Artificial intelligence is changing the nature of work. Tasks that once required substantial amounts of human time – writing reports, analyzing information, developing presentations, generating code, conducting research, and producing strategic recommendations – can increasingly be performed by AI. The implications of this transformation, however, extend beyond productivity. If machines can perform more of the intellectual work traditionally undertaken by professionals, what happens to the human capacity to think, learn, exercise judgment, and develop expertise?
This is not simply a technological question. It is also a question about how human beings learn and how professionals develop. In this respect, the emergence of artificial intelligence creates an unexpected connection between an ancient philosophical tradition and one of modern management’s most influential thinkers. Socrates challenged people to examine what they believed they knew and to develop their own capacity for reasoned judgment. Peter Drucker, writing about the emergence of the knowledge worker, argued that professionals have a responsibility to develop themselves, understand their strengths, acquire knowledge, and continually adapt their contributions as circumstances change.
Viewed through the lens of Management as a Liberal Art, these ideas offer a useful framework for understanding AI. The central issue is not whether artificial intelligence is inherently beneficial or harmful. Rather, it is whether we use it simply to produce more efficiently or whether we use it to expand our capacity to think, learn, and contribute.
The Socratic Problem of Knowing
Socrates did not leave behind a body of written work, but the Socratic tradition has endured because of its approach to knowledge. Learning was not simply a matter of receiving information or memorizing answers. It required examination. Through questioning, dialogue, and the testing of assumptions, individuals were encouraged to discover the limits of their own understanding (Plato, 1997; Plato, 2002).
This approach offers a useful perspective on the contemporary experience of generative AI. AI systems are remarkably capable of producing answers. They can explain concepts, summarize complex material, identify patterns, develop arguments, and propose solutions. They can produce a finished document in a fraction of the time it might previously have taken a professional to develop a first draft.
Yet the existence of an answer does not necessarily demonstrate understanding.
A professional who asks an AI system to explain a concept and accepts its response without further examination has acquired information. A professional who uses the response as the beginning of an inquiry – checking its assumptions, examining the evidence, considering alternative explanations, and identifying its limitations – is engaged in a fundamentally different activity. The former is primarily consumption of information; the latter is learning.
This distinction is becoming increasingly relevant as AI becomes embedded in knowledge work. A 2025 study of knowledge workers found that greater confidence in generative AI was associated with less reported critical-thinking effort (Lee et al., 2025). The researchers also found that AI changes the nature of critical thinking, shifting some effort away from information gathering and problem-solving toward verifying AI outputs, integrating responses, and overseeing the task.
The study’s findings do not suggest that AI necessarily diminishes human intelligence and the ability to think critically. Rather, it points to a more subtle issue: the way we use technology determines which forms of thinking we continue to practice.
The Socratic tradition therefore provides an important reminder. The value of an answer is not limited to whether it is correct. The process through which we arrive at, examine, and understand that answer is itself part of learning and developing critical thinking.
The Knowledge Worker
Peter Drucker approached the development of human capability from the perspective of the modern organization. As economies increasingly became dependent on knowledge workers, the primary contribution of many professionals shifted from physical production toward knowledge, expertise, judgment, and problem-solving.
This transformation created a new responsibility for the individual. In Managing Oneself, Drucker argued that professionals must understand their strengths, how they perform, how they learn, their values, and where they can make their greatest contribution (Drucker, 1999). Professional development could no longer be regarded as something that an organization simply provided. Individuals had to take responsibility for continually developing themselves.
Accordingly, AI makes this argument more relevant, not less.
If artificial intelligence can perform an increasing number of professional tasks, then professional value cannot rest solely on the ability to execute those tasks. The person who produces a standard report may discover that AI can produce it faster. The analyst who spends hours compiling information may find that AI can accomplish the same task in seconds. The writer who produces conventional business content may discover that AI can generate acceptable compositions almost instantaneously.
