To put this 100 kW in context, prototype and near-future generation "AI" datacenter racks range up to 1 MW each. For example, the prototype on the right (NVL576 Rubin Ultra) is a 600 kW rack. You'd need six of the reactors on the left to power that one rack. www.tomshardware.com/pc-component...

This is the secret main essential true 250th anniversary event: the brave voice of democracy and justice; the brutal forces of silencing; it's all here.

Lots of panels keep telling sustainability and consumer communities that data centers are neutral or reduce energy rates, but facts on the ground keep disputing that. The same people get selected for the next panel.

Everything in this thread. And also... Why is this standard only ever applied to scientists whose work is beneficial for progressive causes, but NEVER to those backing eugenics, tobacco, carcinogenic chemicals, and crude oil? Wherefore the double standard and coddling of conservative aims?

Mauritania still practices a slavery remarkably similar to what we had in the US--lifelong, race-based, inherited, chattel slavery. I've tracked 8 deportation flights to Mauritania under Trump. Plus at least 3 Mauritanians subjected to third-country removal. gillianbrockell.com/tracking-all...

A message for the class of 2026: We need more friction!

Google has chosen to ship a product that abets disinformation by degrading their core product, using their voice to make scams cheap and indistinguishable from reliable sources. It is how the technology works, and comes hand-in-hand with misrepresentation, libel, and defamation.

NEW BLOG FOR YOU Google's energy consumption numbers in their new climate report are mind-blowing. 2 years ago they flipped from linear to exponential growth, and their climate impact is blowing out, too. A WILD testament to the obscene bloat and waste of GenAI: ketanjoshi.co/2026/07/01/g...

As activism to curb #edtech makes news around the nation, decades of NEPC’s Commercialism in Education Research Unit research can help guide needed reforms. bit.ly/4p7rBzH

Today we beat Stephen Miller’s lawsuit mill in court! For 3 years I’ve been a defendant in a case filed by America First Legal on behalf of Gateway Pundit & a random antivaxxer we’d never heard of until she sued us. The plaintiffs claimed we “censored” them. It was a lie. They just lost.

The most fundamental duty of university administrators is to protect academic freedom. That requires fighting. Across the country, they are betraying that duty (sometimes gleefully). It falls on faculty, students, and communities, by shared governance, unionization, and politics, to demand better.

The collective journalistic freakout over someone disliking a cafeteria sandwich at Oberlin or some Yale students asking their peers to be less racist at Halloween has generated far more copy and uproar than multiple universities illegally firing faculty for nonconservative speech they disliked.

Commoditization of paper-shaped objects is not actually good for academia. Authorship criteria used to provide friction. It's likely that careful review of these desk-rejections would reveal data/citation misconduct, as we see in legal filings, but there is no incentive to perform that labor here.

If you wanna know what's happening in schools rn...

Behold the future of academia, as administrators increasingly idolize corporate executives and unlock new betrayals of shared governance. Before you say markets will sort it out, consider what that means for a school and who bears the harm. Rationalizing degraded outcomes is now Powered by AI ®.

Up now! I have questions.

Undeserved credulity. Every serious person knew this is what would happen. The purpose of this system has always been to degrade Medicare, making people more precarious and dependent on private insurance while increasing disparities. As always, the function of "AI" is to deflect accountability.

The article makes it sound like the anticipated CMS extortion directive has been issued, but it has not. This is all anticipatory obedience. Also, Childrens Colorado has not resumed care. They had all the providers issue a statement that they would not prescribe.

Teachers and professors who intend to mitigate student use of "AI" by surveilling students (e.g., recording fine-grained edit history), it is an arms race and torches mutual trust. As this thread explains with citations, it also harms student mental health.

Lots of us pointing out the unsupervised algorithmic decision & $39.95 charged for blank page that says, “This article has been withdrawn due to article violation.” But Springer Nature also killed an editorial that the editor-in-chief of Naturwissenschaften planned to write about the retractions.

"Verification [for targeting] meant a human operator reviewed each name for an average of about 20 seconds... Then they signed off". One of my German deceased relatives-in-law was (rightly) a wanted war criminal for signing off on a single civilian execution in occupied France in 1944.

"I feel like AI defeats the purpose of this class." I went into this past school year thinking that it might finally be the year where I had no choice but to begin integrating AI into our classroom as a tool and skill. Walking out of this year? Talking with students? I realized I was wrong.

Avner Gvaryahu is a former IDF officer. The article is powerful. Human systems perform atrocities, but with friction in the form of internal objections and critique, whistleblowers, trauma, and those who show courage in ways that Gvaryahu has. The purpose of automation is to remove that friction.

