AI Mediated Learning Architectures

In the last thirty years, there has been an exponential growth in research on Artificial Intelligence (AI), with applications in natural language processing, automatic programming, robotics, and computer vision. Education has become a valuable field of study in this fast-paced era, with theoretical and epistemological implications emerging through the concept of Bio Educational Technology. In education, AI is employed to personalize learning, provide support, assessment, and enhance communication. Expert systems and intelligent tutors aid in monitoring classroom learning dynamics and detecting student difficulties at an early stage. The application of Artificial Intelligence in education began in the 1960s with Computer Assisted Instruction (CAI), which later evolved into Intelligent Tutoring Systems (ITS), which made it possible to meet subjective learning needs. The focus of research has shifted from metacognitive, motivational, and emotional aspects over time, leading to an evolution towards adaptive and collaborative learning environments. The gap between research and practice persists as a crucial issue, with the possibility that descriptive approaches lack practical impact on educational reality. The idea of Nooscope, conceived by Pasquinelli and Joler (2021), emphasizes AI's role as an epistemic tool, yet criticizes its history, dataset, and algorithmic biases. Concerns arise about phenomena like data cannibalization, where AI models train on data generated by other AIs, resulting in a decrease in quality and reliability. The Human-Centred AI (HAI) approach is proposed as an alternative to traditional AI, emphasizing the augmentation of human intelligence instead of its replacement. There are two viewpoints that come to mind: an AI that is supervised by humans (as opposed to algorithmic training systems) and an AI that is developed to meet human needs. To avoid education that is driven by reductionist and technocentric logics, it's important to balance personalization aspects while maintaining a critical approach. AI has the ability to provide new tools for precise teaching that can adapt to individual needs and improve the quality of teaching, while still keeping the human role at the center of the technology ecosystem.