Hausgeist

Dear #hivemind, I could use some help.

I'm looking for examples/papers/essays/anything that, in your opinion, perfectly exemplifies the detrimental impact of AI use for learning and critical thinking—specifically (but not exclusively) regarding software engineering or in IT.

Context: I've been asked to give a short introduction on the topic of #AI as part of group discussion within my team. The discussion will center around how we, as a software team, can (learn how to) use AI to ImPrOvE PrOdUcTiViTy (sigh, I know). The people organizing the discussion are riding the hype train pretty hard and the org is pushing heavily for it, so my explicit intent is to create space for critical/opposing thoughts/voices. I have about 15 minutes, so I want to use the time wisely. I won't be able to make a complete case on what garbage we're dealing with but at least make a point of not staying silent.

Give me all the links please. And maybe a boost, too! ❤️

Mar 1 at 1:14pmWeb
Show threadAnna e sóMar 1

@hausgeist I’m a researcher at a free software laboratory and a couple of students are developing a study on this very topic. They’re drawing from two ideas from educational psychology: Desirable
Difficulty and Cognitive Load Theory. This is the paper that made them want to investigate that in software development: https://www.researchgate.net/publication/392560878_Your_Brain_on_ChatGPT_Accumulation_of_Cognitive_Debt_when_Using_an_AI_Assistant_for_Essay_Writing_Task

@ploum’s blog post on an experiment conducted in a class is also worth reading: https://ploum.net/2026-01-19-exam-with-chatbots.html

Show threadHausgeistMar 1
@anna Excellent thank you!!
Show threadAlexandre OlivaMar 1
https://sightlessscribbles.com/the-colonization-of-confidence/ is about writing, not software development, but I feel it can be useful as the confidence undermining and outsourcing of autonomy apply
The Colonization of Confidence., Sightless Scribbles

A fabulously gay blind author.

Show threadImaginary RobotsMar 1

@hausgeist here's a great one to use by anthropic (ie one of the places selling the thing) that shows - at best - mixed results in the coding space: https://www.anthropic.com/research/AI-assistance-coding-skills

And here's one from the Harvard Business Review about how LLM use intensifies work rather than reducing it: https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it

How AI assistance impacts the formation of coding skills

Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

Show threadElizabeth AyerMar 1

@hausgeist

From Anthropic itself: "We find that AI use impairs conceptual understanding, code reading, and debugging abilities, without delivering significant efficiency gains on average."

How AI Impacts Skill Formation https://arxiv.org/abs/2601.20245
(ed: just spotted this above, sorry to repeat!)

Also some references here https://www.psychologytoday.com/us/blog/experimentations/202602/accomplishment-hallucination-when-the-tool-uses-you

Good luck!!!!

How AI Impacts Skill Formation

AI assistance produces significant productivity gains across professional domains, particularly for novice workers. Yet how this assistance affects the development of skills required to effectively supervise AI remains unclear. Novice workers who rely heavily on AI to complete unfamiliar tasks may compromise their own skill acquisition in the process. We conduct randomized experiments to study how developers gained mastery of a new asynchronous programming library with and without the assistance of AI. We find that AI use impairs conceptual understanding, code reading, and debugging abilities, without delivering significant efficiency gains on average. Participants who fully delegated coding tasks showed some productivity improvements, but at the cost of learning the library. We identify six distinct AI interaction patterns, three of which involve cognitive engagement and preserve learning outcomes even when participants receive AI assistance. Our findings suggest that AI-enhanced productivity is not a shortcut to competence and AI assistance should be carefully adopted into workflows to preserve skill formation -- particularly in safety-critical domains.

arXiv.org
Show threadjraMar 1

@hausgeist Here's an extremely detailed position post by @catileptic regarding the Romanian Ministry of Education's promotion of GenAI in schools.

https://catileptic.tech/posts/no-genai-in-schools/

Original fedi post: https://chaos.social/@catileptic/115683779728828536

No genAI in Schools

This is catileptic. I move slow and fix things. I'm a tech activist, hacker, and a community-building person.