I don’t want to read LLM output

That is an interesting remark upon which I see no one better suited to reflect than the German poet Heinrich Heine (1797 – 1856) in his Reisebilder (engl. Pictures of Travel), where in the Third part, Chapter 13 he writes about the inhabitants of Norderney, a German island in the North.

Remarks upon the people of Norderney in a translation

That which connects these men so firmly and contentedly, is not so much the inner mystical sentiment of love, as that of custom — that mutual " through-and-above-living" according to nature, or that of social directness. They enjoy an equal elevation of soul, or, to speak more correctly, an equal depression, from which result the same needs and the same desires, the same experiences and the same reflections. Consequently, they more readily understand each other, and sit socially together by the fire in their little huts, crowd up together when it is cold, see the thoughts in each other’s eyes before a word is spoken, all the conventional signs of daily life are readily intelligible, and by a single sound, or a single gesture, they excite in each other that laughter, those tears, or that pious feeling, which we could not awaken in our like without long preliminary explanations, expectorations and declamations. For at bottom we live spiritually alone, and owing to peculiar methods of education, and peculiar reading, we have each formed a different individual character. Each of us, spiritually masked, thinks, feels and acts differently from his fellow, and misunderstandings are so frequent, that even in roomy houses, life in common costs an effort, and we are everywhere limited, everywhere strange, and everywhere, so to speak, in a strange land.

Heine’s text contains a warning against praising shortness in expression as a virtue: conformity and boundedness of thought. “Make not many words, because we (already) know what you will say and what needs to be said,” looms behind that praise of shortness.

Clearly, this community and this place (a club with rules) has every right to demand that “human authorship only” is enforced in a rather strict way. But there is a price to pay, because (as has become clear by now) it is not only about “slop” nor “clarity of exposition” nor “originality of content” that the community cares about. The call for “authenticity” and (visible) “human fallibility” will not only get rid of “slop”, it will likely also get rid of new ideas and maybe even better arguments. This site is named “Discourse” and I am explicitly reminded of Popper’s call to allow for criticism as broadly as possible, irrespective the source. Refute critique for bad reasoning and false facts, but do not ban it for suspicious number of em dashes or strange use of vocabulary (even if correct according to Webster’s).

It seems that the issues this thread has raised are already being recognized and addressed by the social sciences. I have added some interesting papers in this regard below. From cursory reading I find that it is normal that AI use in communication is “penalized”, e.g., you earn less trust. Interestingly, the scientists even found a paradox in this regard: People would call for make AI use transparent, but would penalize people, who did so, providing an absurd incentive to hide the use of AI (cf. my post, which was automatically (!) hidden without me being given any official statement or simply staff approval so far).

There seems to also be some risk of McCarthyism according to the papers, e.g., people use the “AI-found-inside” label to suppress what they do not like or want to see. While it has been denied in this thread, sociology appears to be crisp and clear in this regard: Determining who gets to say what and how is a form of wielding power in societies (and communities). Interestingly, while people mistrust computers and AI, the measured effects of AI assistance in the communication setting studied by Hohenstein et al. (see below) were mostly positive even while suspected AI use itself produced negative evaluations.

Outside this forum, the argument will increasingly be settled differently. Developers, employers, clients, and entrepreneurs will compare what people can accomplish with whatever tools they choose to use. They will not award bonus points for avoidable friction, nor mistake inefficiency for humanity.

Suggestions for further reading
  • Baert, P., Dorschel, R., Hall, M., Higgins, I., McPherson, E., & Philip, S. (2026). Dialogues Towards Sociologies of Generative AI. Social Science Computer Review, 44(1), 59-79. https://doi.org/10.1177/08944393251370354.
  • Jeffrey T Hancock, Mor Naaman, Karen Levy, AI-Mediated Communication: Definition, Research Agenda, and Ethical Considerations, Journal of Computer-Mediated Communication, Volume 25, Issue 1, January 2020, Pages 89–100, https://doi.org/10.1093/jcmc/zmz022.
  • Siavosh Sahebi, Paul Formosa, Sarah Bankins, The AI penalty and disclosure paradox: Trust, authenticity and knowledge uptake in AI-mediated communication,
    Computers in Human Behavior: Artificial Humans, Volume 8, 2026, 100304,
    ISSN 2949-8821, Redirecting.
  • Hohenstein, J., Kizilcec, R.F., DiFranzo, D. et al. Artificial intelligence in communication impacts language and social relationships. Sci Rep 13, 5487 (2023). Artificial intelligence in communication impacts language and social relationships | Scientific Reports.
  • Watson, S., Brezovec, E. & Romic, J. The role of generative AI in academic and scientific authorship: an autopoietic perspective. AI & Soc 40, 3225–3235 (2025). Client Challenge