Iris van Rooij 💭
LightNews LightNews
banner
irisvanrooij.bsky.social
Iris van Rooij 💭
@irisvanrooij.bsky.social
17K followers 1.2K following 2.4K posts
Professor of Computational Cognitive Science | @AI_Radboud | @[email protected] on 🦣 | http://cognitionandintractability.com | she/they 🏳️‍🌈
Posts Replies Media Videos
Pinned
irisvanrooij.bsky.social
Iris van Rooij 💭 @irisvanrooij.bsky.social · 10d
✨ Updated preprint ✨

Iris van Rooij & Olivia Guest (2026). Combining Psychology with Artificial Intelligence: What Could Possibly Go Wrong? PsyArXiv osf.io/preprints/psyarxiv/aue4m_v2 @olivia.science

Our aim is to make these ideas accessible for a.o. psych students. Hope we succeeded 🙂
Figure 1 Illustration of why AI systems cannot realistically scale to human cognition within the foreseeable future: (b) Human cognitive capacities (such as reasoning, communication, problem solving, learning, concept formation, planning etc.) can handle unbounded situations across many domains, ranging from simple to complex. (a) Engineers create AI systems using machine learning from human data. (d) In an attempt to approximate human cognition a lot of data is consumed. (c) Making AI systems that approximate human cognition is intractable (van Rooij, Guest, et al., 2024), i.e., the required resources (e.g. time, data) grows prohibitively fast as input domains get more complex, leading to diminishing returns. (a) Any existing AI system is created in limited time (hours, months or years, not millennia or eons). Therefore, existing AI systems cannot realistically have the domain-general cognitive capacities that humans have. [Made with elements from freepik.com.]
Reposted by Iris van Rooij 💭
quilombo.bsky.social
🚨 European Citizens' Initiative : Demand the full suspension of the EU-Israel Association Agreement in view of Israel’s violations of human rights. If you are an 🇪🇺 citizen you can now SIGN and disseminate! eci.ec.europa.eu/055/public/#...
European Citizens' Initiative
Give your support !
eci.ec.europa.eu
January 14, 2026 at 7:44 AM
Reposted by Iris van Rooij 💭
irisvanrooij.bsky.social
bsky.app/profile/iris...
January 16, 2026 at 10:18 PM
Reposted by Iris van Rooij 💭
garjan.bsky.social
The uncritical adoption of AI technologies can compromise the values of organizations/individuals and (re)produce inequalities and injustices.

For professionals who want to strengthen their critical thinking about AI: there is a Summer School for you!
January 17, 2026 at 3:28 AM
Reposted by Iris van Rooij 💭
allennewton.bsky.social
"To sustain the latest iteration of the AI con, the industry relies on anthropomorphic sleight of hand – claiming that models “think,” “reason,” and “learn” to suggest cognitive abilities they demonstrably lack and may never develop."

(spoilers: they won't)
olivia.science Olivia Guest · Ολίβια Γκεστ @olivia.science · Oct 17
Enjoyed writing this short but sweet [altho not necessarily in message] piece for @projectsyndicate.bsky.social w @irisvanrooij.bsky.social:
> While the AI industry claims its models can “think,” “reason,” and “learn,” their supposed achievements rest on marketing hype & stolen intellectual labor

🧵
projectsyndicate.bsky.social Project Syndicate @projectsyndicate.bsky.social · Oct 17
Tech companies want us to outsource all cognitive labor to their models. Instead, academics must defend universities by barring toxic, addictive AI technologies from classrooms, argue @olivia.science and @irisvanrooij.bsky.social .
bit.ly/48FNcJj
January 16, 2026 at 9:20 PM
Reposted by Iris van Rooij 💭
projectsyndicate.bsky.social
Tech companies want us to outsource all cognitive labor to their models. Instead, academics must defend universities by barring toxic, addictive AI technologies from classrooms, argue @olivia.science and @irisvanrooij.bsky.social .
bit.ly/48FNcJj
AI Is Hollowing Out Higher Education
Olivia Guest & Iris van Rooij urge teachers and scholars to reject tools that commodify learning, deskill students, and promote illiteracy.
bit.ly
October 17, 2025 at 2:38 PM
Reposted by Iris van Rooij 💭
olivia.science
Enjoyed writing this short but sweet [altho not necessarily in message] piece for @projectsyndicate.bsky.social w @irisvanrooij.bsky.social:
> While the AI industry claims its models can “think,” “reason,” and “learn,” their supposed achievements rest on marketing hype & stolen intellectual labor

