Augmented sociomateriality: implications of artificial intelligence for the field of learning technology

Abstract

There has been a conscious effort in the past decade to produce a more theoretical account of the use of technology for learning. At the same time, advances in artificial intelligence (AI) are being rapidly incorporated into learning technologies, significantly changing their affordances for teaching and learning. In this article I address the question of whether introduction of AI and associated features such as machine learning is a novel development from a theoretical perspective, and if so, how? I draw on the existing perspective of sociomateriality for learning and argue that the use of AI is indeed different because AI transforms sociomateriality by allowing materiality to take on characteristics previously associated primarily with a human agent, thereby shifting the nature of the sociomaterial assemblage. In this data and algorithm-driven AI-based sociomateriality, affordances for representation and agency change, thereby modifying representational and relational practices that are essential for cognition. The dualities of data/algorithm, representational/agentic augmentation, and relational/participatory practices act in tandem within this new sociomaterial assemblage. If left unchecked, this new assemblage is prone to perpetuate the biases programmed within the technology itself. Therefore, it is important to take ethical and moral implications of using AI-driven learning technologies into account before their use.