employees who understand not only how to use them, but when to question, override or escalate their recommendations.

For early-career talent, that means accountability should arrive earlier—not disappear.

Replace Slide Decks With Human Apprenticeship

Kiran Palla, Chief Information Officer at CogniwareAI, sees a common substitute for hands-on development emerging as organizations automate work: passive learning.

“Many organizations now rely on slide decks and videos, which are efficient but don’t build judgment,” he says. “To rebuild that foundation, I’ve implemented daily 15-minute touchpoints during a new hire’s first 30 to 60 days with different leaders.”

Those interactions give new employees context that cannot be acquired from a training module alone. They also create repeated exposure to how experienced leaders interpret decisions, culture and organizational priorities. That matters because apprenticeship is not simply instruction. It is observation.

“Today’s workforce is strong in self-learning, but consistent leader engagement ensures structured assimilation and preserves the human apprenticeship needed to grow future leaders,” he says.

For executives, Palla’s model offers a practical starting point: Don’t wait for formal mentoring programs to create development. Put short, recurring interactions between leaders and new employees on the calendar from day one.

“Automate friction, not experience, or today’s productivity gain will become tomorrow’s leadership deficit.”

– Fabio Danze Montini, Investor and Owner of FDM Industrial Sales & Marketing SL

SHARE IT

Preserve the Mistakes That Teach

Fabio Danze Montini, Investor and Owner of FDM Industrial Sales & Marketing SL, has spent nearly 40 years in industrial business. His experience leads him to a deceptively simple conclusion: “Judgment is not taught; it is built through exposure, repetition, mistakes and responsibility.”

That means executives should resist the temptation to optimize away every imperfect learning experience.

“We should not preserve obsolete tasks, but we must preserve those learning experiences.”

Montini describes this through his FDM Method, in which AI strengthens “Run”—the execution layer—while some of the time saved is reinvested in “Change,” including mentoring, customer contact, simulations and decision reviews.

He recommends a “human-first, AI-second” approach for junior employees: “Juniors form their own hypothesis, use AI to challenge it and then justify the final decision.”

The approach prevents AI from becoming an answer machine. Instead, it becomes a debate partner.

Montini also recommends capturing senior experts’ reasoning rather than merely documenting their finished work. This matters because future leaders need to understand not only what decision was made, but how experienced people arrived there.

“Automate friction, not experience, or today’s productivity gain will become tomorrow’s leadership deficit.”

How Leaders Can Build the Next Generation of Talent

• Redesign junior roles around judgment rather than task completion. Give early-career employees responsibility for questioning, applying and improving AI-generated work.
• Make verification a core technical skill. Have junior engineers stress-test AI output, investigate failure modes and audit systems for security, performance and reliability.
• Protect entry-level development programs even as automation expands. Use AI to make training more efficient without eliminating the experiences employees need to build competence.
• Turn AI output into raw material for analytical development. Require employees to fact-check recommendations, identify assumptions and compare AI predictions with actual business outcomes.
• Expose junior talent to business outcomes earlier. Move employees toward problem-solving, cross-functional collaboration, exception management and understanding why decisions matter.
• Reinvest AI-generated time savings in coaching. Use managers’ freed capacity for mentoring, feedback, communication practice and leadership development.
• Build a supervised judgment layer into early-career roles. Let AI handle first drafts while employees validate outputs, investigate exceptions and own recommendations.
• Track judgment, not just productivity. If employees are producing more while taking longer to reach trusted independent judgment, the organization may be borrowing against its future leadership bench.
• Create recurring leader touchpoints. Short, frequent interactions with experienced leaders can provide context, cultural understanding and apprenticeship that self-directed learning cannot.
• Preserve productive mistakes. Automate obsolete work, but keep opportunities for employees to form hypotheses, make decisions, receive feedback and learn from failure.

Build a Better First Rung

The biggest mistake leaders can make with AI is measuring success only by what it eliminates. Every task AI absorbs creates a choice: Let the time disappear into efficiency, or reinvest it in developing the people who will eventually lead the business. If young employees become more productive without gaining experience making decisions, navigating ambiguity and learning from mistakes, that is not an AI problem. It is a leadership-design problem.

The answer isn’t to preserve obsolete entry-level work. It’s to redesign it. Let AI handle execution while emerging talent challenges outputs, investigates failures, makes decisions and learns from experienced leaders. The first rung of the career ladder may be changing, but leaders have a chance to build a better one—one that uses AI not just to accelerate work, but to accelerate the development of future leaders.