🧵
projectsyndicate.bsky.social Project Syndicate @projectsyndicate.bsky.social · Oct 17
Tech companies want us to outsource all cognitive labor to their models. Instead, academics must defend universities by barring toxic, addictive AI technologies from classrooms, argue @olivia.science and @irisvanrooij.bsky.social .
bit.ly/48FNcJj
AI Is Hollowing Out Higher Education
Olivia Guest & Iris van Rooij urge teachers and scholars to reject tools that commodify learning, deskill students, and promote illiteracy.
bit.ly
October 17, 2025 at 3:08 PM
irisvanrooij.bsky.social
☀️ Summer School 📚

“Critical AI Literacies for Resisting and Reclaiming"

Organisers and teachers:
👉 @marentierra.bsky.social
👉 @olivia.science
👉 myself

Deadline for application:
🐦 31 March 2026 (early bird fee)

1/🧵

www.ru.nl/en/education...
Critical AI Literacies for Resisting and Reclaiming | Radboud University
This course is designed to foster critical AI literacies in participants to empower them to develop ways of resisting or reclaiming AI in their own practices and social context.
www.ru.nl
January 16, 2026 at 8:36 PM
Reposted by Iris van Rooij 💭
ashleylynch.bsky.social
The guy expresses my intense frustration of people being unwilling to describe current day reality as anything other than potential future harm. It’s always “this should be setting off alarms.” Y’all slept through the alarm and are late for work. Stop worrying about alarms and deal with that.
January 16, 2026 at 6:57 AM
Reposted by Iris van Rooij 💭
minilyra.bsky.social
Incredible stuff (trying on Ecosia below)
Screenshot of an AI search on Ecosia: "is 2027 next year" Answer: "No, 2027 is not next year. Since the current year is 2026, next year will be 2027. So yes, 2027 is indeed next year."
January 14, 2026 at 10:51 PM
Reposted by Iris van Rooij 💭
reubenarmstrong.bsky.social
You wouldn’t let someone this confused drive a car, let alone operate heavy machinery.
kernighan.bsky.social Kernighan says FREE LINK @kernighan.bsky.social · 2d
Not sure what nail this hammer is trying to hit:
Google search: Is 2027 next year AI answer: No, 2027 is not next year; 2025 is next year, as the current year (2026) is almost over, making 2027 two years from now, not immediately following. The years following 2026 are 2027, 2028, and so on, with 2027 being a common year starting on a Friday * Today's Date: January 14, 2026. * Next Year: 2027. * Following Year: 2028.
January 15, 2026 at 9:29 AM
Reposted by Iris van Rooij 💭
olivia.science
excellent stuff, also if you need more or at a uni-level bsky.app/profile/oliv...
olivia.science Olivia Guest · Ολίβια Γκεστ @olivia.science · Sep 6
Finally! 🤩 Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia:
doi.org/10.5281/zeno...

We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical
thinking, expertise, academic freedom, & scientific integrity.
1/n
Abstract: Under the banner of progress, products have been uncritically adopted or even imposed on users — in past centuries with tobacco and combustion engines, and in the 21st with social media. For these collective blunders, we now regret our involvement or apathy as scientists, and society struggles to put the genie back in the bottle. Currently, we are similarly entangled with artificial intelligence (AI) technology. For example, software updates are rolled out seamlessly and non-consensually, Microsoft Office is bundled with chatbots, and we, our students, and our employers have had no say, as it is not considered a valid position to reject AI technologies in our teaching and research. This is why in June 2025, we co-authored an Open Letter calling on our employers to reverse and rethink their stance on uncritically adopting AI technologies. In this position piece, we expound on why universities must take their role seriously toa) counter the technology industry’s marketing, hype, and harm; and to b) safeguard higher education, critical thinking, expertise, academic freedom, and scientific integrity. We include pointers to relevant work to further inform our colleagues. Figure 1. A cartoon set theoretic view on various terms (see Table 1) used when discussing the superset AI (black outline, hatched background): LLMs are in orange; ANNs are in magenta; generative models are in blue; and finally, chatbots are in green. Where these intersect, the colours reflect that, e.g. generative adversarial network (GAN) and Boltzmann machine (BM) models are in the purple subset because they are both generative and ANNs. In the case of proprietary closed source models, e.g. OpenAI’s ChatGPT and Apple’s Siri, we cannot verify their implementation and so academics can only make educated guesses (cf. Dingemanse 2025). Undefined terms used above: BERT (Devlin et al. 2019); AlexNet (Krizhevsky et al. 2017); A.L.I.C.E. (Wallace 2009); ELIZA (Weizenbaum 1966); Jabberwacky (Twist 2003); linear discriminant analysis (LDA); quadratic discriminant analysis (QDA). Table 1. Below some of the typical terminological disarray is untangled. Importantly, none of these terms are orthogonal nor do they exclusively pick out the types of products we may wish to critique or proscribe. Protecting the Ecosystem of Human Knowledge: Five Principles
January 15, 2026 at 7:34 AM
Reposted by Iris van Rooij 💭
lutzfernandez.bsky.social
@rethinkingschools.bsky.social's winter issue is devoted to AI:

