LLM AI: When Theory Is Threatened by Reality

(Yes, I teach LLM AI and digital marketing, but I'm also interested in Kant, Heidegger, Foucault, and so on. Here's some stuff I wrote this morning before breakfast. -- andreas)

The anxiety over AI-generated text comes out of a contradiction in modern culture. While the theoretical ideas such as Michel Foucault's "Death of Man" and Roland Barthes' "Death of the Author" are commonly accepted, the reality of LLMs reveals the gap between theory and the unspoken deep idea of human exceptionalism. We talked about the theory of the death of the author but didn’t realize it could actually happen.

Foucault's "death of man" questioned the position of human consciousness as the subject and object of knowledge. Barthes argued meaning emerges from readers and language systems, not authorial intention, where texts are cultural codes rather than individual creation.

LLMs turn these theories into reality by generating text without intention. LLMs are post-structuralist theory in practice: LLM AI text is recombination, writing without origin, meaning through difference, and language speaking itself. Yet those who taught these theoretical ideas are terrified of LLMs.

This is clear in the academic response to AI text generation: students are accused (and often dismissed from universities) over AI plagiary. Plagiary is based on the idea of human uniqueness and the privileged position of an author as the origin of knowledge.

The real fear is the threat to institutional authority. A professor says the author is dead… and yet continues to present himself as a professor, a researcher, an educator, and an author of his books as an expert on saying the author is dead. His position evaporates when death of the author becomes reality.

Think about plagiarism detectors. Isn’t that so ironically funny? Professors use AI tools to find out if students used AI! These professors and teachers are admitting they can’t tell if the text was written by humans or not. Should a student get a failing grade for using AI to write? No, the professor or teacher should be dismissed for using AI plagiary detectors.

Instead of relying on AI tools to hunt the users of LLM AI, we should show students how theory has become reality. We can discuss the contradiction of a society that talks about a theory yet clings to the past of human exceptionalism and the power of institutions to enforce that past.

This isn’t just an academic issue in the university. Some of you may say, "well, in the real world…", yes, in the real world, staffers are expected to use their skills to do work, managers and directors get bonus for their work, people get awards for original contributions to the organization, and so on. LinkedIn uses photo IDs to verify profiles. It’s not just academic book authors, it’s experts and skilled people at all levels in work. They are paid for their uniqueness and originality.

The end of human uniqueness cuts to the bedrock of modern nations, which are based on the ideals of the concept of the person, free speech, free assembly, the right to vote, civil rights, and so on. These core ideas are deeply threatened by the reality of LLM AI. The contradiction between ideal and reality and the threat to those ideals is the unacknowledged reason for the resistance to LLM AI.