@artemis When I was teaching freshman writing at a very prestigious university, I had a student who had a list of words he wouldn't use because his high school English teacher told him they were commonly misused. (affect, etc...) He'd go through these awful circumlocutions to get out of using perfectly normal words.

I fear we're going to institutionalize this kind of behavior.

@artemis I read somewhere that neurodivergent is bad and weird, and they deserve it.
No, wait a minute, I'm neurotypical, and I don't read unless I have to.

@artemis my dad has such a poor handwriting that I was accused several times of imitating his signature on various school documents, because of how variable his actual signature was.

Of course we turned this into an advantage at some point as schools required ridiculous amounts of stuff to be signed by parents, and they could not be arsed to do so every evening.

@artemis Fight this. Tooth and nail.

@artemis

I'm devoid of calm about this.

@artemis

Un-AI-ing?
Soon there'll be an AI to do that for you...
Uh. Right.

@artemis I definitely recommend screaming. The hair pulling bit isn't the best experience.

@artemis
We went through this phase with plagiarism checkers, with students submitting endless copies of their reports to reduce the % to zero, and you end up with unreadable garbage.

@artemis oh. Oh dear. I already have people telling me I make up words that are absolutely not made up. Not going to be great for conversation.

@artemis@dice.camp I hope """AI""" destroys homework as something people use.

@artemis I used a lot of unusual words when I was a child because I read a lot of books. Basically, I spoke like a dead author.

@artemis my gosh I love the word devoid though

it's punishing real, human intelligence in all ways

@artemis What does a large language model know about detecting text from large language models?? It's been trained mostly on massive amounts of text from before there were large language models!

So it's bullshitting people about the work of actual other people, and it doesn't even have the data set to fake it.

https://www.scientificamerican.com/article/chatgpt-isnt-hallucinating-its-bullshitting/

@artemis we're so angry. we have no words.

@artemis this hasn’t happened to me yet but i’m honestly a little scared of this happening because people think of these checkers as absolute and objective and it’s hard to defend against someone who thinks that

@artemis

Seriously, I... I read the dictionary for _fun_ as a kid. I know that's not normal but I don't need robots bullying me for it.

Also, using a bot to determine if a bot wrote something is definitely a second layer of stupid.

"Well, the machine said it was written by a machine. Who else would know so well what a machine wrote than a machine? Hmm?"

The most reasonable response I can come up with to any of this is to burn it down. Just... Burn it down.

@artemis As a kid, my vocabulary was well above average. To punish someone for being too proficient at language is absurd. How about we go back to having actual human teachers evaluate students?

@mspro ^^

@artemis indeed. Autistics like myself often develop a more formal style of communication.

@artemis

This would have me camping out in the principle's office roasting marshmallows over the fire that was once their desk.... 🤬

(I wonder how adding "Please FOAD, Mr. AI, you & whoever signed on the vendor" to the edit would change the "AI score"?)

@artemis @PragmaticAndy this makes me want to delve further.

@artemis
A co worker has tried an AI checker on one of his papers that he did long before ChatGPT was even a thing. The "AI checker" claimed the paper to be likely written by AI.
"AI checkers" are complete nonsense and do not work.

@artemis Harrison Bergeron vibes

@artemis This is also true of people who have learned English as a second (or third! or more!) language, particularly from certain parts of the world, and particularly people who have read a lot. They are then punished for their erudition.

@artemis One of the things that's not appreciated is that if AI development breaks a certain threshold, there will come a point where human cognition is less accurate and more prone to hallucination than AI cognition, but it will still be humans judging AI by how human it is.

@artemis @nicklockwood I’m confused. What does the original post have to do with neurodivergent people?

@artemis fuck that shit. We will NEVER surrender to mediocrity.

@artemis

Reducing your vocabulary to prove you're not AI makes it easier for AI to imitate humans.

@artemis

AI is about making computers smarter, but "AI" is about making people dumber.

Two approaches, one goal:

Make computer intelligence equal to human intelligence.

Scientist employ the first, while charlatans employ the second.

@artemis in addition, growing up it was not uncommon for peers to call me a "robot" because of my formal language and undetectable tone of voice. I imagine the use of LLMs is leading to even more dehumanizing comparisons of autistic people to them.

Which is worse. I at least could retort with examples of all the great fictional robots. After all, where and what would Picard be without Data? There's no witty comeback against a comparison to a glorified random text generator.

@artemis
Truly sad the school wants you to prove the negative that you did not use AI.

@artemis @sudaksis So, we have to bring us down to AI's stupidity-level in for the AI to tell us that we are not an AI? Umm, I believe we took a wrong turn somewhere along the road, I mean as a species...

@artemis Damn I remember having to self edit in school because of this very thing

also shout out to the comments... all of you are right and scary

Because the thing is this is how they view education and PEOPLE

something that can be produced on the factory floor and nit picked until "right" 🤔

@artemis This is devoid of any rhyme and reason

@artemis
The horrible thing flying under the radar is a pressumtion of being guilty. C'mon. When kids get used to being guilty until proven innocent, who the majority can easily vote (look at Russia)? How the minority will act, when you are apriori guilty? (crime)

@artemis This stochastical parrot finder should be devoid of life.

Yes, that is a play on Monty Python's Dead Parrot Sketch, where they used the phrase "devoid of life".

@artemis So... it's "reproduce the drivel of your surroundings, or be called a faker", then...?