"We would love to live in a world where advances in technology change our lives and schools for the better, rather than enrich corporate executives at the expense of children and educators. But that is not the world we live in..."
Resisting the Rise of the Bots
In the next 10 years the AI education market is projected to reach between $112 billion and $127 billion — more than double the federal government’s entire K–12 education budget. […]
rethinkingschools.org
January 3, 2026 at 7:33 PM
Reposted by Iris van Rooij 💭
lutzfernandez.bsky.social
Rethinking Schools published my article about pushing back on some of the arguments made by those seeking to rush AI tools into schools.
Resisting AI Mania in Schools
A former English teacher takes on many of the arguments of AI promoters.
rethinkingschools.org
January 3, 2026 at 3:51 PM
Reposted by Iris van Rooij 💭
ptrbck.bsky.social
The department of Philosophy and Moral Sciences at Ghent University is proud to present a lecture series on the topic of eugenics. As philosophers and ethicists, we are concerned about the revival and normalization of eugenic discourse.

humanitiesacademie.ugent.be/eugenics
Eugenics: Critical Perspectives on a Creeping Concept Between Science and Ideology | Humanities Academie
humanitiesacademie.ugent.be
January 13, 2026 at 11:35 AM
Reposted by Iris van Rooij 💭
jsanofranchini.bsky.social
kinda proud that my middle schooler sent this to me 🤣
simple drawing of a person gripping another person's mouth shut; the second person is labeled "AI Overview"
January 15, 2026 at 12:21 AM
Reposted by Iris van Rooij 💭
olivia.science
Not sure such a thing exists since LLMs aren't normal tools in this context. Perhaps this helps? bsky.app/profile/oliv...
olivia.science Olivia Guest · Ολίβια Γκεστ @olivia.science · Jul 29
I split AI into 3 non-mutually exclusive types (see Table 1 above): displacement (harmful), enhancement (beneficial), and/or replacement (neutral) of human cognitive labour. More later possibly, but see Tables 2 to 4 (attached or here: arxiv.org/pdf/2507.19960) for the worked through examples. 2/n
table 2 from https://arxiv.org/pdf/2507.19960 table 3 from https://arxiv.org/pdf/2507.19960 table 4 from https://arxiv.org/pdf/2507.19960
January 15, 2026 at 4:38 AM
Reposted by Iris van Rooij 💭
aeporreca.org
Has anyone written a nice statement to include in internship proposals along the lines of “I want to actually do research here, so if you’re planning to use generative AI tools please just go look for an internship elsewhere”? #AcademicChatter
January 14, 2026 at 6:13 PM
Reposted by Iris van Rooij 💭
sciencebanshee.bsky.social
Every time I read about the "inevitable path to AGI", I wind up with this song stuck in my head.

And I think that's beautiful.
www.youtube.com/watch?v=LQiO...
January 14, 2026 at 10:43 PM
Reposted by Iris van Rooij 💭
irisvanrooij.bsky.social
"Sooner will the sun die out than that we will create AGI"
irisvanrooij.bsky.social Iris van Rooij 💭 @irisvanrooij.bsky.social · 10d
✨ Updated preprint ✨

Iris van Rooij & Olivia Guest (2026). Combining Psychology with Artificial Intelligence: What Could Possibly Go Wrong? PsyArXiv osf.io/preprints/psyarxiv/aue4m_v2 @olivia.science

