"The proliferation of references to fake articles threatens to undermine the legitimacy of institutional research across the board" -- by Miles Klee www.rollingstone.com/culture/cult... open access link: archive.ph/qt59c 1/🧵

Le manifeste « Face à l’IA générative, l’objection de conscience » par @atecopol.bsky.social a désormais 3252 signatures ! doi.org/10.58079/1591s

Particularly their position piece "Against the Uncritical Adoption of 'AI' technologies in Academia" bsky.app/profile/iris...

Little signs of humanity that must be preserved. Cheers! @irisvanrooij.bsky.social

Such a useful table. "The term ‘Artificial Intelligence’ (AI) means many things to many people (see Table 1 for different types of meanings)." from Reclaiming AI as a Theoretical Tool for Cognitive Science (Iris van Rooij, Olivia Guest, et al. 2024) link.springer.com/article/10.1...

And finally, ask them why they privilege statistical models in such a simplistic way... They are importantly no different to AI models at the level they are talking, i.e. the "maths" itself — and importantly, they have an extremely sordid history; show them: bsky.app/profile/oliv... 3/3

When people complain about the label "AI" being used instead of I guess "statistical methods" for whatever reason, maybe they assume "cancer can be cured" through such a tool as opposed to through human cognition, feel free to send them this point, if useful: bsky.app/profile/oliv... 1/3

Once you see, you cannot unsee. destaatvanhetweb.nl/2025/07/12/t...

Summarising requires figuring out what is important LLMs can certainly produce a response containing items that *sound* like they *would be* important, but in the end importance is a pretty subjective judgment that's hard to delegate to someone unless your personal ideas of importance overlap

Iris has great references in the thread below. A very short rebuttal is that a summary should include the important facts and ideas, and LLMs have access to neither, they only have access to text. If asked for a summary they will give you a shorter text which probably includes frequently used words.

Interessant gesprek (met wetenschappelijke argumenten) over waarom LLM’s niet kunnen samenvatten.

Published with some minor changes! Please share widely. 😊 Guest, O., Suarez, M., ... & van Rooij, I. (2026). Against the Uncritical Adoption of 'AI' Technologies in Academia. Digital Culture & Education, 16(2), 85–118. doi.org/10.5281/zeno... Journal: www.digitalcultureandeducation.com/volume-162

Ik ben met een groep geestverwanten The Firewall begonnen, omdat ik weiger toe te kijken hoe techgiganten onze democratie ondermijnen. The Firewall vecht terug. Met het vrije woord en de rechtspraak. #StaOpTegenBigTech Word medestander van @thefirewall.eurosky.social www.thefirewall.eu/doe-mee

> In 1895 the Coast and Geodetic Survey grew concerned because Tide-Predicting Machine No. 1 had developed considerable wear [so] decided to construct a new machine > Tide-Predicting Machine No. 2, also known as Old Brass Brains > Old Brass Brains en.wikipedia.org/wiki/Tide-Pr...

Drawn by Dr Seuss over 50 years ago.

It feels many only know my work pushing back against AI, which is very fair — but I have a broader research interest, which also forms a protective belt against such forces, as well as being what I have been thinking about since discovering "cognitive science" in ~2006: 🍃 metatheory.space 1/

CAIL = prerequisite knowledge for a critical perspective, such as to tell apart nonsense hype from true theoretical computer scientific claims. For example, the idea that human-like systems are a sensible or possible goal is the result of circular reasoning and anthropomorphism.

This is so good I'M CRYINGGGGG

Re your bit on Islamic algorithms, I have this illuminated piece where I compare the algorithmic art of the empire to LLMs as political propaganda logicmag.io/supa-dupa-sk...

1️⃣ That's @irisvanrooij.bsky.social's tractable cognition thesis which helps constrain theories 2️⃣ Also something the dynamical systems cognitive scientists get confused about when they think they have rejected computationalism 3️⃣ Still doesn't imply we know or can ever know the computations* 1/

It just doesn't seem sustainable for LLM vendors to keep training on copyrighted materials

ˢᵐᵒˡ ᵖᶦⁿᵉᵃᵖᵖˡᵉ

Useless comment from me, but those fruits looks so good 🌝

for passersby, I would also recommend this recent piece where @emilymbender.bsky.social reflects on how people have (mis)understood and (mis)used the phrase "stochastic parrots" over the years: medium.com/@emilymenonb...

