The Psychology of Learning from Machines: Anthropomorphic AI and the Paradox of Automation in Education

TL;DR

This paper integrates automation psychology, human factors, HCI, and philosophy of technology to explain how learners psychologically relate to anthropomorphic AI tutors and the risks of over-reliance. It combines a cross-disciplinary theoretical synthesis with a large-scale naturalistic analysis of YouTube discourse to identify three amplified challenges: trust calibration failures, the anthropomorphism paradox, and automation ironies that can erode skills and professional judgment. The study finds domain-specific trust patterns, distinct affective engagement, and temporal evolution where deeper analytical engagement often emerges only after time, suggesting a complementary AI–human design where AI handles routine content and humans preserve judgment, ethics, and relational learning. These insights yield design principles emphasizing transparency, calibrated anthropomorphism, and learner agency, with protective measures for vulnerable populations and calls for longitudinal, cross-cultural validation. Practically, the work supports differentiated deployment in engineering education, leveraging AI for foundational topics while maintaining robust human mentorship for ethics, design, and professional identity formation."

Abstract

As AI tutors enter classrooms at unprecedented speed, their deployment increasingly outpaces our grasp of the psychological and social consequences of such technology. Yet decades of research in automation psychology, human factors, and human-computer interaction provide crucial insights that remain underutilized in educational AI design. This work synthesizes four research traditions -- automation psychology, human factors engineering, HCI, and philosophy of technology -- to establish a comprehensive framework for understanding how learners psychologically relate to anthropomorphic AI tutors. We identify three persistent challenges intensified by Generative AI's conversational fluency. First, learners exhibit dual trust calibration failures -- automation bias (uncritical acceptance) and algorithm aversion (excessive rejection after errors) -- with an expertise paradox where novices overrely while experts underrely. Second, while anthropomorphic design enhances engagement, it can distract from learning and foster harmful emotional attachment. Third, automation ironies persist: systems meant to aid cognition introduce designer errors, degrade skills through disuse, and create monitoring burdens humans perform poorly. We ground this theoretical synthesis through comparative analysis of over 104,984 YouTube comments across AI-generated philosophical debates and human-created engineering tutorials, revealing domain-dependent trust patterns and strong anthropomorphic projection despite minimal cues. For engineering education, our synthesis mandates differentiated approaches: AI tutoring for technical foundations where automation bias is manageable through proper scaffolding, but human facilitation for design, ethics, and professional judgment where tacit knowledge transmission proves irreplaceable.

Figures (2)

• Figure 1: Domain-dependent sentiment patterns ($n=72{,}178$). Audiences readily accept AI for abstract topics but show strong polarization for moral-ethical content, mirroring differential acceptance of AI in technical instruction versus professional ethics.
• Figure 2: Comparative analysis of AI debates vs. engineering tutorials ($N=104{,}984$).