Why Universities Must Teach Judgment: Prof. Pushpinder Singh Gill, Vice Chancellor of The Maharaja Bhupinder Singh Punjab Sports University, Patiala, examines how artificial intelligence is reshaping professional decision-making. He argues that universities must move beyond teaching answers and instead cultivate judgment, ethical responsibility, and reflective thinking—skills that AI cannot replace -- Editor (SamvadPatar)
The Difference Between Using AI and Understanding It
A hiring manager screens two hundred resumes using an AI tool. The system shortlists twelve candidates in seconds. One profile rises to the top—polished credentials, uninterrupted employment, a perfect keyword match. Another, ranked lower, shows gaps, career switches, uneven progression. One manager clicks “approve” and moves on. Another pauses. She asks what the algorithm rewards, whose career paths it treats as “normal,” what kinds of experience it quietly penalises, and what the team actually needs beyond efficiency. She wonders who will grow, who will adapt, and who will take responsibility when things go wrong. Both use the same technology. Only one has been trained to see beyond the output.
This difference is becoming common and consequential.
AI Shapes Decisions; But Universities Still Teach for Answers
Artificial intelligence is now embedded in everyday professional life. It screens candidates, flags risks, drafts responses, predicts outcomes, and ranks performance. It no longer merely supports decisions; it increasingly shapes them. Yet universities still educate students as if the central challenge ahead is finding answers. That assumption no longer holds. The real challenge now is deciding how much to trust an answer, when to question it, and when not to act on it at all.
This is the gap universities must address.
Where Knowledge Ends and Judgment Begins
Consider a medical intern on a night shift. An AI system recommends a standard course of treatment. One intern follows the protocol without hesitation. Another notices the patient’s history is incomplete and that the data reflects averages rather than individuals. She pauses. Both interns know the medicine. Only one recognises where knowledge ends and judgment begins.
Or consider a law graduate using AI to draft a contract. The document is legally sound and impressively efficient. One graduate submits it untouched. Another reads it closely and sees how certain clauses shift disproportionate risk to the weaker party. He asks how the agreement will function under stress, not just on paper. Again, the difference is not legal skill. It is judgment.
The Risk of Speed Without Reflection
These situations are already unfolding across professions. As AI systems grow faster and more confident, decisions that once took days are now made in minutes. Speed feels productive. It feels decisive. But speed without reflection does not improve outcomes. It produces faster mistakes.
Teaching Tool Use Is Not Enough
Teaching students how to use tools is no longer enough. Most students master that quickly. What they struggle with—and what universities rarely train them for—is knowing when to override a tool, when to slow down despite pressure, and when to accept uncertainty rather than force a decision.
Universities will not outrun AI’s answers. But they can still forge irreplaceable minds. That distinction matters because evidence now shows the cost of unreflective automation. A 2024 McKinsey report found that, in high-stakes contexts, AI-assisted decisions made without deliberate human reflection were 20 per cent more likely to be wrong. The failure was not technological. It was human—an absence of trained judgment at the point of decision.
Judgment Is Trained, Not Assumed
Some of the most effective training systems have already solved this problem. Military officer training, for instance, assumes students arrive with basic knowledge. Classrooms are not used to deliver information, but to use it. Learning progresses deliberately—from blackboard exercises to sand-model simulations, from controlled war-gaming to outdoor field exercises. Each stage introduces uncertainty and evolving constraints. Decisions are made, consequences observed, and then examined. The goal is not correctness, but readiness: the ability to think clearly, adapt, and take responsibility under pressure. This is not about hierarchy. It is about sequencing learning so judgment is trained, not assumed.
This insight reshapes how universities must think about teaching.
The Teacher’s New Role in the Age of AI
When AI can explain concepts and generate practice endlessly, teachers are no longer most valuable as explainers. Their real value lies in helping students examine their own thinking. A well-trained teacher no longer asks only, “Is this correct?” but presses further: Why do you trust this output? What assumptions are hidden? What information is missing? Who carries the risk if this decision fails?
These questions surface daily in professional life. They cannot be automated because they depend on context, responsibility, and human consequence.
From Lectures to Guided Thinking
If universities are serious about preparing students for this reality, the change must be visible in classrooms.
Classrooms organised around lectures belong to a time when explanations were scarce. Today, explanations are abundant. What is scarce is guided thinking. Classroom time must therefore be used for analysing situations, weighing trade-offs, and defending choices under challenge.
Students should arrive having gathered information—often with AI—and spend time making sense of it. Instead of problems with a single correct answer, they should confront scenarios with competing priorities and unclear outcomes. They must explain not just what they chose, but why they rejected other options. A student who can articulate uncertainty and revise a position under scrutiny is far better prepared for real work than one who produces a flawless answer without reflection.
Designing Classrooms for Disagreement and Reflection
Universities must instead create spaces where disagreement is expected and handled seriously. Faculty-led discussions on real dilemmas, interdisciplinary problem labs, and small forums without grades but with high expectations can change how students think. When students see teachers disagree respectfully and admit uncertainty, they learn that thinking is not about winning arguments, but understanding reality.
Curriculum Must Teach Consequence, Not Just Output
Curriculum design must follow through. A single course on ethics or AI awareness is not enough. Students must repeatedly practise decision-making with consequences within their disciplines. Engineering students should justify safety margins, not just efficiency. Business students should examine how incentives shape behaviour over time, not only quarterly results. Data science students must question data sources and bias—not merely accuracy. Across fields, students must learn to think about impact, not just output.
Assessment Must Reward Reasoning, Not Just Results
Assessment must reflect this shift. If grades reward final answers alone, AI will dominate. If students are evaluated on reasoning and response to feedback, AI becomes a support rather than a substitute.
Preparing Teachers to Train Judgment
Teacher preparation is the final—and most neglected—piece. Facilitating discussion and evaluating reasoning fairly require training and institutional backing. Universities must treat this as core professional development, not an optional interest.
The Harder Responsibility of Universities
Artificial intelligence will continue to improve at producing answers. Universities will not outrun it. Their relevance will depend on whether they accept a harder responsibility: shaping judgment. This means teachers who challenge assumptions, classrooms that slow thinking, and curricula that force students to confront consequences. In a world flooded with intelligence, education’s true task is not to add more answers, but to teach when not to trust them. That task now sits squarely with universities.
Dr. Pushpinder Singh Gill, VC (MBSPSU), Patiala
