It is argued that intelligence should be assessed not by output quality or efficiency alone, but by its impact on intellectual character, arguing that intelligence should be assessed not by output quality or efficiency alone, but by its impact on intellectual character.
Abstract
Abstract The rapid adoption of generative AI tools has intensified debate about whether these technologies enhance or diminish human thinking. While proponents argue that AI can support learning through efficiency, personalization, and feedback, emerging research suggests that routine AI use may reduce recall, creativity, and original reasoning. This paper argues that the central risk posed by generative AI is not cognitive decline per se, but the erosion of epistemic humility – an intellectual virtue essential to learning, inquiry, and human flourishing. Drawing on Aristotelian and contemporary virtue epistemology, the paper conceptualizes epistemic humility as accurate awareness of one’s epistemic limits, openness to correction, and willingness to engage in sustained cognitive effort. Generative AI systems, by producing fluent and authoritative-seeming outputs, can foster an illusion of understanding in which users mistake access to information for genuine comprehension. This illusion encourages epistemic overconfidence, passive dependence, and the outsourcing of judgment, undermining the conditions for responsible knowing. The paper situates these concerns within broader debates about cognitive coupling between humans and AI, arguing that intelligence should be assessed not by output quality or efficiency alone, but by its impact on intellectual character.
The concept of “alienated intelligence” is introduced to critically examine contemporary cultural perceptions of artificial intelligence to argue that certain visions of AI function as cultural projections of human cognitive capacities that become externalized, idealized, and reified in technological systems.
It is argued that this understanding of human-GAI engagement, as explained through epistemological beliefs, lays the foundation for alternative approaches to teaching and assessment, student interactions, professional development, and AI governance and policy, while noting that the framework remains an exploratory heuristic requiring empirical validation.
S. Strydom· Journal of Applied Learning...· 0 citations
The proposition that agentic artificial intelligence may precipitate a depletion of collective cognitive capital has circulated with unusual velocity in both scholarly and public discourse. The present paper offers a deliberately heterodox reading of the dynamic model advanced by Acemoglu, Kong and Ozdaglar (2026). Rather than reconstructing the formal apparatus or replicating its notation, we reposition the argument within three underutilized scholarly streams: the cognitive ergonomics of human-machine collaboration, the institutional ecology of knowledge stewardship, and the developmental psychology of novice expertise formation. We introduce a phase-space taxonomy that maps commons trajectories as functions of effort elasticity and knowledge complementarity, and we advance a governance typology calibrated to distinct cognitive levels - declarative, procedural, causal, and metacognitive. Drawing upon recent experimental evidence on neural offloading (Kosmyna et al., 2025), educational neuroscience (Lodge and Loble, 2026), and critical-thinking erosion under AI assistance, we argue that the collapse narrative, while theoretically coherent, overstates uniformity and understates adaptive capacity. The paper supplies a governance matrix organized by cognitive level and institutional lever. We conclude that the salient policy challenge is not the prevention of an inevitable collapse but the design of polycentric stewardship regimes that render the commons robust to heterogeneity in human responsiveness.
Artificial intelligence is being adopted in educational settings faster than its consequences are understood. We argue that the central risk is misalignment: AI that eliminates human effort erodes the very capacities education is meant to build. We organize this risk into an integrative framework of four interrelated dimensions -cognition, agency, emotional well-being, and ethics- linked by a self-reinforcing cycle where cognitive offloading reduces effort, weakens agency, and compounds emotional and ethical harm. We ground the framework in the perspective of a small cohort of students: an exploratory analysis of 49 International Baccalaureate argumentative essays about the impact of AI reveals that learners perceive these risks, with $80\%$ of essays reporting that AI reliance reduces thinking. At the same time, the essays articulate a consistent vision of the AI the students want: systems that support rather than replace learning by withholding immediate answers, prompting recall, and encouraging reflection through questions instead of solutions. These desiderata closely align with established principles from the learning sciences. Building on these insights, we propose a single design principle, scaffold, do not substitute. We argue that this principle extends beyond education. It represents a broader challenge for the AI ecosystem: any system that mediates human thinking can either weaken human capabilities through substitution or strengthen them through scaffolding. We conclude by outlining a research agenda for developing AI systems that foster enduring human capacity, an imperative not only for learners but, ultimately, for democratic societies.
Lucile Favero, J. A. Pérez-Ortiz, Tanja Käser et al.· 0 citations
Generative artificial intelligence (GenAI) is rapidly reshaping engineering education, yet prevailing debates frame its impact narrowly as either pedagogical innovation or epistemic threat. In this article, we argue that such framings misdiagnose the problem by treating GenAI as a tool rather than as a sociotechnical reorganization of epistemic authority, agency, and professional formation. Drawing on sociological theory, philosophy of science, labor economics, and recent empirical studies, we show that GenAI lowers epistemic floors by expanding access to competent performance while leaving epistemic ceilings, those associated with judgment, authority, and problem framing, largely intact. This asymmetry produces new forms of stratification, redistributes responsibility without redistributing control, and stabilizes plausibility‐based forms of knowledge in place of warrant‐based engineering reasoning. We advance three central contributions. First, we conceptualize a shift from warrant to plausibility in engineering knowledge, in which fluent, algorithmically generated outputs increasingly circulate as legitimate despite bypassing core practices of verification, critique, and accountability. Second, we extend the concept of the AI wall beyond productivity to describe a ceiling of epistemic development, where further learning and judgment are not institutionally rewarded without deeper expertise, authority, or complementary organizational resources. Third, we theorize how engineering education participates in stabilizing these dynamics through what we term the social campus, the institutional space in which curricular norms, assessment regimes, professional signaling, and peer comparison converge to normalize AI‐mediated epistemic authority. Together, these dynamics position engineering education at a critical juncture. We argue for a reorientation toward epistemic stewardship, in which engineering programs deliberately cultivate the capacities needed to interrogate, govern, and responsibly deploy AI systems rather than merely operate within them. The question we want to highlight is not whether engineering education will adapt to AI, but whether it will do so as a steward of epistemic responsibility or as an accelerator of epistemic stratification.
