Chapter 10: Artificial Intelligence

Part V — Mastery in the Digital Age

How AI Changes Expertise, Knowledge Creation, Automation, and the Human Meaning of Mastery

Artificial intelligence represents one of the most significant technological developments in the history of education, knowledge production, professional practice, and human learning. Earlier technological revolutions changed the tools through which people worked, communicated, calculated, travelled, and manufactured. Artificial intelligence is different in one particularly important respect: it increasingly interacts directly with activities traditionally associated with human cognition. It can classify information, generate language, summarize documents, recognize patterns, produce images, write computer code, assist with research, translate languages, analyse data, generate hypotheses, and participate in forms of reasoning. Generative artificial intelligence has therefore created a fundamental question for education: If machines can produce answers, what does it mean for a human being to become a master of knowledge?

This question reaches directly to the central theme of this book. Mastery has never simply meant possessing information. As discussed in earlier chapters, mastery involves knowledge, understanding, judgment, application, experience, ethical responsibility, professional identity, and the capacity to act appropriately in situations where no simple answer exists. Artificial intelligence may therefore weaken the importance of certain forms of information recall while simultaneously increasing the importance of interpretation, judgment, verification, creativity, contextual understanding, ethics, and human responsibility.

UNESCO's guidance on generative AI in education emphasizes precisely this tension. It describes the rapid emergence of generative AI and argues for a human-centred approach involving ethical, safe, equitable, meaningful, and pedagogically appropriate use. The OECD's 2026 Digital Education Outlook similarly identifies generative AI as reshaping education beyond merely classroom teaching and learning.

The central issue, therefore, is not whether artificial intelligence will replace education. The deeper question is whether education itself will be redesigned around a new relationship between human intelligence and machine intelligence.

17.1 From the Information Age to the Intelligence Age

The twentieth century was often described as the Information Age because computing, telecommunications, databases, television, satellites, and the internet dramatically expanded humanity's capacity to create, store, transmit, and retrieve information. The twenty-first century may increasingly be understood as an age in which the principal challenge is no longer access to information but the ability to interpret and use enormous quantities of information intelligently.

The internet transformed information availability. Search engines transformed information retrieval. Smartphones transformed information accessibility. Cloud computing transformed information storage and collaboration. Artificial intelligence is transforming information processing and generation.

This distinction is fundamental.

A student once had to search a library for several books, locate relevant chapters, take notes, organize information, and construct an essay. Search engines reduced the time required to locate information. Generative AI can now assist with summarization, organization, drafting, translation, coding, and other intellectual tasks.

The educational consequence is profound. If students can obtain an immediate explanation of a concept from an AI system, education cannot remain centred exclusively on memorizing information that machines can retrieve or reproduce.

This does not mean that knowledge has become unnecessary. Quite the opposite is true.

The more information becomes available, the more important knowledge structures become. A person cannot critically evaluate an AI-generated explanation of quantum physics, constitutional law, theology, economics, medicine, engineering, or organizational behaviour without possessing sufficient background knowledge to recognize errors, omissions, assumptions, and inappropriate conclusions.

Artificial intelligence therefore creates what may be called the paradox of knowledge:

The easier information becomes to obtain, the more important deep understanding becomes.

A person who knows nothing about a subject may be impressed by an AI-generated answer. A person who possesses mastery can interrogate that answer.

This distinction separates the consumer of AI output from the expert user of AI.

17.2 Artificial Intelligence and the Meaning of Expertise

Expertise has traditionally developed through prolonged exposure to a field, formal education, practice, mentorship, feedback, reflection, and increasingly sophisticated problem-solving. Research on expertise has demonstrated that experts develop rich mental representations that allow them to recognize patterns, prioritize information, anticipate problems, and select appropriate strategies (Chi et al., 1981; Ericsson & Lehmann, 1996).

Artificial intelligence challenges this model because a machine may demonstrate highly sophisticated performance without possessing human expertise in the traditional sense.

An AI system may generate an explanation of a legal principle without being a lawyer. It may generate a medical differential diagnosis without being a physician. It may produce a financial model without being a financial analyst. It may generate computer code without being a software engineer.

This creates a crucial distinction between performance and professional mastery.

A machine may perform a task.

A professional is accountable for the consequences of performing the task.

That distinction is central.

A physician who uses AI to assist diagnosis remains responsible for professional judgment. A university professor who uses AI to help construct educational material remains responsible for academic accuracy. A researcher who uses AI to analyse literature remains responsible for the integrity of the research. A corporate executive who relies on AI-generated analysis remains responsible for the organizational decision.

AI therefore changes expertise without necessarily eliminating expertise.

Instead, expertise may move upward.

