The Negative Impact of Advanced Generative Textual AI Models on Education

It's still fairly early into the widespread adoption of LLMs, and other kinds of advanced generative textual AI models, and there are already a selection of studies showing the negative impact on people's ability to independently think critically or two write well without the assistance of a chatbot or other generative textual AI model; however, this note is about the interplay between these effects, and the last 70 years of progress in education via the adoption of writing across the curriculum. See also Writing to Learn by William Zinsser and the books central thesis around writing being a vehicle for structured thought, good writing being evidence of well-structured thought, and writing across the curriculum as a method for instructing well-structured, critical thought. The entire point of the argument in favor of introducing writing across the curriculum in Zinsser's book is completely undermined by students using AI models for any form of assistance—aside from, perhaps, for spelling and grammar checks.