Part Two: Relational Judgment and the Limits of Automation in Education
In part one of this series, I argued that education is not merely a system for delivering content, but a relational practice structured by trust, recognition, and authority. In this second part, I want to extend that argument by focusing on relational judgment: the educator’s capacity to interpret context, respond to uncertainty, and make ethically informed decisions in the moment. This becomes especially important in an age increasingly shaped by artificial intelligence, platformization, and data-driven educational design. Contemporary debates about education often emphasize efficiency, personalization, and scalability. These concerns are not insignificant. Yet they also reflect a broader tendency to view learning as a process that can be optimized through technical systems alone. Such assumptions sit uneasily with long-standing theories of learning that understand education as socially situated, ethically charged, and irreducibly interpretive. As Biesta has argued, education cannot be reduced to measurement or output; it involves a qualitative dimension that resists full proceduralizing (Biesta, 2015). Similarly, Schön’s account of the reflective practitioner emphasizes that professional action is not merely rule-based but depends on situational judgment and reflection-in-action (Schön, 1983). This is where the limits of automation become visible.
AI systems may assist with explanation, feedback, or content generation, but they cannot assume responsibility for the relational conditions of learning or discern whether a student’s silence signals confusion, resistance, reflection, or exclusion. They cannot determine when a learner needs challenge rather than reassurance or when a classroom interaction has become misaligned in ways that require repair. These are not peripheral concerns. They are central to the practice of education. Judgment in teaching is contextual, ethical, and relational. This distinction is important because much of education involves navigating ambiguity. Educators work with incomplete information, shifting emotional states, diverse learning histories, and unequal access to resources. Learning environments are therefore never fully transparent. In such contexts, the educator’s role is not only to convey knowledge but also to interpret relations. This requires awareness of what Lave and Wenger described as situated learning: the idea that knowledge is inseparable from the social and institutional contexts in which participation occurs (Lave & Wenger, 1991). It also requires recognizing, following Rogers, that meaningful learning depends on an environment marked by empathy, authenticity, and trust (Rogers, 1969).
These claims become even more significant when we consider the scale of contemporary education. In digital and AI-mediated environments, learners are increasingly mediated by systems that classify, recommend, and predict. Williamson has shown how data-driven infrastructures reshape education by turning learners into objects of measurement and intervention (Williamson, 2017). Zuboff’s account of surveillance helps clarify the broader logic behind this transformation: the extraction of behavioral data to predict and shape future conduct (Zuboff, 2019). When applied to education, this logic risks reducing students to data profiles and teachers to managers of algorithmically organized flows rather than interpreters of human development.
Against this backdrop, relational judgment becomes a crucial counter-practice. Judgment in education includes the capacity to read the learner as a situated subject, to balance structure with flexibility, and to recognize when a pedagogical intervention is necessary. In this sense, judgment is inseparable from relationality. The teacher must decide not only what is academically correct, but what is educationally appropriate for a particular learner at a particular moment. This is a form of practical wisdom that cannot be fully encoded into systems of prediction or automation.
The literature on therapeutic relationships clarifies how relational work depends on responsiveness to rupture, uncertainty, and asymmetry. Safran and Muran’s work on the therapeutic alliance shows that relational breakdowns are not exceptional failures but recurring moments that can be repaired through reflection and engagement (Safran & Muran, 2000). Education functions similarly. Misunderstanding, disengagement, or resistance are not always signs that teaching has failed. They may indicate that the pedagogical relationship requires recalibration. What matters is whether the educator can respond with interpretive attentiveness rather than procedural rigidity. This also requires a more robust account of power. Education is structured by asymmetry: teachers select material, set expectations, and evaluate performance. But asymmetry need not produce domination. Buber’s “I–Thou” relation complements this model because it offers a philosophical vocabulary for encountering the other as a subject rather than an object, even within unequal relationships (Buber, 1923/1970). Freire similarly insisted that education should be dialogical rather than banking-based and that students should be recognized as co-participants in the production of knowledge (Freire, 1970). These traditions insist that hierarchy must be made accountable to the conditions of learning.
This is especially relevant to education in the age of AI. If educational technology is to serve learning rather than merely efficiency, it must be embedded in pedagogical relationships capable of ethical interpretation. AI may assist with pattern recognition, but it cannot replace the judgment that comes from embodied presence, contextual knowledge, and responsibility toward the learner. Riess’s work on empathy reminds us that human understanding is not only cognitive but also communicative and affective, involving both verbal and nonverbal attunement (Riess, 2017). Teaching requires precisely this kind of attunement. The broader implication is that the future of education will depend not on whether AI can take over more teaching functions, but on whether educational institutions continue to value the forms of relational knowledge that technology cannot reproduce. If education is reduced to content delivery, the teacher’s role becomes increasingly narrow and mechanical. If, however, education is understood as a practice of relational judgment, the teacher remains indispensable—not as a mere transmitter of information, but as a responsible interpreter of human learning. In this sense, relational judgment is not an optional supplement to education. It is one of the conditions that make education possible in the first place.
References
- Biesta, G. (2015). Good Education in an Age of Measurement.
- Buber, M. (1923/1970). I and Thou.
- Freire, P. (1970). Pedagogy of the Oppressed.
- Lave, J., & Wenger, E. (1991). Situated Learning: Legitimate Peripheral Participation.
- Riess, H. (2017). The Empathy Effect.
- Rogers, C. R. (1969). Freedom to Learn.
- Safran, J. D., & Muran, J. C. (2000). Negotiating the Therapeutic Alliance.
- Schön, D. A. (1983). The Reflective Practitioner.
- Williamson, B. (2017). Big Data in Education.
- Zuboff, S. (2019). The Age of Surveillance Capitalism.



