As mentioned before, I hate bringing this up because I have no evidence or expertise here, just a gut feeling. But I just can't help feeling like, aside from everything else aside about LLM chatbots, they're quickly becoming the leaded gasoline of our time.
Something doing real damage to human cognition, but in this diffuse and difficult to measure kind of way.
Many, not nearly all but *many*, folks using this things seem (again, as a gut feeling) to just talk differently after contact with chatbots? I can't even quite put my finger on it, but it scares the shit out of me.
It's not even an argument against chatbots, I have plenty of arguments that are far better substantiated, it's a personal fear about what they're doing.
1. I have absolutely seen this kind of—i hate using this term but there's not really any other word for it—"cognitive decline" from many people, and I am collecting a file on documented public instances of it. It's definitely fucking scary. I will say that it is selective, and I don't know why *some* users seem to suffer from it and others don't. I certainly haven't seen a pattern. It seems to be a general pattern of which the infamous "AI psychosis" is a subcategory
3. It's also because of (1.) that my own usage has shrunk to nothing. I think that some of the people I am presently arguing with will end up being "safe" but I don't know which ones, or why. I haven't seen any plausible safety protocols; I do not know how to experiment with it safely. So every time I open up a chat prompt I feel like I'm asking myself "how big of a swig from this flask of luminous radium paint do I feel comfortable drinking in one sitting".
@glyph @SnoopJ @aburka @xgranade
Does the SCP Foundation have any general material around environmental and occupational infohazards? Because model based conversation entities would seem to fit.
(joking, but only barely so: SCP is satire as much as it is speculative existential horror)
Unless a more helpful approach would be anthropology or sociology: "tools shape users" or a power and consent analysis..
@glyph @aburka @xgranade coined last week in this which might have missed you: https://www.theregister.com/2026/02/16/semantic_ablation_ai_writing/
(coined to describe the "AI" writing but here I'm using it to describe the bleed of that into the user's writing/speech)
@SnoopJ @glyph @aburka @xgranade I *really like* the term semantic ablation, so I do not appreciate my brain barging in with 1) a George Carlin bit from "Parental Advisory: Explicit Lyrics" about how we bury meaning in euphemism over time, relevant because of 2) the post-it note my brain's waving at me that reads "semantic ablation is the mechanism, but just say Newspeak"
stupid brain
remembering things
@microblogc @glyph @aburka @xgranade I rather prefer how Ted Chiang put it 3 years ago now (!) but since this is attracting attention, just in case anyone of present missed that one when it still smelled of fresh bits:
https://www.newyorker.com/tech/annals-of-technology/chatgpt-is-a-blurry-jpeg-of-the-web
Cassandra is only carbon now
aburka 🫣
SnoopJ
Glyph
Jeff Miller (orange hatband)
Random Geek
Microblog Castellano
Talya (she/her) 🏳️⚧️✡️
the gripe ibis