Humanist educational approach in a world with Large Language Models (Artificial Intelligence): Reflexions and experiences from a university professor

Abstract

Large Language Models (LLM’s) are statistical programs able to generate text extremely rapidly, known colloquially as “Artificial Intelligence” (AI). Due to the fact that they tend to be trained with huge amounts of online information, results are often quick and surprising, to the point that many of them are already passing the famous “Turing test”; however, most experts still claim these models are far from being considered “Artificial General Intelligence (AGI). The main advancements that we are interested in are the speed of the new models, as well as the availability of use for any person with an internet connection, which has deep implications in several fronts, specifically for educators and students. This article uses hermeneutical phenomenology to reflect on, and re-signify the role of the university professor, in an environment where most people, having at least one device that can be connected to internet, have access to LLMs. Firstly, we offer the technical definition of LLM, along with some statistics, then we explore the scientific literature on the use of LLMs in university students and finally, we reflect on its consideration, for the rethinking of activities inside the classroom, the redesign of educational programs and even the modification of whole educational policies. We conclude on the importance of the humanist approach, and mediation on the university professor in the use of LLMs.

Keywords

• humanist approach
• large language models
• artificial intelligence
• higher education