Despite technological advancements in AI capabilities for education, current AI implementations in K-12 settings are not creating fundamentally new pedagogical practices but rather following the same principles of previous educational technologies.

Objective: The main goal of this study was to assess whether artificial intelligence has led to new pedagogical trends in education by examining the pedagogical applications of AI in K-12 contexts through a Human-Centered AI framework.

Methods: The researchers conducted a systematic literature review following PRISMA guidelines, involving:

• A comprehensive search across Web of Science, Scopus, and EBSBU databases
• Analysis of publications from 2019 to 2023
• Initial identification of 3,277 papers, narrowed down to 183 that met inclusion criteria
• Inductive coding analysis to identify emerging pedagogical themes in AI implementation
• Development of a Pedagogical-Centered AI (PCAI) framework to analyze findings

Key Findings:

• Six pedagogical categories emerged from the analysis: Behaviorism, Cognitivism, Constructivism, Social Constructivism, Experiential Learning, and Communities of Practice
• Current AI implementations in education largely follow traditional pedagogical principles rather than creating novel approaches
• The failure to transform education through AI stems from a lack of consideration of "pedagogical intelligence" (proposed as Gardner's ninth intelligence type)
• Behaviorist AI implementations typically focus on mathematics but extend to early education to enhance writing or creativity
• Constructivist AI systems prioritize understanding students' existing knowledge to tailor personalized learning experiences
• Social constructivist AI strategies focus on facilitating rich social environments where students collaborate
• Communities of practice AI implementations create environments where students engage at their own pace and build social learning communities

Implications:

• The study proposes a new Pedagogical-Centered AI (PCAI) framework that reconceptualizes Human-Centered AI for educational contexts
• This framework distinguishes between teacher-centered and learner-centered control in AI implementation
• To transform teaching and learning through AI, technology must be developed to mimic "pedagogical intelligence"
• Effective AI implementation requires understanding how different pedagogical approaches align with AI capabilities
• AI has potential to support personalized and learner-centered education when properly designed with pedagogical principles

Limitations:

• The study acknowledges limitations regarding the scope of search and analysis
• Restricting the search to English-language publications may have excluded relevant studies
• The focus on K-12 educational contexts may limit generalizability to other educational levels
• Many studies analyzed lacked solid theoretical learning frameworks or justifications for their AI implementations
• Current research emphasizes tool development without adequately addressing implementation strategies

Future Directions:

• Develop technology that mimics pedagogical intelligence to transform education
• Explore how AI can support students' decision-making and foster critical thinking, collaboration, and other 21st-century skills
• Scale research on AI systems designed for students with specific needs across diverse contexts and education levels
• Investigate how to translate learning theories into AI algorithms effectively
• Develop new theoretical learning frameworks to better understand AI's role in supporting student learning
• Further explore AI implementation through the lens of the PCAI framework

Title and Authors: "Artificial Intelligence for Teaching and Learning in Schools: The Need for Pedagogical Intelligence" by Brayan Díaz and Miguel Nussbaum

Published On: May 9, 2024 Published By: Computers & Education (Volume 217, 2024, 105071)

The research highlights a critical gap between technological advancement and pedagogical transformation in educational AI. While AI tools are increasingly sophisticated, they typically reinforce existing pedagogical approaches rather than creating truly innovative teaching methods. The proposed PCAI framework offers a new way to conceptualize AI implementation in education by considering both the level of control (teacher vs. learner-centered) and the pedagogical intelligence embedded in the AI system. For AI to meaningfully transform education, developers must focus not just on the technology's capabilities but also on incorporating genuine pedagogical intelligence that aligns with effective teaching and learning principles.