The implications are therefore significant because many of these activities have traditionally served two purposes: they produce an organizational output, but they also develop the person performing the work. Searching for information teaches us how to identify relevant evidence. Analyzing a problem teaches us how to recognize patterns. Writing forces us to organize and clarify our thinking. And revising an argument exposes weaknesses in our reasoning.
When AI performs these activities, it may improve productivity (completing a task much faster) while simultaneously reducing some of the opportunities through which professionals develop their capabilities.
This is where Drucker’s argument becomes particularly relevant. When the environment changes, professionals must reconsider not only how they perform their existing work, but also the nature of the contribution they should make.
The Productivity Paradox
This creates a paradox at the heart of AI-enabled work. Artificial intelligence can make us more productive precisely because it reduces the amount of effort required to perform certain tasks. Yet some of the effort removed from the task may have been contributing to the development of the person performing it.
Consider a young professional asked to prepare a market analysis. Traditionally, the individual might spend hours searching for information and data, comparing sources, organizing evidence, identifying patterns, writing an argument, and revising the final document. The process might be inefficient in some respects, but it is also educational. The professional learns about the market and the factors that influence it, by doing the work.
Now let us suppose that AI performs most of those activities. The professional receives a polished market analysis within minutes. The immediate productivity gain is obvious. The less obvious question concerns what has been learned in the process.
This is the challenge of cognitive offloading. When we transfer cognitive work to a machine, we may also transfer some of the opportunities through which we develop the capabilities required to perform that work. The danger is not that every task must remain difficult. Nor should technological progress be resisted simply because effort carries educational value. We cannot ignore that the history of technology is largely a history of removing unnecessary effort. The more important distinction and point is between removing unnecessary work and removing necessary learning.
That particular distinction should influence how organizations introduce AI into professional work. An organization that uses AI to eliminate repetitive administrative tasks may free employees to spend more time on analysis, creativity, relationships, and strategic thinking. An organization that uses AI primarily to eliminate the need for employees to understand their work may create a different outcome: professionals who become increasingly dependent upon systems they cannot adequately evaluate. The technology is the same. Yet, the developmental consequences are not.
Management as a Liberal Art
This is where Drucker’s broader conception of the management function becomes especially important. Drucker described management not simply as a technical discipline concerned with efficiency, but as both a social function and a liberal art. Management draws upon economics, psychology, history, philosophy, ethics, and other fields because organizations ultimately involve people, knowledge, values, and society. This perspective offers an important counterweight to a purely technological interpretation of AI.
If work is understood only as the production of outputs, then the most ‘efficient’ system will naturally be the one that produces those outputs with the least human effort. But if work is also a context in which people develop knowledge, judgment, responsibility, and purpose, then efficiency cannot be the only measure that matters.
The liberal-arts perspective embodied in Management as a Liberal Art consequently introduces a deeper consideration: organizations do not simply produce goods, services, and financial results. They also develop people. This makes the introduction of AI a management question rather than merely a technology or tool question.
A manager deciding how to deploy AI is also deciding what employees will continue to learn through their work. If AI writes the first draft, what will the employee learn from drafting? If AI performs the analysis, what will the analyst learn about the problem? If AI makes the recommendation, what opportunity remains for the professional to develop judgment?
These questions do not imply that AI should be excluded from such activities. They suggest that organizations should be intentional about preserving the developmental dimension of professional work.
From Socratic Dialogue to AI-Assisted Thinking
AI does not necessarily have to replace the intellectual process through which professionals develop. It can also become part of that process.
Consider the difference between asking AI for a finished answer and using it as an intellectual counterpart, a colleague of sorts. Instead of simply requesting a conclusion or finished product, a professional can use AI to test assumptions, identify weaknesses in an argument, generate alternative explanations, expose missing evidence, or present a competing perspective.
Under this approach, the resulting interaction begins to resemble a modern form of Socratic dialogue.
