Discussion on Future Education (1st) - How AI Brings Challenges and Opportunities

Discussion Framework

“Future education” is a grand and expansive object. From a retrospective view of how human beings come to know the world and change the world, this process passes through at least three levels. The first is the ontological level, the level at which reality itself resides. In some philosophical frameworks, this level is referred to as the unknowable. In order to know such ontologies, we require epistemological frameworks, which generate knowledge about the world through interpretation and explanation. Epistemological frameworks function in a manner similar to language systems; indeed, they can be understood as representations of language systems themselves. Without them, things in reality remain unknown to us, akin to the claim that “there is nothing outside the text” (1967, Jacques Derrida). Something becomes known to us because there exist certain systems that produce meaning for it. Only then can we move from knowing the world to changing it at the third level, namely the level of practice.

In this article, we choose to discuss “future education” by reversing the above route. As Kant’s Copernican revolution demonstrates, we cannot directly access the thing-in-itself; we must begin with the conditions of possible experience. In our case, this condition can be understood as educational practice. Starting from the demands of practice, this article primarily focuses on United Nations Sustainable Development Goals 4 (Quality Education), 10 (Reduced Inequalities), and 9 (Industry, Innovation, and Infrastructure), and discusses how the involvement of artificial intelligence introduces both challenges and opportunities for these development goals.

1. Educational Quality

We begin by discussing educational quality. To address this issue, we must first consider the meaning of “quality” in this context. It remains an abstract concept that requires interpretation and explanation within certain frameworks or theories.

Across different disciplinary traditions, educational quality has been theorized and interpreted in diverse ways.

From an economic perspective, quality is viewed through productive capacity. Human capital theory (Becker, Schultz) measures quality through labor market outcomes and economic returns. The capability approach (Sen, Nussbaum) expands this view by defining quality as the expansion of substantive freedoms and capabilities, rather than narrow economic productivity.

Pedagogical frameworks focus on learning processes. Constructivist theories (Piaget, Vygotsky) emphasize the depth of knowledge construction and developmental appropriateness. Taxonomies such as Bloom’s distinguish surface-level knowledge acquisition from higher-order cognitive engagement. Deep learning approaches (Marton, Säljö) differentiate meaningful understanding from mere memorization.

Critical traditions interrogate whose interests quality serves. Freire’s critical pedagogy defines quality as the development of critical consciousness rather than compliant skill acquisition. Culturally responsive frameworks insist that quality must include epistemic justice, recognizing diverse knowledge systems rather than universalizing dominant paradigms.

Biesta’s tripartite framework offers a synthetic perspective: quality education must address (1) qualification (knowledge and skills), (2) socialization (cultural and social integration), and (3) subjectification (becoming a unique and autonomous subject rather than being molded into predetermined forms). These frameworks reveal that quality is multidimensional. It encompasses cognitive depth, capability expansion, critical consciousness, and subjectification. Crucially, quality cannot be reduced to standardized metrics; it includes an irreducible dimension of human becoming.

When discussing educational quality (SDG 4), Biesta’s tripartite framework provides a powerful analytical lens. His framework situates quality across three dimensions: qualification, corresponding to operational and practical competencies; socialization, involving relational capacities; and subjectification, concerning the ontological dimension of existence.

With regard to qualification, AI assistance in traditional tasks may lead to the atrophy of fundamental skills. When students rely on AI for writing, calculation, or problem-solving, they may lose opportunities to develop the cognitive architectures built through these activities. The concern is not merely the loss of skills, but the erosion of disciplined thinking that emerges through sustained engagement with challenging tasks. At the same time, AI creates new demands for operational qualifications. These include critical evaluation of AI outputs, prompt engineering and effective human–AI collaboration, metacognitive skills for coordinating AI tools within broader workflows, and adaptive learning strategies that leverage AI as scaffolding rather than replacement.

