Generative Artificial Intelligence and Extended Cognition in Science Learning Contexts

Abstract This paper philosophically examines the impact of generative artificial intelligence on learning processes from the perspective of extended cognition. The central problem addressed is how these technologies can transform students into passive or active learners, influencing the development of cognitive skills. It will be argued that generative artificial intelligence presents risks of diminishing cognitive activity among students, as it is likely to substitute—rather than complement—the cognitive subject. It will also be argued that there are ways to leverage generative artificial intelligence so that learners are not passive but rather active cognitive subjects. Three cases will be presented, with empirical support, to show how this leveraging is possible: the production of feedback, assistive technologies, and gamification. In these cases, generative artificial intelligence is a complementary cognitive artifact rather than a substitutive one. To achieve this goal, the paper presents the framework of the extended mind thesis as a conducive scenario for analyzing the relationship between generative artificial intelligence and learning contexts, and it analyzes specific cases of science education. An analysis of the types of cognitive artifacts will also be conducted to examine how generative artificial intelligence intervenes in learning in both substitute and complementary ways.