Organizations are investing heavily in artificial intelligence (AI), but many are approaching implementation with the flawed assumption that AI replaces jobs. According to Kason Morris, Head of Future of Work and Talent Intelligence Strategy at Merck, that mindset misses where value is actually created and risks weakening the very capabilities that businesses will need in the future.
“AI does not automate jobs. It automates tasks,” Morris says. “The job title itself is really a container. Inside that container are the tasks, the decisions, the relationships, the handoffs, and the different levels of judgment.” Rather than treating AI as a tool for reducing headcount, Morris argues leaders should view it as an opportunity to redesign how work happens. Organizations that understand work at the task level will make smarter automation decisions while strengthening human capability. Those that don’t may create efficiencies today at the expense of expertise tomorrow.
Looking Beyond the Job Title
For decades, organizations have organized work around job descriptions and organizational charts. Morris believes AI requires leaders to rethink that model entirely. “When leaders can see that mix, the conversation about what to do with AI becomes a bit more precise,” he says.
Instead of asking whether a role should continue to exist, leaders should examine the individual activities within it. Which tasks create value? Which create friction? Which can AI perform effectively? And perhaps most importantly, which responsibilities develop human judgment?
Those questions shift AI from being a workforce reduction exercise to an operating model discussion. They also bring together leaders from across the business. Technology teams understand the tools, business leaders understand where value is created, and workforce leaders understand how people develop capability. Only when those perspectives come together can organizations redesign work responsibly, changing the objective to “moving from reducing labor to redesigning execution.”
Why Execution Visibility Matters
Central to Morris’ thinking is a discipline he calls Execution Visibility: a framework that makes the real mechanics of work visible before organizations automate them. “Execution Visibility is not about documenting every task for the sake of documentation,” he says. “It’s about giving leaders enough clarity to redesign work without unintentionally weakening the organization.”
The speed of AI transformation has made that clarity increasingly important. Automation decisions are being made faster than ever, but without understanding where judgment, expertise, and decision-making actually occur, organizations risk removing capabilities they never intended to lose.
Execution Visibility helps leaders identify where value is created, where human judgment enters the system, and which experiences build expertise over time. Rather than asking only what AI is technically capable of doing, Morris encourages organizations to ask a more strategic question: “What kind of organization are our decisions around AI creating?”
Designing Work That Builds Human Judgment
One of the most common mistakes organizations make is assuming that if AI can complete a task, it should, but technical capability and organizational wisdom are not the same thing. When evaluating work, he considers factors such as predictability, context, accountability, and the consequences of error. Just as important, however, is understanding what employees gain by performing that work.
“What capability does the human build in this work?” Many routine tasks appear inefficient when measured only by productivity, but they often develop pattern recognition, commercial understanding, professional judgment, and trust, all of which prepare employees for increasingly complex decisions later in their careers.
“It’s not simply about: ‘Do we automate or retain?'” Morris says. “It’s about designing the right human-AI relationship around the work.” In practice, that means automating highly predictable, repeatable work, while deliberately protecting or replacing experiences that help people develop stronger decision-making capabilities.
The Expertise Paradox
Perhaps nowhere is this challenge more visible than in entry-level hiring. As generative AI increasingly performs foundational work, many organizations are reducing junior roles. Morris believes that approach creates an unintended consequence, which he describes as the “expertise paradox.”
Entry-level work has always served two purposes. It produces business outcomes, while exposing employees to the experiences that eventually develop expertise. Junior professionals learn by preparing analyses, reviewing cases, observing experienced colleagues, making lower-risk decisions, and receiving feedback. While these activities may appear routine, they build the judgment organizations rely on years later. “If AI is performing those foundational tasks while senior experts continue to make difficult decisions, the organization may become more productive in the near term but they’re weakening their pipeline for future experts and future judgment.”
Rather than preserving legacy work for nostalgia, Morris believes organizations must create new development pathways through apprenticeships, simulations, guided AI experiences, and structured decision reviews. “We can’t just remove that bottom rung of the ladder without designing another way for people to climb.”
Every Automation Decision Is a Capability Decision
Looking toward 2030, Morris believes the organizations that pull ahead will be those that used AI to strengthen human capability rather than consume it. Automation decisions made today determine where judgment remains human, how future talent develops expertise, and whether productivity gains become opportunities for innovation or simply reductions in headcount.
“If every productivity gain becomes a headcount reduction, an organization may become smaller and leaner but not necessarily smarter,” he says. Instead, the capacity AI creates can be redirected toward customer understanding, experimentation, mentorship, innovation, and solving increasingly complex problems. “The dividing line won’t simply be between companies that have adopted AI versus those who didn’t. It will be between organizations that use AI to compound human judgment and create that depth, and organizations that use AI just to consume the capability they’ve inherited.”
For Morris, the future of AI is ultimately less about technology than leadership. Organizations that take the time to understand how work creates value before redesigning it will be far better positioned to thrive. “Every automation decision is also a capability decision. The organizations that recognize that first will build an advantage AI alone cannot create.” Follow Kason Morris on LinkedIn or visit his website.
