"I can tell you wrote this yourself"

Universities and workplaces are increasingly nudging people to use large language models for tasks like writing reports, raising worries that core skills such as critical thinking, clear writing, and genuine human expression will atrophy. Commenters contrast AI-assisted spellchecking or code review with delegating whole papers or design docs to models, arguing that in many domains the learning and value lie in the process, not just the polished output. Others counter that AI is simply the next stage of automation—akin to calculators, CAD, or industrialization of manual crafts—and that the real challenge is to use these tools selectively without surrendering understanding or responsibility.

Human-written style, errors, and “I can tell you wrote this yourself”

  • Many fixate on the “right/write” typo as evidence of genuine human authorship; some find it charming, others think it undercuts the point.
  • Several argue visible imperfections and idiosyncrasies will become markers of non‑LLM text, potentially even deliberately faked later.
  • Others say grammar tools (AI or not) are low‑risk helpers and that obvious errors should be caught, especially in formal work.

AI as writing assistant vs erasing voice

  • One camp sees LLMs as powerful grammar/style reviewers, idea‑bouncers, or “cog manufacturers” that free humans to focus on design and higher‑level thinking.
  • Another camp warns LLMs normalize tone, soften opinions, and blur individual voice; they find LLM text “flat,” like reading a transcript instead of a live conversation.
  • Some adopt a rule of never pasting AI sentences directly, only using suggestions, to preserve voice.

Thinking, learning, and cognitive offloading

  • Recurrent worry: outsourcing writing and problem‑solving dulls critical thinking, especially among students; the point of a report or design doc is the thinking it forces.
  • Others counter that tools have always removed grunt work (calculators, CAD, FEA); understanding and validation still matter, not doing everything by hand.
  • Several emphasize a middle ground: use AI for boilerplate and mechanics, but do the reasoning, research, and structuring yourself.

Education, cheating, and “prompting as a skill”

  • Some describe courses that effectively require prompting LLMs for reports or code, interpreting this as academia “giving up” on teaching fundamentals.
  • Others see responsible AI use as a necessary professional skill and compare banning it to banning WolframAlpha for integrals.
  • There is concern that undisclosed AI use in assignments turns education into credential‑chasing rather than skill‑building.

Industrialization of knowledge work and jobs

  • Multiple comments liken LLMs to industrialization of intellectual labor: faster, cheaper output, but less craft, pride, and deep understanding.
  • Debate over whether displaced workers will find new “niches” as with past automation or whether AI’s breadth makes this transition uniquely harsh.

Social conflict and ethics around AI

  • Some AI users feel “othered” and attacked as lazy or unethical; critics say backlash is a response to real harms (theft of training data, spam, enshittification).
  • Broader worries include environmental costs, centralization of power, declining literacy, and the difficulty of maintaining human‑only spaces online.