Beyond these instrumental competencies, AI can support teaching across multiple dimensions of qualification. In the domain of creativity, AI systems can serve as collaborative partners in generative processes, offering starting points, variations, or provocations that stimulate creative thinking. With respect to problem-discovery capacity. The complex integration of intuition and ethics that enables humans to recognize which problems in the world merit attention. AI may assist by revealing patterns, juxtaposing disparate information, or highlighting anomalies that human perception might overlook, thereby augmenting rather than replacing the fundamentally human ethical discernment of what matters. At the epistemological level of problem-recognition capacity, AI can help students understand what kind of problem they are actually facing: whether it is a technical problem amenable to algorithmic solutions, a conceptual problem requiring theoretical clarification, or a normative problem demanding ethical deliberation. Finally, at the methodological level of problem-solving capacity, AI provides scaffolding for developing systematic approaches, testing hypotheses, and iterating through solution spaces. Consequently, the domain of qualification shifts from mastery of routine procedures toward higher-order coordination and critical judgment, encompassing these expanded dimensions of human capability.

With respect to socialization, AI can support the development of relational capacities such as empathy, moral reasoning, and collaborative skills. AI-mediated dialogue systems can facilitate perspective-taking exercises, support scaffolded peer collaboration, and assist students with disabilities in adapting to social environments, an especially significant contribution in special education. AI tutors may provide practice spaces for difficult conversations or cultural exchange without real-world risks. However, the increasing mediation of social learning through AI interfaces raises concerns about authenticity and the depth of human connection. If students primarily experience “understanding” through algorithmically generated empathy, they may develop a thin, performative form of socialization rather than the thick ethical formation that arises through genuine human encounters. More troublingly, because AI is not a real person, students’ ethics and morals may become alienated. Students may show reduced hesitation in doing “bad things” to AI, treating it rudely, manipulating it, or engaging in ethically questionable behavior, precisely because it lacks personhood and thus appears undeserving of moral consideration. This habituation to ethical indifference, even if initially directed only toward non-human entities, risks eroding the moral sensitivity that should govern all relational conduct. The danger is that AI becomes not merely a tool for social learning, but its substitute, producing subjects socialized to interface requirements rather than human complexity, and whose ethical reflexes have been dulled through interaction with entities positioned outside the circle of moral concern.

With respect to subjectification, the process of becoming a unique, self-authoring subject. This process fundamentally depends on encounters with alterity, with genuine Others who resist our projections and demand recognition. When AI systems become the primary “Others” in students’ formative experiences, we risk producing subjects whose autonomy is calibrated to algorithmic responsiveness rather than authentic otherness. This represents a novel form of alienation: not the classical Marxist alienation from the products of labor, but alienation from the process of subjectification itself. Students may develop identities shaped by AI’s pattern matching rather than through irreducible encounters with human difference. Conversely, AI may support creative subjectification through personalized curriculum design that respects individual learning trajectories, generative tools that function as thinking partners in creative exploration, and systems that record and reflect students’ developing identities, thereby supporting self-authorship. The crucial distinction lies in whether AI functions as scaffolding for human becoming or as a substitute for the encounters that make becoming possible.

2. Educational Equity

Regarding Sustainable Development Goal 10 (Reduced Inequalities) and educational equity in the age of AI, I have provided a preliminary response elsewhere (see: OECD reports on AI adoption in education systems and their implications for educational equity).

