Laura

@codingcatgirl@chaos.social
797 Followers
83 Following
6.2K Posts

cuddly trans lesbian, doing Python and IT security. Made @c3nav.

Generative AI trash will be blocked on sight

this is my less interesting German account. the other one is not publicly linked.

Pronounsshe/her, fae/faer
Websitehttps://codingcatgirl.de
GitHubhttps://github.com/codingcatgirl

Wisst ihr noch wenn irgendwelche Spinner nur mit Leuten die keine mRNA Impfung gekriegt haben interagieren wollten weil sie noch nicht "verseucht" sind oder sowas?

Das, aber mit LLMs, könnte die tatsächliche Zukunft werden…

“Over four months, LLM users consistently underperformed at neural, linguistic, and behavioral levels.”

https://www.media.mit.edu/publications/your-brain-on-chatgpt/

Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task – MIT Media Lab

 This study explores the neural and behavioral consequences of LLM-assisted essay writing. Participants were divided into three groups: LLM, Search Engine…

MIT Media Lab

Since I've left my last job, I've been thinking about the guy who used me as an alternative to ChatGPT whenever he hit a problem that he couldn't vibe code the answer out of at work.

He basically rotted his own brain by compulsively using ChatGPT in lieu of actually thinking with most any of the projects he was working on. Instead of taking the time to read through code in our framework, look up documentation, or do any sort of debugging, he instead just begged and pleaded with ChatGPT to try and get somewhere because "it was faster." Basically just really hammering his brain with the Programmer's Slot Machine. (@davidgerard wrote a really good article here about this specific gambling addiction angle here. I highly highly recommend reading/watching the corresponding YouTube video:
https://pivot-to-ai.com/2025/06/05/generative-ai-runs-on-gambling-addiction-just-one-more-prompt-bro/ )

Back to the story; When that wasn't working, which was a significant portion of the time, he'd then just turn and use me as a "more informed alternative" to ChatGPT.

I worked fully remote and the majority of our interactions was via a Teams chat. which apparently crossed some wires in his monkey brain and made him start just... Basically verbally barraging me like he would with the company ChatGPT instance. No thoughts at all, just an immediate process of:

- Ask vague question
- Get guess for an answer with a request for more details
- Try applying the guess blindly without thinking if it's applicable at all
- Have it not work and just report back that it didn't work.
- No follow-up details, no further explanation of what was going on or what he's trying to do. Nothing added past the original vague situation
- If lucky, I might get a screenshot of part of the error, meticulously sliced before it gave something useful in the output because he stopped reading error output to things and made no attempt to understand it. (Why? ChatGPT can do that part!)
- Rinse and Repeat until I get fed up and get into a call with him
- Fix the thing in less than a minute, pointing out that he should have been able to tell what was wrong almost immediately if he actually dropped a break-point and debugged the code at *literally any point* along the way
- Fuck off immediately after getting his fix, no thank you or anything
- start the process anew the following day when he vibe coded himself into a corner all over again

I literally had to go to leadership and make them have a talk with him and get him to leave me the fuck alone at work, after repeated attempts to establish boundaries about it, due to how much time it sucked out of me being able to work on other projects. Effectively just doubling up my work and slamming me with burn out right at the start of the year for absolutely no reason other than his belligerent insistence to just Not Do His Job Without His Hand Being Held By A Chat Window.

It rapidly went from a "He sometimes asks informed questions that I can answer and help him with. I enjoy working with him" to "The dude isn't even trying in the slightest and is now basically offloading his work onto me because he broke his capacity to actually do work independently of an external chat window. I fucking hate him and I hope he gets in a car wreck so I can get a break from the bleakness of dealing with him every goddamn morning"

ChatGPT has basically just been an absolute blight for me since it's inception. Going from the team being generally pro-crypto to intensely pro-genAI/LLM because their favorite scammers (er.. I mean YouTubers) had them hooked on a fantasy of some day making it Big by jumping from one Hype cycle to the next. I sincerely was very close to just finding an entirely different career path altogether because of just how incredibly shitty it was working with that team on just about anything, but lacking the job experience on the resume to land someplace else.

Nobody wanted to be an actual expert, nobody really wanted to learn anything. They had their degree and ChatGPT, which means they learned all they ever will need. ...While working in an industry that tends to re-invent itself every half decade or so while half-assing solutions with an outsourced bullshit generator. 🫠

All in the name of "Well it got me from point A to point B faster." and leaving it at that, despite taking significantly longer than they should have from the get go over it.

