The Erosion of Expertise: Why Automating the “Labor of Thinking” Threatens Professional Mastery

In the contemporary landscape of professional development, we often treat work as a series of deliverables—artifacts to be produced, forms to be filled, and boxes to be checked. However, a growing chorus of experts argues that this perspective is fundamentally flawed. Instead, we must view our work as a "practice": a complex, interrelated system of skills, knowledge, attitudes, and—most crucially—habits of mind.

These elements of a practice are what allow professionals to navigate the rhetorical and technical demands of their fields. Yet, as generative AI moves from the fringes into the core of professional life, we face an existential risk. By prioritizing efficiency through "cognitive offloading," we are inadvertently dismantling the very mechanisms of expertise.

The Anatomy of a Practice

At the heart of professional competence lies the "habit of mind"—the internal cognitive architecture that governs how we plan, research, draft, revise, and refine our work. In education, these habits are notoriously difficult to assess. Two students may submit identical essays, but the path taken to reach that result can be vastly different. One student may have followed a rigid template, effectively "filling in the blanks," while the other engaged in a deep, iterative process of discovery and critical analysis. Only the latter is truly developing a robust, adaptable practice.

This distinction holds true across nearly every high-level profession. Whether we are discussing doctors, lawyers, musicians, or engineers, the "practice" is not merely the final artifact; it is the process of synthesis that creates it. As I have argued in my previous work, such as Why They Can’t Write: Killing the Five-Paragraph Essay and Other Necessities, writers share more with chefs than one might expect. Both require a blend of technical skill, sensory awareness, and a cultivated, reflexive habit of mind that evolves only through the crucible of experience.

The Mirage of Efficiency: The Case of Medical Scribes

Nowhere is the tension between efficiency and expertise more visible than in the medical profession. A recent, highly regarded piece by Dr. Helen Ouyang in The New York Times Magazine serves as a sobering case study. Dr. Ouyang explores the evolution of the clinical note—the "chart"—which has historically served as a space for physicians to process patient interactions, synthesize diagnostics, and formulate a care plan.

The structure of this note, formalized in the 1960s as the "SOAP" method (Subjective, Objective, Assessment, Plan), was never intended to be merely clerical. It was a tool for cognitive labor. However, the rise of AI scribes—technologies capable of recording and summarizing patient interactions in real-time—has promised a utopian future: the removal of the administrative burden.

Initially, for practitioners like Dr. Ouyang, the AI scribe seemed like a triumph. It promised to return time to the physician, allowing them to focus on "the important work." Yet, a profound sense of "discomfit" set in as she realized that by outsourcing the drafting of her notes, she had bypassed the very cognitive friction required to process her patient’s condition. She had effectively removed the opportunity to think through the case, settling instead for a summary that had been "already made for me."

Chronology of a Shifting Paradigm

To understand how we reached this point, we must look at the timeline of cognitive automation in the workplace:

• 19th Century: The formalization of the medical chart as a vital, written record of clinical observation and institutional memory.
• 1960s: Dr. Lawrence Reed codifies the SOAP note structure at Case Western Reserve University, standardizing the diagnostic reasoning process for physicians across the United States.
• 2024: Generative AI scribes move from experimental pilot programs to widespread, public-facing adoption in clinics and hospitals.
• 2025: The publication of More Than Words, highlighting the early signs of professional disruption caused by automated linguistic tools.
• 2026: Leading medical institutions, including Johns Hopkins, begin moving toward policies where third-year medical students are no longer required to draft their own notes, relying entirely on AI-generated summaries.

The "Invisible" Cost of Automation

The danger of this shift is that it is often presented as a net positive. When Ben Gooch, a general practitioner in the U.K., reviewed an AI-generated note for a patient he had seen six weeks prior, he noted that the summary was technically accurate and comprehensive. There were no factual errors. Yet, he remarked, "I did not recognize it."

Gooch’s experience highlights a critical failure in the logic of "cognitive offloading." By delegating the writing to an algorithm, the practitioner loses the connection to the patient that is forged through the act of reflective documentation. Writing is not merely a method of transmission; it is a method of thinking. When we offload the writing, we offload the thinking.

This leads to a paradox: as our habits of mind become more efficient, they become more invisible. When a practitioner is well-trained, their skills function almost subconsciously. But if we automate the process before that mastery is achieved—as is now happening with medical students—we never build the muscle memory required to exercise judgment in the first place.

Implications for Future Expertise

The implications for higher education and professional training are profound. If we eliminate the "labor" of the practice in favor of the "artifact" of the result, we are essentially training professionals who are capable of managing AI, but incapable of functioning without it.

1. The Erosion of Critical Thinking

If students, doctors, and lawyers are never forced to struggle with the synthesis of information, they will lose the ability to spot errors in the AI’s output. Expertise is built on the foundation of having "done the work" repeatedly. Without that foundational experience, the next generation of professionals will be "automation-dependent" rather than "automation-empowered."

2. Efficiency vs. Quality

In both health care and education, efficiency is frequently conflated with quality. However, efficiency is merely a metric of speed, not of value. A doctor who saves ten minutes on documentation by using an AI scribe but loses the deeper cognitive engagement with the patient’s diagnostic history is not necessarily providing a better, or safer, standard of care.

3. The Institutional Responsibility

Institutions must pivot from a focus on output to a focus on the process of inquiry. In faculty development workshops, I often challenge experts to unpack the "when" and "how" of their own learning. We need to ask if our current technological adoptions are helping students develop their practices or if they are simply helping them bypass the struggle of learning.

Conclusion: Reclaiming the Labor of Thinking

The allure of AI is, in many ways, an extension of the broader societal drive toward convenience. But if the future of work is to involve the systematic sidelining of the most important human cognitive contributions, we must ask ourselves: what are we doing here?

We are at a crossroads. We can choose to view AI as a partner in a human-centric practice, or we can allow it to replace the very cognitive habits that define our expertise. The path forward requires a deliberate, skeptical, and reflective approach. We must protect the "labor" of our professions—the messy, time-consuming, and difficult work of thinking, writing, and synthesizing—because it is in that very labor that our humanity, and our professional value, resides.

The goal of education and professional development should not be to make the work "fast and easy." It should be to make it deep, meaningful, and deeply personal. We must ensure that the tools we adopt serve to enhance our habits of mind, not replace them. If we lose the ability to think through our own work, we lose the ability to own the outcomes of our labor—and that is a price we cannot afford to pay.