The expert of the future may not necessarily be the person who can perform every routine intellectual task manually. The expert may increasingly be the person who understands which tasks should be delegated to AI, how they should be delegated, how the results should be evaluated, and when the machine should not be trusted.

This is a new form of mastery.

17.3 Human Expertise and Machine Capability

It is tempting to frame AI as a competition between humans and machines. Such a binary approach is often inadequate.

The more productive framework is human–machine complementarity.

Humans possess forms of intelligence associated with embodied experience, social relationships, moral responsibility, contextual understanding, emotional interpretation, cultural meaning, and lived experience. AI systems possess extraordinary capabilities for processing large volumes of information, detecting statistical patterns, generating content, and performing computational tasks.

Neither description should be romanticized. Humans make serious errors, while AI systems can produce sophisticated but incorrect outputs.

The important question is therefore:

What should humans do, what should machines do, and what should humans and machines do together?

A useful conceptual distinction can be made among four categories.

Human-dominant activities

These include:

• ethical responsibility;
• moral decision-making;
• interpersonal trust;
• pastoral and relational care;
• leadership accountability;
• culturally sensitive judgment;
• interpretation of human suffering;
• political and social responsibility;
• professional accountability.

Machine-dominant activities

These may include:

• processing enormous datasets;
• repetitive classification;
• rapid information retrieval;
• pattern recognition in structured datasets;
• high-volume computation;
• routine transcription;
• repetitive administrative tasks.

Collaborative activities

These include:

• research;
• writing;
• data analysis;
• curriculum development;
• software development;
• translation;
• strategic planning;
• literature exploration;
• scenario development.

Human-verification activities

These are tasks in which AI may produce an initial result but a qualified human must validate it.

Examples include:

• medical decisions;
• legal interpretation;
• academic assessment;
• scientific conclusions;
• financial decisions;
• engineering safety;
• research ethics;
• university policy.

This final category may become one of the most important domains of future professional education.

17.4 The Expert as Validator

In the traditional model of expertise, the expert produced the answer.

In an AI-supported environment, the expert may increasingly become the person who evaluates the answer.

This changes the cognitive requirements of professional education.

Consider a doctoral researcher.

Previously, the researcher might spend considerable time locating literature, organizing sources, constructing tables, proofreading text, and conducting preliminary data analysis. AI may assist with several of these activities.

But this does not eliminate the researcher's responsibility. Instead, the researcher must become better at determining:

• whether the source actually exists;
• whether the citation supports the claim;
• whether the statistical interpretation is correct;
• whether the research design is appropriate;
• whether the AI has introduced bias;
• whether the data have been misunderstood;
• whether the conclusion exceeds the evidence;
• whether confidential information has been exposed;
• whether ethical requirements have been followed.

Thus, AI can reduce the mechanical burden of research while increasing the importance of epistemological judgment.

The doctoral researcher of the future therefore needs not merely research skills but AI-mediated research literacy.

17.5 Artificial Intelligence and Knowledge Creation

For centuries, knowledge creation followed recognizable institutional pathways. Scholars observed phenomena, collected evidence, developed theories, conducted experiments, communicated findings, and subjected their work to criticism and peer review.

AI introduces a new participant into this knowledge ecosystem.

AI systems can search, synthesize, classify, generate hypotheses, identify patterns, compare large datasets, and assist with modelling. Machine learning systems can identify statistical relationships that may not be obvious to human observers.

This creates opportunities for scientific discovery.

At the same time, the generation of an output does not automatically constitute the creation of reliable knowledge.

Knowledge requires justification.

A machine-generated statement may be linguistically persuasive but epistemologically weak. Generative AI systems can produce what are commonly described as hallucinations: plausible-looking but unsupported or incorrect claims. Therefore, the future of scholarship cannot be based on the assumption that fluent output equals truthful knowledge.

The distinction between information, interpretation, evidence, and knowledge becomes more important than ever.

Information is data or content.

Interpretation gives information meaning.

Evidence provides grounds for accepting or rejecting a claim.

Knowledge is justified understanding developed through appropriate methods.

Wisdom involves knowing how that understanding should be used.

AI can assist with information processing and interpretation, but it does not remove the need for human epistemic responsibility.

17.6 AI and the Transformation of Research

Research universities are likely to experience some of the most significant effects of AI.

AI can assist researchers in:

• literature discovery;
• bibliographic organization;
• qualitative coding;
• statistical programming;
• simulation;
• language translation;
• transcription;
• image analysis;
• pattern recognition;
• research-question generation;
• data visualization;
• manuscript preparation.

The research process may therefore become faster.

But faster research is not necessarily better research.

The danger is that researchers may confuse efficiency with quality.

A doctoral student might use AI to generate a literature review containing hundreds of apparently relevant references. Yet if those references have not been checked individually, the apparent efficiency may conceal significant weaknesses.