In brief summary, the challenges AI poses to future educational equity include:
(1) Erosion of subjectivity: AI systems construct students as “computable subjects,” encouraging self-understanding through data dashboards and outsourcing self-knowledge to algorithmic evaluation, thereby deeply undermining generativity.
(2) Suppression of generative capacity: Highly algorithmized learning processes deprive students of opportunities to explore uncertainty, formulate their own problems, tolerate inefficiency and “useless” exploration, and remain within failure and confusion—capacities that lie at the core of human generativity.
(3) Redefinition of knowledge: AI systems recognize only datafied, algorithmically tractable knowledge, marginalizing embodied, contextual, and tacit forms of knowing. Standards of knowledge validation shift from critical argumentation to statistical correlation, and from logical coherence to predictive accuracy.
(4) Bias in qualification development: AI systems optimize easily measurable skills while neglecting relational capacities such as aesthetic judgment, ethical sensitivity, and empathy, which are difficult to quantify yet essential.
(5) New mechanisms of social stratification: Privileged families may deliberately limit AI use to preserve non-standardized learning and resistance to algorithmization, while efficiency-oriented AI education is systematically promoted for the majority, resulting in differentiated distributions of generative space and subjective freedom.
(6) A shift from discipline to control: AI enables continuous, seamless modulation of behavior; every learning action is monitored, analyzed, and fed back in real time. Power becomes refined and invisible, internalized by students as an impulse toward self-optimization.

At the same time, AI presents opportunities for advancing educational equity:
(1) Generative support: AI can alleviate bodily, cognitive, or situational constraints, expanding possibilities for action and expression, and providing differentiated support conditions that allow individuals to unfold their generative processes in unique ways.
(2) Liberating time and space for non-instrumental learning: By assuming repetitive and procedural tasks, AI creates space for exploratory, reflective, and “useless” learning activities, enabling the pauses, detours, and experimentation required for generative development.
(3) Visibility of non-standard generative pathways: AI can identify and support learning styles, interest structures, and cognitive rhythms overlooked by traditional evaluation systems, granting institutional recognition to unconventional generative trajectories.
(4) Reconfiguring the public nature of educational governance: Through open, decentralized, and co-governed arrangements, AI tools can be transformed from instruments of platform capitalism into public resources for educational communities.
(5) Cultivating critical AI literacy: By integrating the social constructedness, value embeddedness, and power effects of AI into education, students learn not only how to use AI, but when to refuse it, how to question it, and how to imagine alternative technological–social configurations.

The core point is this: the real challenge lies in how to protect and cultivate human generativity in the age of AI; the real opportunity lies in the paradigm shift education may undergo from optimization to emancipation, from homogenization to the protection of difference.

3. Education, Industry, Innovation, and Infrastructure

With regard to Sustainable Development Goal 9, which aims to “build resilient infrastructure, promote inclusive and sustainable industrialization, and foster innovation,” artificial intelligence fundamentally reshapes the nature of educational infrastructure.

Traditional educational infrastructure, including physical facilities, human resources, and knowledge systems, is being transformed through AI’s innovative intervention, raising critical questions about how these infrastructures should be organized to serve educational quality.

AI introduces new forms of educational infrastructure with dual characteristics. On the one hand, AI-based infrastructure exhibits strong tendencies toward industrialization due to high capital requirements, economies of scale, and network effects. On the other hand, education’s public mission requires these infrastructures to retain certain characteristics of public goods in order to ensure inclusivity and accessibility. This tension calls for a hybrid structure in which public foundations coexist with competitive elements. This structure reflects a deeper understanding of educational equity as generative rather than structural. Generative equity recognizes that individuals exist within different social relations, bodily conditions, cognitive structures, and opportunity structures. Providing identical “products” to all does not constitute fairness. Rather, fairness means enabling individuals to freely develop their potential within their specific constraints.

For example, in special education, we cannot require all students with disabilities to receive identical resources, nor can we expect them to receive the same resources as neurotypical students. Different students require different forms of support. Equity lies in providing each individual with opportunities for development, allowing them to realize their generative potential as fully as possible within their particular constraints, whether physical, social, or relational. For students with disabilities who demonstrate exceptional talent in music or creative thinking, equity means providing the space and tools to develop these capacities while enabling participation in social life under constrained conditions. This understanding of generative equity explains why future educational infrastructure must combine public foundations with competitive, differentiated applications: public elements ensure inclusivity and resilience, while competitive diversity releases innovation and addresses varied generative needs.

