Historical threads, missing links, and future directions in AI in education

Analysis

Why This Paper Matters

This paper is a seminal critical review of artificial intelligence in education (AIED), published in 2020 when AI was becoming increasingly embedded in everyday life and educational technologies. It matters because it challenges the prevailing techno-optimistic narrative that AI will automatically improve education, and instead situates AIED within historical, social, and political contexts. By tracing the 'historical threads' from early cybernetics and intelligent tutoring systems to contemporary machine learning and learning analytics, the authors reveal how current AIED systems are not neutral but carry forward assumptions about learning, measurement, and control.

The paper's significance is amplified by its timing—just before the rapid expansion of generative AI in education—and its call for 'missing links' such as attention to equity, data justice, and the voices of marginalized students. It has become a reference point for researchers and policymakers seeking to critically evaluate AIED, and its 734 citations attest to its influence in shaping a more cautious and socially aware research agenda.

Technical Contributions

The paper's technical contributions are primarily conceptual and critical, rather than algorithmic. Key innovations include:

• Historical periodization: It identifies distinct eras of AIED, from rule-based systems to machine learning, and shows how each era's technical capabilities shaped pedagogical assumptions.
• Critical framework: It introduces a sociotechnical lens that examines AIED as a 'black box' of algorithms, data, and power, urging researchers to open this box.
• Identification of missing links: It highlights gaps such as the lack of research on AIED's impact on teacher autonomy, student agency, and systemic inequalities.
• Future directions: It proposes a research agenda that includes participatory design, algorithmic transparency, and policy interventions to ensure AIED serves public good rather than commercial interests.

Results

As a review paper, it does not present quantitative results or experimental comparisons. Instead, its 'results' are the synthesis of historical and contemporary literature, leading to a set of critical insights. The paper demonstrates that AIED research has been dominated by cognitive and individualized models of learning, often ignoring social and cultural dimensions. It also shows that current AIED systems, such as adaptive learning platforms, can perpetuate biases and surveillance, and that there is a lack of robust evidence for their effectiveness in improving learning outcomes. The paper's main outcome is a call for a 'critical AIED' that prioritizes equity and ethics.

Significance

The broader impact of this paper lies in its role as a catalyst for critical AIED scholarship. It has influenced subsequent research on algorithmic fairness in education, data privacy, and the ethics of learning analytics. It has also informed policy discussions about the regulation of AI in education, particularly in Europe and beyond. By framing AIED as a sociotechnical phenomenon, the paper encourages interdisciplinary collaboration among educators, computer scientists, sociologists, and policymakers. Its legacy is a more nuanced understanding that AI in education is not just a technical challenge but a matter of social justice, and that future developments must be guided by democratic and participatory principles.