Similarly, automated qualitative coding may identify linguistic patterns but misunderstand context, irony, cultural meaning, or the significance of silence.

Consequently, the future researcher needs methodological mastery plus AI literacy.

UNESCO's guidance specifically emphasizes the need for educational institutions to consider ethical validation, pedagogical design, data protection, inclusion, and human agency when integrating generative AI into education and research.

17.7 Automation and the Changing Nature of Work

Artificial intelligence is also changing the relationship between education and employment.

Historically, education prepared people for occupational roles. Industrialization created demand for engineers, accountants, managers, administrators, technicians, physicians, teachers, lawyers, and other professionals.

Automation has always changed occupations.

The difference today is that automation increasingly reaches into cognitive and administrative work.

The World Economic Forum's Future of Jobs Report 2025, based on responses from more than 1,000 employers representing more than 14 million workers across 55 economies, identifies technological change among the major forces expected to transform labour markets through 2030. The report also examines expected changes in skill requirements and training needs.

The educational consequence is that universities cannot simply prepare students for a static occupation.

They must prepare students for occupational transformation.

A graduate may enter a profession in which some of the tasks learned at university are subsequently automated. The enduring value of education must therefore include capabilities that enable graduates to learn, adapt, interpret, communicate, collaborate, and exercise judgment.

17.8 What AI Cannot Easily Replace

The discussion of AI should not be reduced to technological optimism or technological fear.

There are important dimensions of human activity that remain deeply connected to human relationships and responsibility.

Consider teaching.

An AI system can explain calculus. It can generate examples. It can produce quizzes. It can translate instructions. It can provide individualized explanations.

But teaching is more than information transmission.

A teacher recognizes when a student is confused, discouraged, embarrassed, enthusiastic, or disengaged. A teacher develops relationships. A teacher models intellectual character. A teacher creates community. A teacher makes judgments about when to challenge, encourage, correct, or listen.

Likewise, leadership is not simply the generation of strategic recommendations.

Leadership involves responsibility for people.

Healthcare is not merely diagnosis.

It involves trust, communication, compassion, consent, professional responsibility, and care.

Theology is not merely textual interpretation.

It involves questions of meaning, faith, community, suffering, hope, morality, and spiritual formation.

Education must therefore preserve the human dimensions of professional practice even while incorporating AI.

17.9 AI Literacy as a New Dimension of Mastery

Future mastery will increasingly include several forms of literacy.

Technical AI literacy

Professionals should understand, at an appropriate level:

• machine learning;
• generative AI;
• large language models;
• data;
• algorithms;
• model limitations;
• prompt design;
• automation;
• cybersecurity;
• privacy.

Critical AI literacy

Professionals must learn to question AI outputs.

They should ask:

• What evidence supports this?
• What assumptions were made?
• Could this contain bias?
• Is the source authentic?
• Is the answer appropriate for this context?
• What information might be missing?

Ethical AI literacy

Professionals must understand:

• privacy;
• consent;
• intellectual property;
• fairness;
• transparency;
• accountability;
• discrimination;
• responsible use.

Professional AI literacy

Professionals must know how AI changes their own discipline.

The AI literacy required by a physician differs from that required by a theologian, engineer, lawyer, teacher, researcher, accountant, or business executive.

Thus, AI literacy should become increasingly embedded within disciplines rather than treated solely as a separate computing subject.

17.10 The Future of Mastery

The emergence of AI does not necessarily mean the death of mastery.

It may instead produce a more sophisticated conception of mastery.

The nineteenth-century master demonstrated that he or she could perform difficult tasks.

The twentieth-century professional demonstrated specialized knowledge and technical competence.

The twenty-first-century master may increasingly demonstrate the ability to orchestrate human and machine intelligence responsibly.

The future master therefore needs:

Knowledge + Understanding + Experience + Judgment + AI Literacy + Ethics + Creativity + Responsibility.

This represents an expansion rather than a reduction of mastery.

The ultimate objective of education should not be to create humans who compete with machines at being machines.

It should be to develop human beings who can use machines while remaining deeply human.

17.11 Conclusion

Artificial intelligence forces universities to reconsider what they mean by knowledge, expertise, learning, assessment, research, and professional competence.

The central educational challenge is not simply teaching students how to use AI. It is teaching them how to think when AI is available.

The future university must therefore protect deep learning while embracing appropriate technology. Students must still learn foundational knowledge. They must still read deeply. They must still write. They must still conduct research. They must still debate. They must still experience failure and correction. But they will increasingly perform these activities in partnership with intelligent digital systems.

Mastery in the AI age will not mean knowing everything.

It will mean knowing what matters, what can be delegated, what must be verified, what cannot be automated, and when human judgment must remain decisive.

That may become the defining intellectual challenge of the twenty-first-century university.