* What this means is that even if we accept / assume computationalism, i.e. that cognition is computationally understandable / describable, we still may never ever be able to write those computations down. The search for them is intractable. See: bsky.app/profile/iris...

“We must protect and cultivate the ecosystem of human knowledge. AI models can mimic the appearance of scholarly work, but they are (by construction) unconcerned with truth — the result is a torrential outpouring of unchecked but convincing-sounding “information”. — as quoted in Guest et al. (2026)

> Noor van der Heijden of the Auteursbond called the scale unprecedented and alarming: “AI models [use] illegally obtained, copyrighted material. Those texts have been stolen from our writers and now such an AI book maker is competing with the robbed writers. We as the Auteursbond are against that”

Ga je eerst even verdiepen in de stad waar je het over hebt. Nijmegen heeft herhaaldelijk een van de grootste noodopvangplekken in Nederland ter beschikking gesteld. in no time veel vrijwilligers ter ondersteuning Ook voldoet Nijmegen ruimschoots aan de spreidingswet.

Welkom in Nijmegen! Geweldig om te zien hoeveel mensen laten zien dat mensen die vluchten voor oorlog en geweld, hier welkom zijn. ❤️

"I am suddenly less interested in their policy on LLMs and more interested in their policy on ethics. Excluding ethical concerns as a basis for policy is what you do when you know your preferred policies are unethical. Any project which accepts such an exclusion should be treated as a threat."

Já passamos por isso antes! Como assim?, por Olivia Guest Tradução parcial de “We’ve been here before!”, de @olivia.science , 18 de fevereiro de 2026. adolfoneto.elixiremfoco.com/blog/posts/2...

I need to start making videos again that teach programming in a natural way, without AI

Now is a period of time when the folks in charge are very big into offering false binaries & forced choices. Everyone else ends up becoming reactive because they don’t feel they can take the time to stop & think, but it’s necessary. It’s an inefficiency dressed up as efficiency. To what end?

6. the extremely unhinged series of claims that without training them on how to be users of such systems that we somehow fail as teachers — truly ludicrous, utterly bizarre, and in fact directly contradicts other industry selling points. (Section 3.6 here doi.org/10.5281/zeno...) 8/n

exactly! bsky.app/profile/oliv...

That’s exactly what will happen. Current frameworks being proposed for education begin with the assumption that students should use whatever systems they’re provided. They leave out a couple of steps before that: should you use them at all, what kind, for what purposes.

I’m concerned about people who want students to critique ‘AI’ systems by making them use those systems.

This is a serious problem that will end w victim blaming & forcing students to take responsibility for things that are not their fault (next in line for culpability: teachers who didn’t teach them well enough to know that something went wrong —everyone but the system & the profiteers).

> teaching students how to question and critique [AI] should not be the panacea for biased, unregulated, and problematic AI. There is a risk that a focus on digital literacies can responsibilise young people in dealing with the significant problems of generative AI. www.ox.ac.uk/news/2025-11...

The coolest thing about programming (not with LLMs, actually coding) is how close it is to spells. When you know how to code, the computer does what you tell it to. It's pleasing and beautiful. Chatbots take away that ability and skill, in totality forever, since they cannot follow instructions.

@jelmerevers.com maybe useful, from US union

In case useful: www.digitalcultureandeducation.com/volume-162-p... bsky.app/profile/oliv...

Universities' "top-down implementation ignores severe institutional risks: imminent workforce displacement, inherent, unfixable data hallucinations, work intensification, systematic intellectual property theft, and widespread student skepticism."

The AAUP is proud to support the Sanders/AOC AI Data Center Moratorium Act because the future of technology must be governed democratically, sustainably, & in service of working people—not corporate profit. ➡️https://www.aaup.org/news/aaup-backs-data-center-moratorium

The AAUP supports this AI Data Center Moratorium because the future of technology must be governed democratically, sustainably, and in service of working people—not corporate profit.

Resistance to reckless implementation of AI is growing. Local communities have successfully blocked data center construction projects & faculty unions have blocked vendor contracts & codified protective language in collective bargaining agreements.

yes, this is very important and relatedly to hold the field(s) to account when they do not take their role seriously as academic disciplines!

I'm happy you mention this. I've been encouraging people who ask: how do we resist 'AI' to use your work in that space as a guide for action. But ALSO, to see cognitive science as a living research program producing insights that refute tech industry reductionism.