leon

The infuriating part is that people apply the pop-psychology definition of "trauma-bond", but apply the actual psychology judgement of it, as in, "obviously bad, duh, everyone knows that".

In pop psychology, "trauma bonding is when you're bonding with someone over shared trauma or similar past experiences." In actual psychology, "A trauma bond refers to the situation where someone develops strong feelings of sympathy, loyalty, dependance, and attachment towards their abuser."

In pop psychology, "Gaslighting is when someone lies to you, questions you, or disagrees with you over what happened." In actual psychology, "Gaslighting is an abuse tactic where you are manipulated to question your reality, memory, or sanity—not just a disagreement over what happened."

In pop psychology, "Being triggered means that something makes you uncomfortable or annoyed." In actual psychology, "A trigger is something that causes a sudden increase in symptoms—like flashbacks in PTSD or compulsions in OCD. Its not just a source of general discomfort."

In pop psychology, "Intrusive thoughts are random impulses you might want to act on; e.g. wanting to do something unexpected like dye your hair pink or eat cake for dinner." In actual psychology, "Intrusive thoughts and unwanted, involuntary thoughts that go against your real value and desires. You dont want to act on them, and them coming true would be your worst nightmare. They can be very prominent and distressing in disorders like OCD and PTSD."

princessgirls -> OUT

insectoid mercenarygirls -> IN

doggirls -> OUT

insectoid assassingirls -> IN

mechsploitation pilotgirls -> OUT

insectoid warriorgirls -> IN

I am once again begging people to realize that AI checker doesn’t work. it’s never worked. it’s notoriously known to have flagged human-made works as AI and AI-generated works as human-made. and by feeding it people’s works, you are feeding more works to AI, because apparently the machine itself is AI.

What would be the economic advantage of reliable AI checkers even if that were technically possible? Why would the company giving individuals subsidised AI to make them dependant consumers want those people punished for being dependant consumers?

Humans cannot tell LLM text from human made text either - however good you think you are it is highly probable you are over estimating . I believe LLMs have affected how whole cohorts write - I suspect but obviously cannot prove that trying to keep up with study while working several jobs or being a carer is likely to make a human write superficially and vacuously, citing work they have not read and understood (just blagging having skimmed an abstract) - but more than this LLMs are human creations and humans are merging with them stylistically - for example LLM created text is easy to grade high if you are not paying attention because there are millions of examples of professional publications chewed up in there that have consistently filtered for white middle class forms of English, specific forms of not saying much but appearing fancy, mass economic and social pressure about what a 'good' literary novel, research article, romance, educational resource for undergrads etc sounds like.

The problem of AI and LLMs is not someone who put someone else's fic in a chatbot to get a new but very similar story because they like them, it is not someone writing their resume because they don't know how, it is not someone turning in a paper at the last minute because they were up all night with a sick baby.

I feel like a lot of anti-AI is snobbery and disgust - the class politics of the real, the handmade, the organic. A fantasy of cottagecore meritocracy.

The desire to hunt outsiders with pitchforks. To say to the monster you do not know how to fit in here.

Instead I want us to say to individuals - this does not work well for learning, what would be more fun but just as easy?

This does not work well for getting over your feeling you cannot make art - what could we do to help people not be ashamed of their drawing, handwriting, fiction.

this does not help the way education is a factory where you have to make stuff that sounds white and middle class - that nobody cares about your individual skills they just want you to have enough student debt that you don't go on strike if you get a job - how can we organise for better working condituons

AI and LLMs are digital colonialism - Real people are being exploited and mentally harmed - particularly in West and Southern Africa - being "Digital Janitors" for silicone valley to keep their illusion going till they can cash out (very soon) as IPOs - when they sell the shares to become'public' companies with billions of debt, unbuilt data centres, and those shares will be bought by pension companies and local government. The whole economic edifice is likely to come crashing down soon and like - yay end to LLMs at scale - but real people, real Black and Brown, and poor people, Disabled people, people needing medical care - those people are going to get crashed on the way down too

look up