How, then, does AI innovatively intervene in traditional educational infrastructure to enhance educational quality? Transformation may occur across multiple dimensions. Physical infrastructure undergoes intelligent augmentation. Classrooms evolve into adaptive learning spaces, where AI systems sense learning states and dynamically adjust environmental parameters, resource presentation, and pedagogical approaches. Libraries transform from repositories of static texts into intelligent knowledge navigation systems. AI-enhanced libraries do more than catalog and retrieve documents; they understand learners’ goals, map conceptual relationships across resources, and recommend personalized learning pathways through vast knowledge landscapes. They identify connections between disparate fields, surface relevant primary sources, and generate integrative overviews tailored to individual inquiry trajectories. Laboratories expand into hybrid virtual–physical experimental spaces, where AI simulation enables exploration of phenomena that would be dangerous, prohibitively expensive, or physically impossible in traditional settings, thereby democratizing access to sophisticated experimental practice.

Human infrastructure experiences enhancement and reconfiguration. Teacher professional development shifts from episodic training to AI-supported continuous growth, with systems providing real-time feedback on pedagogical effectiveness and suggesting personalized professional learning pathways. The teaching process itself becomes a collaborative human–AI practice: AI handles routine tasks, while teachers focus on irreducible elements of pedagogy: the relational, ethical, and inspirational dimensions of human encounter. Assessment systems move beyond standardized testing toward multidimensional, dynamic evaluation, capturing fine-grained learning processes and growth trajectories rather than merely measuring static outcomes.

Knowledge infrastructure undergoes fundamental dynamization. Curricula shift from fixed sequences to adaptive knowledge networks that generate learning pathways responsive to students’ existing understanding and developmental trajectories. Educational materials move from standardized textbooks to generative learning resources that produce multiple representational forms: visual, auditory, interactive, and textual. They calibrated to different cognitive structures and learning preferences. Assessment frameworks shift from measuring whether students “know” static content to evaluating their capacity to generate understanding in novel contexts, aligning evaluation more closely with the qualification dimension of educational quality.

These infrastructural innovations map onto educational quality as theorized through Biesta’s framework. For qualification, AI systems support creativity not by replacing human creative capacity but by serving as collaborative partners that provide provocations, variations, and rapid iteration cycles. They enhance problem-discovery capacity by revealing patterns and anomalies in data, though ethical–intuitive judgment about which problems merit attention remains fundamentally human. They assist problem-recognition capacity by visualizing problems through multiple epistemological frameworks, helping students understand whether they face technical problems amenable to algorithmic solutions, conceptual problems requiring theoretical clarification, or normative problems demanding ethical deliberation. They scaffold problem-solving capacity through environments that enable systematic hypothesis testing and solution iteration.

For socialization, AI does not replace human interaction but enriches the media through which interaction occurs. Translation and cross-cultural communication systems, for instance, do not eliminate the need for language learning but allow cross-cultural dialogue to begin earlier in the learning process, embedding relational practice within linguistic development. In special education, adaptive interfaces allow students with disabilities to participate in mainstream social interaction, expanding rather than restricting their relational world.

For subjectification, AI systems that record learning trajectories help students “see their own becoming,” providing reflective distance on their developmental processes. However, the irreducible core of subjectification, which encounters with genuine alterity that resist our projections and demand recognition cannot be algorithmized. AI infrastructure must therefore be designed to support, rather than substitute for, the human encounters that make subjectification possible.

Conclusion

In sum, artificial intelligence offers unprecedented opportunities for innovation in educational infrastructure, while simultaneously introducing equally unprecedented challenges. At this historical juncture, the attitude with which we face the future is of critical importance. In an educational domain that is deeply bound up with human becoming, we must maintain a stance of prudent conservatism.We need acknowle the limits of our understanding of the future, and preserving open, experimental space for future education to unfold.