@artemis bwahhahahaaha! i probably read as a THE CULTURE Hub Mind

@artemis The fact that many autistic coded characters are robots makes this funny, but it is also sad and infuriating.

@artemis aaaaaa this is awful, I use uncommon vocabulary sometimes without even thinking. Thankfully my school doesn't use this crap

@artemis

"Hey, won't this just automate bias and persecution of anyone who deviates from a flawed arbitrary baseline that was created to intentionally embody the most bland, generic, and soulless presentation of human facsimile?"

"Yeah but 💲💲💲"

"Ah I see. My mistake. Carry on."

@artemis@dice.camp What makes a sentence "AI-written" anyway? I remembered reading a report of professor feeding university students' assignment to AI checker, and all those assignment that has less complex grammar and vocabilary are marked as AI-generated, which harmed English-as-second-language students. It's so ridiculous.

@artemis Another form of neurodivergent discrimination. I wonder how may neurodivergents actually wrote that AI engine. Probably a lot.

In verbal conversation, either you have to wait for specific word selection, or endure my replacement curse words for near real time communication. 😁

@artemis @Enalys This proves that both teachers and AI checker devs have low English fluency.

@artemis that was definitely what I was like as a kid, trying out words that I'd discovered. This is infuriating!

@artemis Regard yourself to proob hooman

@artemis this is extra infuriating for me as I have several distinct memories of having to fight my teachers to prove that words I used in writing assignments were a) real words, b) I did know the meaning, and c) had used them correctly. That was bad enough. The kids using gen ai are wrecking themselves, and this sort of crap is going to wreck the ones who don't need it

@artemis
Yeah, that's not an issue with AI. That's an issue with humans.

When I started working in customer service, I was given the advice to reduce the size and complexity of my lexicon. The words I was using caused a perception of condescension. And we don't want the customers to think I'm condescending!

Masking takes many, many forms.

#ActuallyAutistic

@artemis I still remember the eye-rolls of colleagues when I'd suggest that "plagiarism checkers" (Turnitin) were a wrong-headed waste of time. I doubt that will be any different with "AI checkers."

I'm comfortable in the world I occupy, but I've been forced to realized there aren't many people here with me.

@artemis Ah, the thing I learned to do in middle school for survival 😕

@artemis I like to think me mum would've told that school to fsck itself.

@artemis When I was in high school and college, Microsoft Word had a grammar checker that would estimate the grade level of your document. I always tried to get that as high as possible. I guess now my papers would be rejected as cheating.

@artemis i am 100% positive that, if i were still in school, my neurotic ass would be filming a timelapse of myself writing papers and doing homework as proof of life

@artemis Several thoughts, including: (1) memories of the times a teacher assumed I had copied based on my language use, (2) wondering to what extent careful word use is actually "masking" (I agree it's a factor), and (3) fantasies of what I want to do to any teacher who claims to be under the impression that AI checkers are capable of doing what it says on the tin.

@artemis@dice.camp the world should be devoid of ai ngl

@artemis Spending time “un-AI-ing” is some thing that @neil_selwyn addressing in his fantastic podcast.
This simply reality is that educators and by extension parents will find themselves increasingly putting in invisible labor to make human-work conform to automation requirements so that it “just-works,” because it really doesn’t without all the extra human effort.

@artemis I learned a big word when I was eleven or twelve and used it in a report and got accused of copying from the encyclopedia

(if that didn't give away how long ago it was, it was more than 35 years)

the AI thing feels worse though, because my teacher just went on vibes, where now they're using algorithms that are based on other people's vibes

@artemis Shit like this is why I refuse to let my CV get eaten up by "#AI" from recruiting companies!

#NameThemBlameThem

@artemis

Why can't the teacher be bothered to read the essay himself, I asked myself. Since the teacher knows the writing style of the kid it'll be easily to determine whether it's AI or not

@artemis @Nigel_Purchase
The future is devoid of intelligence. Hence the need for artificial kind.

@artemis https://soundcloud.com/farewell-official/devoid-thats-what-i-think

@artemis

Why the fuck did we ever tolerate robots deciding if we're human enough. This started with captchas.

@artemis I hate AI checkers with a passion, they just don't work and if someone forces me to use them I lose all motivation to write because then I have to then modify it until the checker is happy with it

I'm not even in the spectrum, I just have ADHD but it's still a big annoyance

@artemis rule of thumb: if GBoard can't guess the word when you swipe, you're an AI.
I am an AI, btw.

@artemis @eloy

@artemis
Or don’t to show that you’re literate, well read & articulate. Just shows what AI thinks of humans…

@artemis

Comme on disait chez moi :
Il y a l'intelligence artificielle
- et la connerie naturelle !"

As we used to say back home:
There's artificial intelligence
- and natural bullshit!"

@artemis 😭

@artemis I'm a teacher. If a kid gave me an essay with an unusual word, I'd just ask them what it means in general and in this context, and if they know, no problem. Using the automated check before any human check is, well, inhuman.

@artemis AI has to go... it has to be stamped out... there are far more urgent causes than the greed of corporates and CEOs aspiring for eye-watering personal wealth.

@artemis AI checkers are absolute scams, schools are being conned into forcing their students to submit content to be used as AI training datasets against their will, this shit should be illegal

@artemis
Wrong hammer for this type of screw.

@artemis The real scandal here is that teaching the kid how to defeat broken automated checks is a much more useful life skill in 2024 than writing the essay.

@artemis This is middle management nothings trying to make us all talk like middle management nothings, the ergot of corpospeak. Mediocrity enforced by a template.