Every executive leadership team is talking about it. The memos have been sent, task forces have been formed, the town halls have been held, and the mandates are clear: Everyone must embrace artificial intelligence. Leaders are eager to solve the crisis by closing the skills gap, reskilling the workforce, and racing to teach a generation of employees how to prompt-engineer their way to productivity.

They are unfortunately focusing on the wrong crisis.

The real crisis was there all along, hidden in plain sight. It’s not an AI problem; it’s an organizational and philosophical one. Generative AI doesn’t just automate tasks. It performs a brutal, unblinking audit of a company’s work design, its training programs, and the very value of its human capital. The story isn’t that robots are taking jobs. The story is that AI is revealing how many of those jobs were never designed to harness human potential in the first place.

For decades, organizations were built on a foundation of low-value, repetitive tasks, politely calling it “paying your dues.” Now, AI is holding a mirror up to organizational structures, and the reflection reveals not a need for reskilling but a long-overdue reckoning with what constitutes “work.”

One only needs to look at any entry- to mid-level job description. The verbs tell the story: compile, review, summarize, track, format, update, coordinate. These aren’t strategic activities. They’re biological API calls—humans serving as middleware between disconnected enterprise systems.

Employers have come to accept a system where people spend 60% of their time on repetitive work that exists primarily to compensate for fragmented technology and bloated processes.

This isn’t new. When the internet arrived, companies layered digital tools onto analog processes, creating more bureaucracy. When cloud computing emerged, they moved servers to AWS but kept the same approval workflows that require printing and processing of paper documents. Each technological wave exposed the same misalignment between tools and organizational design.

Despite trillions of dollars invested in digital transformation, cloud computing, automation, and enterprise software over the past 15 years, white-collar productivity growth has effectively flatlined over the same period. Companies gave their workforces Slack, Zoom, SharePoint, Workday, and Salesforce, yet output per hour barely budged.

Why? Because companies didn’t use technology to eliminate low-value work; they used it to accelerate low-value activity. The real shock isn’t that AI can do a lot of white-collar work.

The real shock is realizing how much of a company’s headcount was focused on work that never should have required a human in the first place.

THE MYTH OF “GRUNT WORK” AS APPRENTICESHIP

For generations, leaders justified the corporate rite of passage known as “paying your dues.” They told themselves, and their new hires, that years of mind-numbing, repetitive tasks were a necessary apprenticeship—a crucible that forged judgment and built character. Senior executives have deep nostalgia for the “junior associate” grind built on these beliefs:

• You “learn the business” by doing reconciliations.
• You “build judgment” by drafting versions 12 through 27 of the same slide.
• You “earn your stripes” by staying late to assemble reports no one will remember reading.

In reality, this was a convenient fiction. It was hazing by spreadsheet.

Grunt work was never a deliberate or effective training strategy. It was a symptom of organizational inertia and a failure to properly design roles and workflows. Everyone knew this work was questionable, but it was easier to hire another body than fix the underlying problem.

There is always great enthusiasm to hire more staff. Leaders are eager to build large organizations because at many companies, one’s gravitas, seniority, and standing are measured by team size. Departmental budgets are often allocated based on headcount. While leaders at early-stage companies are incentivized to run lean operations, leaders at large enterprises experience the perverse incentives to constantly grow their teams—because steady hiring is perceived to correlate with company health and performance. However, when layoffs and workforce reductions take place, departments often shed 5%, 10%, or even 15% of their workforce, yet somehow do not experience even the slightest hint of reduced output or productivity.

The data backs up what many employees quietly knew long before large language models entered the boardroom. Multiple studies have found that knowledge workers spend more than half their day on repetitive, low-value tasks (email, routine reporting, formatting documents, chasing status updates). A 2025 study from Eagle Hill Consulting (“Are Employee Ideas the Hidden Key to Operational Efficiency?”) found that 68% of employees regularly spend time on low-value, inefficient tasks. Other research, including a 2025 study from Voucher Cloud, paint a bleaker picture, suggesting that in an eight-hour day, the average worker is “truly productive” for less than three hours. The rest of the time is consumed by coordination, duplication, and work activities that add little to no value to the enterprise.

AI AS AN X-RAY FOR JOB DESIGN AND THE BUSYWORK ECONOMY

Companies normalized inefficiency and built entire career paths around it. Each new system, process, regulation, or compliance requirement added a little bit of manual glue work. A spreadsheet here, a workaround there, an extra approval email, a document printed in triplicate, a status meeting “just to align” all slowly contributed to building a large “hidden economy” of people whose main job is compensating for bad processes and fragmented systems.

Entry-level roles have become manual interfaces between broken systems—endless cycles of copying data from Excel to PowerPoint, summarizing documents, sending emails, and tracking information that should have been automated years ago. Companies installed a human patch for bad process design.

Employees are well aware of their roles as cogs in an inefficient machine. According to a 2024 Gallup article, employee engagement in the U.S. has hit an 11-year low, with “active disengagement” rising. The workforce didn’t check out because of AI. They checked out because companies had built jobs that treated humans like robots long before ChatGPT arrived.

The single biggest productivity prize from AI may not be deploying virtual assistants to glamorous use cases. It may be eliminating this invisible tax of busywork—the unexamined patchwork of low-value tasks that has quietly claimed so much of a company’s time and energy.

THE REAL WORKFORCE CRISIS: A FAILURE OF IMAGINATION, NOT TECHNOLOGY

The real crisis isn’t about teaching employees to write better prompts. It’s about a fundamental failure of leadership. The current panic over AI training is a distraction from the much harder work of rethinking job design, career paths, and the very definition of value creation. A recent Forbes survey (“The Forbes Research 2025 CxO Growth Survey”) found that while 93% of companies plan to increase AI investment, only 49% of their HR leaders are prioritizing training employees in AI and data analysis.