How Generative AI Can Harm Higher Education, With Special Emphasis on Legal Education by Scott Fruehwald

Abstract

Generative AI has the potential to do catastrophic harm to higher education.   This is because learning is a biological process that requires years and years of effortful labor.  Anything that disrupts this labor will severely damage an individual’s ability to function in the world.  Gen AI is such a disruption because it substitutes efficiency for learning.  Nothing can compensate for the losses incurred when Gen AI is used instead of the human brain.The problem with Gen AI performing learning tasks rather than the student is that there is no learning.  This lack of learning applies to both basic and advanced learning.  Over reliance on Gen AI produces less factual knowledge and process knowledge than traditional learning in long-term memory—potentially much less knowledge. In addition, it could significantly harm advanced learning.  To begin, advanced learning requires a foundation in the brain.  Experts have much, much greater factual and procedural knowledge than novices.  If such basic knowledge does not exist in the brain (long-term memory), one cannot become an expert.  There is no way that an individual can make up for a lack of basic knowledge in long-term memory—not by using Wikipedia or Gen AI.  Basic knowledge must be internal for advanced thinking.  Similarly, students cannot imprint advanced thinking techniques on their brains if Gen AI does the work for them.

Fruehwald says this is the first chapter of an AI book.