I've seen and lived what an AI Fueled future looks like:
Mediocre men harassing their talented and likely autistic peers until their peers just up and fuckin leave to a different organization out of frustration and exhaustion.

I think down the road, we'll be able to measure the negative impact using LLMs has on people's cognitive faculties by comparing it to horse kicks to the head, and only be exaggerating it by a little bit.

Generative AI runs on gambling addiction — just one more prompt, bro!

You’ll have noticed how previously normal people start acting like addicts to their favourite generative AI and shout at you like you’re trying to take their cocaine away. Matthias Döpm…

Pivot to AI

[…]
"I don't see why you keep bringing up the MK4S, they are nothing alike (assuming it is not enclosed)"

[it's literally the exact same print head sitting on two different motion systems]

"It is enclosed, like i said before, and it works fine"

"Ah, yeah, we generally don't recommend printing TPU inside an enclosure."

"As i said earlier, the MK4S in the enclosure works fine, with a higher ambient temperature than the Core One, which is not working fine"

…????

What the fuck is going on

"It's working well in my MK4S, not in my Core One"

"This might be because the Core One is enclosed"

"My MK4S is also enclosed and it'S working fine"

"Yes, that's why we generally recommend to print TPU without an enclosure"

"But my MK4S is sitting in an enclosure and printing fine."

"I understood that perfectly fine the first time, thank you very much"

[ … ]

Like this is so absurd i expect them to tell me to feed the filament into the fan next

I asked to be transfered to a new agent, never had to do that before, only to have her tell me i should losen the idler screws as far as i can without them getting fully disengaged from the thread. The resulting completely lose idler lever looked so wrong that i sent a video of it, and she replied to by saying that that looks too lose… ?!?!? why give me such instructions then what the heck is going on?

This makes no sense. This isn't incompetence, it's just… nonsense…?! Do they have a gas leak?

What the fuck happened to Prusa support, i just talked to someone who kept talking to me not like AI but like fucking Cleverbot, full of typos and randomly going off tangents based on words in my messages, i told him things about my setup opposite times, he got annoyed at me repeating myself and then still assumed the opposite of what i told him, i was genuinely concerned he was having a stroke or was dehydrated…

Wissenschaftler:innen haben herausgefunden: Wer ChatGPT oder andere Bullshit-Generatoren nutzt, verblödet innerhalb kurzer Zeit.

https://arxiv.org/abs/2506.08872

#LLM #GenAI #ChatGPT #Copilot #Verblödung

Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task

This study explores the neural and behavioral consequences of LLM-assisted essay writing. Participants were divided into three groups: LLM, Search Engine, and Brain-only (no tools). Each completed three sessions under the same condition. In a fourth session, LLM users were reassigned to Brain-only group (LLM-to-Brain), and Brain-only users were reassigned to LLM condition (Brain-to-LLM). A total of 54 participants took part in Sessions 1-3, with 18 completing session 4. We used electroencephalography (EEG) to assess cognitive load during essay writing, and analyzed essays using NLP, as well as scoring essays with the help from human teachers and an AI judge. Across groups, NERs, n-gram patterns, and topic ontology showed within-group homogeneity. EEG revealed significant differences in brain connectivity: Brain-only participants exhibited the strongest, most distributed networks; Search Engine users showed moderate engagement; and LLM users displayed the weakest connectivity. Cognitive activity scaled down in relation to external tool use. In session 4, LLM-to-Brain participants showed reduced alpha and beta connectivity, indicating under-engagement. Brain-to-LLM users exhibited higher memory recall and activation of occipito-parietal and prefrontal areas, similar to Search Engine users. Self-reported ownership of essays was the lowest in the LLM group and the highest in the Brain-only group. LLM users also struggled to accurately quote their own work. While LLMs offer immediate convenience, our findings highlight potential cognitive costs. Over four months, LLM users consistently underperformed at neural, linguistic, and behavioral levels. These results raise concerns about the long-term educational implications of LLM reliance and underscore the need for deeper inquiry into AI's role in learning.

arXiv.org
"TRI RE11" sind jetzt auch zwei Begriffe die man eher nicht nebeneinander erwartet