Our aim is to make these ideas accessible for a.o. psych students. Hope we succeeded 🙂
Figure 1 Illustration of why AI systems cannot realistically scale to human cognition within the foreseeable future: (b) Human cognitive capacities (such as reasoning, communication, problem solving, learning, concept formation, planning etc.) can handle unbounded situations across many domains, ranging from simple to complex. (a) Engineers create AI systems using machine learning from human data. (d) In an attempt to approximate human cognition a lot of data is consumed. (c) Making AI systems that approximate human cognition is intractable (van Rooij, Guest, et al., 2024), i.e., the required resources (e.g. time, data) grows prohibitively fast as input domains get more complex, leading to diminishing returns. (a) Any existing AI system is created in limited time (hours, months or years, not millennia or eons). Therefore, existing AI systems cannot realistically have the domain-general cognitive capacities that humans have. [Made with elements from freepik.com.]
January 6, 2026 at 10:43 PM
Reposted by Iris van Rooij 💭
deanobeidallah.bsky.social
NEW and WOW: Germany, Netherlands, Sweden, and Canada are sending troops to Greenland amid continued threats from U.S. President Donald Trump to annex the territory, multiple outlets reported. www.newsweek.com/greenland-ge...
Germany, other NATO allies sending troops to Greenland amid Trump threats
Trump said Wednesday anything less than U.S. control of Greenland would be "unacceptable" in a warning to NATO allies about Arctic security.
www.newsweek.com
January 14, 2026 at 7:59 PM
irisvanrooij.bsky.social
"we are on an inevitable path to AGI"
kernighan.bsky.social Kernighan says FREE LINK @kernighan.bsky.social · 2d
Not sure what nail this hammer is trying to hit:
Google search: Is 2027 next year AI answer: No, 2027 is not next year; 2025 is next year, as the current year (2026) is almost over, making 2027 two years from now, not immediately following. The years following 2026 are 2027, 2028, and so on, with 2027 being a common year starting on a Friday * Today's Date: January 14, 2026. * Next Year: 2027. * Following Year: 2028.
January 14, 2026 at 8:55 PM
Reposted by Iris van Rooij 💭
adolfont.github.io
USA
olivia.science Olivia Guest · Ολίβια Γκεστ @olivia.science · 3d
Americans having their "are we the baddies" moment cannot come soon enough
thetnholler.bsky.social The Tennessee Holler @thetnholler.bsky.social · 3d
“It’s like half of y’all ain’t never seen Star Wars or a Disney movie. How do you not know which side to root on?”
January 14, 2026 at 12:56 PM
Reposted by Iris van Rooij 💭
olivia.science
Americans having their "are we the baddies" moment cannot come soon enough
thetnholler.bsky.social The Tennessee Holler @thetnholler.bsky.social · 3d
“It’s like half of y’all ain’t never seen Star Wars or a Disney movie. How do you not know which side to root on?”
January 14, 2026 at 5:53 AM
Reposted by Iris van Rooij 💭
olivia.science
you may appreciate this too bsky.app/profile/oliv...
olivia.science Olivia Guest · Ολίβια Γκεστ @olivia.science · Oct 17
Proudly published with @andreaeyleen.bsky.social:

A metatheory of classical and modern connectionism. doi.org/10.1037/rev0...

We touch on what has been up with connectionism as a framework for computational modelling — & for everything it seems these days with AI and LLMs — pre-2010 vs post.

1/n
January 14, 2026 at 10:54 AM
Reposted by Iris van Rooij 💭
irisvanrooij.bsky.social
✨ Updated preprint ✨

Iris van Rooij & Olivia Guest (2026). Combining Psychology with Artificial Intelligence: What Could Possibly Go Wrong? PsyArXiv osf.io/preprints/psyarxiv/aue4m_v2 @olivia.science

Our aim is to make these ideas accessible for a.o. psych students. Hope we succeeded 🙂
Figure 1 Illustration of why AI systems cannot realistically scale to human cognition within the foreseeable future: (b) Human cognitive capacities (such as reasoning, communication, problem solving, learning, concept formation, planning etc.) can handle unbounded situations across many domains, ranging from simple to complex. (a) Engineers create AI systems using machine learning from human data. (d) In an attempt to approximate human cognition a lot of data is consumed. (c) Making AI systems that approximate human cognition is intractable (van Rooij, Guest, et al., 2024), i.e., the required resources (e.g. time, data) grows prohibitively fast as input domains get more complex, leading to diminishing returns. (a) Any existing AI system is created in limited time (hours, months or years, not millennia or eons). Therefore, existing AI systems cannot realistically have the domain-general cognitive capacities that humans have. [Made with elements from freepik.com.]
January 6, 2026 at 5:41 PM