“This is the ghost at the center of our field: we keep being right, but wrong keeps winning. Being right is not the same as being enough.” #FAccT2026

It is interesting that everyone talks about this center as having a political philosophy, it's right there in the name, and they have never hired someone who does not align with that philosophy. And yet... www.cu.edu/regents/law/9

A phenomenon I've noticed recently is people trying to occupy some untenable middleground wrt to the use of systems sold as "AI" -- this is a position where people try to recognize the harms of this tech but also hold space for "responsible" or "ethical" use. 🧵>>

The "chain of custody of why" is a load-bearing element of science. In numerical solutions of PDE (as in fluid and structural mechanics), the modern practice of verification (solving the equations right) and validation (solving the right equations) as distinct emerged from painful epistemic flailing

Honoured as ever to be in conversations with some truly special friends in @marwasf.bsky.social and @alixdunn.com, and to round off this critical series from @themaybe.org! Check out and sign the statement here as well 👇🏽👇🏽👇🏽 www.accessnow.org/press-releas...

No—blue books, in-class exams, locked down browsers, etc. won't fix this. Nothing will replace the research, reading, thinking, organizing, writing, and revising skills that students developed through the good old fashioned research essay that has been the core of higher ed for countless decades.

One overlooked advantage of the rapidity of the digital age is that you can see someone do something stupid and evil, say "hey I have identified a really obvious risk or failure here" and then you only need to wait like a couple weeks to be proven right. Used to take years!

The purpose is to sabotage higher ed. University administrators are indifferent or incapable of defending their institutions, so that means sabotaging careers. Promotion and tenure policies and norms that expect sufficient external funding will now do the 🗡️ work to gut a generation of scholars.

Time to shine, Harper's Letter signatories? Heterodox academy? ... Bueller? hechingerreport.org/proof-points...

"generative AI to me feels like a corporate DDoS on society in a way that is really disorienting" "we're running on the fumes of knowledge produced in the past and underfunding and undermining knowledge production and serious science right now" Perfectly stated, @alixdunn.com

have reviewed a few AI citation correction packages, and they would all amplify the problem - if invented male authors are present, hit crossref, pick the first match. Authors are more exact matches than titles. "Better LLM tooling" does not fix this, you can't dig up.

by request i'm going to share a quick thread of collections where people have pulled together articles / resources / arguments about ai harms and how to resist them this is by alphabetical order in my bookmarks, it's not comprehensive, feel free to add to the list of lists

As in the 19th century, these companies are selling both the counterfeiting machine and the (purported) counterfeit detector. Despite critical-washing, in their framing, all substantive learning has been excised. Policing and surveillance will only exacerbate the root problems of transactionalism.

The second half of this article really heats up. It shows that when a journalist does the most basic part of their job - to figure out if what people say has any correspondence with truth - they bust a lot of AI ed tech scams. Gift Link:

Hard to justify products that amplify discrimination and falsehoods in citations. In Living a Feminist Life, @saranahmed.bsky.social writes: "Citation is feminist memory. Citation is how we acknowledge our debt to those who came before; those who helped us find our way…" @dukepress.bsky.social

The last thing we need is the US government having an even greater interest in the success of these grifters.

At the end of the term I asked my college creative wriing students to submit anonymous thoughts on AI. No real surprises: Mood ranges from resignation to despair, capitulation from embittered erosion of standards to total, feelings of betrayal from deep to furious. 1/

One of the things I research is the cooperation of cultural institutions like museums in pitching tech products--specifically AI--to the public. This is advertising for the capacity of AI to educate and thus to reshape education itself as a project better left to private tech companies. 1/n

An excellent interview with @heidykhlaaf.bsky.social on safety engineering (in contrast to the clown show of "AI safety"). This discussed abdication of standards and safety revisionism is paralleled in computational science & engineering with abandonment of verification and validation (V&V).

“(…) the threat from AI worthy of highlighting here is that a culture of frictionless practice, which such models (in name only) promote, is not only harmful to people's mental health and skills, but also directly undermines scholarly praxis." -- Olivia Guest (2026)

I would love to see the contract hospitals have with this vendor. What data is obtained and communicated, and how can it be used? It's a surveillance company and contracts often have carve-outs for "de-identified" data. I bet no nurses were involved in evaluating the product prior to purchase.

You can double down on surveillance and policing or you can build mutual trust. What are students focusing on here, as they self-surveil, suppress their voice, and deliberately de-professionalize their work? And who has resources to hire a lawyer for the crime of applying their learned knowledge?

This KPMG report contributed to the inevitability and workforce-development narratives that have been used to pressure academia. This is what businesses need, workforce development, etc. University administrators and business school faculty are exceptionally gullible marks.

1/Congress has made itself very clear that working to retain students under civil rights law once they attend is fine. Moreover, this is whole “racial proxy” theory is only working because it has yet to touch judicial grass. What conservatives are saying is that Black people don’t belong at Yale.

The commencement speech by @sifill.bsky.social at UC Berkeley Law provides important context behind the Fourteen Amendment and inspiring words to the graduates for the future of our democracy. youtu.be/MHuGJ66o4hE?...

"What I really would like to see unions do is push as strongly as they can on their racial and social justice commitments. [...] Once you really center [racial and social justice], you can't really go back. If you keep it marginalized, you can play with it." 💯