Critical AI https://read.dukeupress.edu/critical-ai/issue
@criticalai-journal.bsky.social
5.9K followers 750 following 1.2K posts
Critical AI's new issue is out! https://read.dukeupress.edu/critical-ai/issue Email us at [email protected] Our website and blog: https://criticalai.org/
Posts Media Videos Starter Packs
Reposted by Critical AI https://read.dukeupress.edu/critical-ai/issue
propublica.org
A former DOGE staffer developed an AI tool that hallucinated the size of Veterans Affairs contracts.

For example, it concluded that more than a thousand deals were each worth $34 million, when in fact some were for as little as $35,000.

(Published June)
DOGE Developed Error-Prone AI Tool to “Munch” Veterans Affairs Contracts
We obtained records showing how a Department of Government Efficiency staffer with no medical experience used artificial intelligence to identify which VA contracts to kill. “AI is absolutely the wron...
www.propublica.org
Reposted by Critical AI https://read.dukeupress.edu/critical-ai/issue
lucyosler.bsky.social
Chatbots are explicitly being designed “to elicit intimacy and emotional engagement in order to increase our trust in and dependency on them” (me in Wired magazine 🤭). This designed intimacy combined with AI sycophancy creates serious risks for delusional thinking

www.wired.com/story/ai-psy...
AI Psychosis Is Rarely Psychosis at All
A wave of AI users presenting in states of psychological distress gave birth to an unofficial diagnostic label. Experts say it’s neither accurate nor needed, but concede that it’s likely to stay.
www.wired.com
criticalai-journal.bsky.social
None of us has yet fully read this; but it looks to be an important preprint from @olivia.science + @irisvanrooij.bsky.social

Critical AI literacies means standing back from "AI first" stances that outsource critical activities to opaque systems, flattening and anthropomorphizing H-C relations.
olivia.science
New preprint 🌟 Psychology is core to cognitive science, and so it is vital we preserve it from harmful frames. @irisvanrooij.bsky.social & I use our psych and computer science expertise to analyse and craft:

Critical Artificial Intelligence Literacy for Psychologists. doi.org/10.31234/osf...

🧵 1/
Cover page of Guest, O., & van Rooij, I. (2025, October 4). Critical Artificial Intelligence Literacy for Psychologists. https://doi.org/10.31234/osf.io/dkrgj_v1 Table 1 Guest, O., & van Rooij, I. (2025, October 4). Critical Artificial Intelligence Literacy for Psychologists. https://doi.org/10.31234/osf.io/dkrgj_v1 Table 2 Guest, O., & van Rooij, I. (2025, October 4). Critical Artificial Intelligence Literacy for Psychologists. https://doi.org/10.31234/osf.io/dkrgj_v1
Reposted by Critical AI https://read.dukeupress.edu/critical-ai/issue
edzitron.com
10K Word Premium: OpenAI is another unprofitable, desperate AI startup building products on its own models, spending $2.60 to make $1. Sora 2 is a reckless attempt to grow ChatGPT that costs them at least $5 per generated video. They need to raise $20bn+ before EoY.
www.wheresyoured.at/sora2-openai/
OpenAI Is Just Another Boring, Desperate AI Startup
What is OpenAI? I realize you might say "a foundation model lab" or "the company that runs ChatGPT," but that doesn't really give the full picture of everything it’s promised, or claimed, or leaked t...
www.wheresyoured.at
Reposted by Critical AI https://read.dukeupress.edu/critical-ai/issue
Reposted by Critical AI https://read.dukeupress.edu/critical-ai/issue
edzitron.com
I cannot express how wild this quote is. This is a top tech analyst at a well-regarded firm saying that there is not enough capital to actually pay for all the stuff that NVIDIA and OpenAI has promised. It is so important that everybody realizes that *there is not enough money to do this*
edzitron.com
I spoke with analyst Gil Luria at D.A. Davidson, asking if the capital existed to build OpenAI's promised 17GW of data centers.

He said "of course there isn't enough capital for all of this," but "enough capital to do this for a at least a little while longer."

www.wheresyoured.at/openai-onetr...
Shortly before publishing this newsletter, I spoke with analyst Gil Luria, Managing Director and Analyst at D.A. Davidson, and asked him whether the capital was there to build the 17 Gigawatts of capacity that OpenAI has promised.