The value of the technology then lies not simply in the answer it produces but in the additional intellectual possibilities it yields. AI can help a professional examine a problem from multiple perspectives, challenge an initial interpretation, or identify questions that might otherwise have been overlooked. Used in this way, AI becomes less a substitute for thinking and more an instrument for thinking.
This distinction is critical because the goal of professional development should not be to preserve every traditional task. It should be to ensure that professionals continue to develop the capabilities that allow them to exercise judgment over increasingly complex problems. AI can accelerate learning when it expands inquiry. And it can undermine learning when it eliminates inquiry.
The Changing Nature of Professional Contribution
Drucker’s concept of self-development becomes particularly important in this environment. As technology changes the tasks associated with professional work, individuals must continually reassess what they know, what they can do, and where they can contribute most meaningfully.
The financial analyst may spend less time manipulating data and more time interpreting what the data means. The marketer may spend less time producing conventional content and more time understanding customers and developing strategy. The professor may spend less time transmitting information and more time facilitating inquiry, discussion, application, and judgment. The manager may spend less time producing reports and more time making decisions and developing people. In each case, the professional’s contribution moves upward from task execution toward interpretation, judgment, and responsibility.
This is consistent with Drucker’s broader understanding of innovation. Innovation changes the nature of what is possible, but organizations and individuals must determine how those possibilities should be used (Drucker, 1985). The professional’s responsibility is therefore not simply to acquire technological competence. It is to understand how technology changes the nature of his or her contribution.
This is perhaps the most important implication of AI for the knowledge worker. The challenge is not simply learning to use a new tool. It is continually redefining what it means to be useful when the tools themselves are becoming increasingly capable.
The Human Responsibility
The emergence of AI ultimately brings Socrates and Drucker into an unforeseen conversation. Socrates reminds us that knowledge requires examination. Information is not the same as understanding, and an answer is not necessarily the end of inquiry. Drucker on the other hand, reminds us that professionals have a responsibility to develop themselves and to continually reconsider the contribution they make as their environment changes. Management as a Liberal Art introduces another dimension: organizations must consider not only what technology allows them to accomplish, but also what kind of people and professionals their systems of work are creating.
Taken together, these perspectives suggest that the central challenge of AI is not simply automation. It is human development in an age of cognitive automation. AI will certainly perform more of the work that professionals once performed themselves. That is not necessarily a problem. Indeed, it may free people from routine activities and create opportunities for more meaningful forms of work.
But those opportunities will not emerge automatically, on their own. They will depend on how individuals and organizations choose to use the technology. AI can be used to produce answers without understanding, or it can be used to deepen inquiry and critical thinking. It can eliminate the effort involved in routine tasks while creating space for higher-order thinking, or it can gradually eliminate the very experiences through which professionals develop judgment.
The choice about how AI is used ultimately rests with us. The future professional may not be the person who knows how to perform every task independently. Nor will it necessarily be the person who knows how to delegate every task to AI. The more valuable professional may be the one who understands which tasks should be delegated, which should be learned, which should be questioned, and which ultimately require human judgment. That is a distinctly Socratic and Druckerian challenge.
AI can produce the work. It can accelerate analysis, generate ideas, and extend our capabilities. But it cannot assume responsibility for what we choose to learn, how carefully we think, or what contribution we ultimately make. The technology may change the work. But, the responsibility for developing the human being remains ours.
References
Drucker, P. F. (1985). Innovation and Entrepreneurship: Practice and Principles. Harper & Row.
Drucker, P. F. (1999). Managing Oneself. Harvard Business Review, 77(2), 64–74, 6(1).
Lee, H. P., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., & Wilson, N. (2025). The Impact of Generative AI On Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects from a Survey of Knowledge Workers. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (pp. 1-22).
Plato. (1997). Complete Works (J. M. Cooper, Ed.). Hackett Publishing.
Plato. (2002). Five Dialogues: Euthyphro, Apology, Crito, Meno, Phaedo (G. M. A. Grube, Trans.; J. M. Cooper, Rev. 2nd ed.). Hackett Publishing.
