#Tag · Kakapo Social

"Over four months, LLM users consistently underperformed at neural, linguistic, and behavioral levels."
https://arxiv.org/abs/2506.08872

arXiv.org

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.

Kevin Russell

@kevinrns@mstdn.social replied · 2 days ago

@lilithsaintcrow

Research shows cross the board cognitive loss using AI.

Using AI is akin to eating lead paint.

The degradation of thinking is comparable, the only question is if it is a permanent loss, like lead.

#ai

Robin Jayasinghe and 1 other boosted

Michał "rysiek" Woźniak · 🇺🇦

@rysiek@mstdn.social · 3 days ago

New blogpost: AI will compromise your cybersecurity posture
https://rys.io/en/181.html

The way “AI” is going to compromise your cybersecurity is not through some magical autonomous exploitation by a singularity from the outside, but by being the poorly engineered, shoddily integrated, exploitable weak point you would not have otherwise had on the inside.

LLM-based systems are insanely complex. And complexity has real cost and introduces very real risk.

1/🧵

#AI #InfoSec

Songs on the Security of Networks

AI will compromise your cybersecurity posture

Yes, “AI” will compromise your information security posture. No, not through some mythical self-aware galaxy-brain entity magically cracking your passwords in seconds or “autonomously” exploiting new

⁂

More from

Michał "rysiek" Woźniak · 🇺🇦

Breschtli aus'm Allgäu boosted

fly-Ing

@flyIng@mastodon.bachgau.social · 3 days ago

https://www.heise.de/news/Analyse-Zeitweise-tausende-sexualisierte-Deepfakes-pro-Stunde-von-Grok-11134092.html
(german language)
It's disgusting. Really disgusting. So AI seems to be a tool made from and used by white male supremacists.
I did'nt study AI LLMs as @Tattie did - yet I'm familiar enough with "AI" to be careful using it - if I ever see an advantage for using an AI compared using my own brain or conventional ways collecting information.
#noAI #AI

heise online

"Abgründe der Verderbtheit": Neue Einblicke in KI-generierte Deepfakes von Grok

Eine Forscherin hat nachgezählt, wie viele sexualisierte Deepfakes Grok pro Stunde generiert. Zudem gibt es Einblicke in das, was die KI nicht öffentlich macht.

fly-Ing

@flyIng@mastodon.bachgau.social · 3 days ago

heise online

"Abgründe der Verderbtheit": Neue Einblicke in KI-generierte Deepfakes von Grok

Eine Forscherin hat nachgezählt, wie viele sexualisierte Deepfakes Grok pro Stunde generiert. Zudem gibt es Einblicke in das, was die KI nicht öffentlich macht.

stux⚡

@stux@mstdn.social · 4 days ago

Currently I’m with my mom for a few days and I noticed her Facebook and Instagram feed..

I think about half is #AI generated slop..

And she has no clue whatsoever

This is such a massive difference from just a few years ago, personally I really think AI slop is a plague, I certainly don’t wanna host it myself

Driftini and 5 others boosted

Michał "rysiek" Woźniak · 🇺🇦

@rysiek@mstdn.social · 6 days ago

Microsoft renamed the whole Office suite to Copilot. 🤣

Please bear in mind that calling "AI" output "slop" makes Satya Nadella sad. Under no circumstances should you call their office suite "Microslop Slopilot". That would just be mean and you're better than that.

*) Copilot might not be your favorite app.

#Microsoft #AI #Slop #Microslop

Screenshot of a fragment of office.com landing page. Text:

The Microsoft 365 Copilot app (formerly Office) lets you create, share, and collaborate all in one place with your favorite apps now including Copilot.* — Screenshot of a fragment of office.com landing page. Text: The Microsoft 365 Copilot app (formerly Office) lets you create, share, and collaborate all in one place with your favorite apps now including Copilot.*

Michał "rysiek" Woźniak · 🇺🇦

@rysiek@mstdn.social · 6 days ago

Microsoft renamed the whole Office suite to Copilot. 🤣

*) Copilot might not be your favorite app.

#Microsoft #AI #Slop #Microslop