He said the following:

No of course there isn't enough capital for all of this. Having said that, there is enough capital to do this for a at least a little while longer.
There is quite literally not enough money to build what OpenAI has promised.
Reposted by Critical AI https://read.dukeupress.edu/critical-ai/issue
talkingpointsmemo.com
Oklahoma Schools Chief Ryan Walters made news almost daily for controversial actions threatening to take over the Tulsa schools and mandating Trump-endorsed classroom Bibles. He's leaving his post to run an anti-teachers union think tank.
talkingpointsmemo.com/news/scandal...
Scandal-Plagued Oklahoma Schools Chief Ryan Walters Steps Down to Lead Anti-Teachers Union Group
This story first appeared at The 74, a nonprofit news site covering...
talkingpointsmemo.com
criticalai-journal.bsky.social
Compare to driving: you don't learn it immediately and, for many, you get better w/ time and experience. You might add a new skill late in life (parallel parking a a dense city).

But if you're handed the keys and switch on the autopilot button you likely won't ever learn any of these things.
criticalai-journal.bsky.social
The question is not wholly to do with quality but with the process. The mantra in teaching of writing (with which I agree) is process not product.

Not every student who takes an intro course will achieve these things overnight, ofc. But they won't even try if they outsource to automation. /a
criticalai-journal.bsky.social
I think "cognitive debt" is an aspect of the decline in self-realization.
Lack of attention span is huge, amplifying weak reading skills, lack of ability to evaluate evidence or use it effectively (w/ attribution) to support one's own claims, declining interest in or ability to research robustly
criticalai-journal.bsky.social
If thinking for yourself and cultivating the ability to express your own ideas in writing is an "obsolete skill" then heaven help us.

Although you didn't mean it this way, comparisons to calculators are flawed b/c we STILL teach students the basics of arithmetic.
criticalai-journal.bsky.social
For evidence that this is the actual end game--go no further than the @nytimes.com where Natasha Singer, @tressiemcphd.bsky.social and Jessica Grose are publishing strong work from teacher and parent perspectives. /end
criticalai-journal.bsky.social
Finally the essay ignores evidence that industry WANTS students addicted to bots; wants to monetize the ELIZA effect. They don't care about cognitive debt!

Maybe reach out to other tech reporters outside the narrow channel of bothsidesing edtech

@bcmerchant.bsky.social or @karenhao.bsky.social /8
criticalai-journal.bsky.social
As for the below, the logic of the last sentence seems a lot like the bromides against gun control.

Yes, people COULD use chatbots in ideal ways, just like they COULD use social media w/o becoming addicted or harmed.

/7
criticalai-journal.bsky.social
Suggestions:

Can we drop the guise of "panic" which is a way of talking down to readers.

Can we drop tendentious language like "brain rot."

/6
criticalai-journal.bsky.social
What exactly does that mean: Que sera sera?

Look: active learning really is a thing; and not just in the lab.

Moreover, the implication that the "cognitive debt" documented by Kosmyna and colleagues will lessen as "people adapt" is McMurtrie's own theory (no evidence).

/5
criticalai-journal.bsky.social
McMurtrie provides good quotations from experts, but her own conclusions are weak.

Kosmyna et al's evidence that bot-users remembered less about their writing months later, is followed this gem:
"But life is not a laboratory. Learning is a multifaceted process. People adapt to technology." /4
criticalai-journal.bsky.social
Um, no, there is no evidence of "panic" over "brain rot" obscuring complexity.

What there is in this article is a false opposition between "panic" and "complexity" which tries to tie up complexity in a bow. /3
criticalai-journal.bsky.social
Yes: learning IS complex; also the Python study wasn't proffered in DIRECT contrast to the Kosmyna's on writing.

That said, as so often in ed tech reporting, McMurtrie's article leans toward, "It's complicated" (which it is) under cover of a non-existent strawman ("panic" over "brain rot"). /2
criticalai-journal.bsky.social
Chronicle author @bethmcmurtrie.bsky.social discussed @nataliyakosmyna.bsky.social's study of "cognitive debt" re essay-writing with ChatGPT. McMurtrie then writes:
criticalai-journal.bsky.social
read.dukeupress.edu/critical-ai/... In her intro column to CAI 3.1, editor Lauren Goodlad talks about teaching critical AI literacies including new DESIGN JUSTICE LABS web resources, and using of COMPARATIVISM as a creative counter to "AI first" approaches to education.
Humanist in the Loop: Teaching Critical AI Literacies, Episode 1 | Critical AI | Duke University Press
read.dukeupress.edu