Now is probably a good time to remind folks of the carefully-worded statement on AI that a the IMU and many mathematicians, including some active AI users, signed onto. They lay out principles, and concerns about second-order effects, that probably apply across many fields. leidendeclaration.ai

Leiden Declaration on Artificial Intelligence and Mathematics

This declaration calls for action to address the challenges posed by the use of artificial intelligence within mathematics research.

leidendeclaration.ai

Quanta Magazine@quantamagazine.org · 2d ago

On Saturday, OpenAI announced solutions to three Erdős problems. These solutions, along with May’s high-profile counterexample to the “unit distance” problem, are quickly changing how mathematicians approach their work. @kkakaes.bsky.social reports: www.quantamagazine.org/why-the-lege...

If only the story ended there

e.w. niedermeyer@niedermeyer.online · 5d ago

well, a couple of Moda Center "Guest Experience" stooges just tried to tell my friend that she couldn't wear a shirt that says "F*CK TRANSPHOBIA" in the building because "there had been complaints" and "you can't wear shirts with profanity" we laughed and went to our seats

The AI project “talkie produces bigoted responses then frames them as a representation of 1930; an outcome of the project’s methodology. The hard work of expanding the archive is being subverted.” activehistory.ca/blog/2026/07...

The Archive is Not a Toy: The Hidden Problems of a ‘Vintage’ AI

talkie’s poor representation of history is a side effect of a larger problem. The model was conceptually built as an experimental tool to see if language models can accurately predict the future, a…

activehistory.ca

A new AI data center can be up and running in a year. Building new energy infrastructure to power it could take a decade.

Data Centers Are Easy to Build. Powering Them Is Complicated, Slow, and Expensive

A new AI data center can be up and running in a year. Building new energy infrastructure to power it could take a decade.

404media.co

i wouldn’t say i Know Ball per se but imo the court at powhatan community center has one of the sickest backdrops in all of richmond

finally got a link to share with people who try and say "oh you DON'T LIKE AI??? what about ALPHAFOLD!!!!" as if it has anything to do with any of this AI capex

Engadget@engadget.com · last wk.

DeepMind's award-winning AlphaFold team is no more.

Turned this rant into a blog post. davekarpf.beehiiv.com/p/the-data-c...

The data center backlash Is working. That's why the New York Times is so concerned.

I have a

davekarpf.beehiiv.com

Dave Karpf@davekarpf.bsky.social · last wk.

I'm reading the "environmentalists are wrong about data centers" piece from the NYT and ohmygoditssofuckingbad. Apparently, so long as you're punching left, they'll just let you say any goddamn thing you please... www.nytimes.com/2026/07/25/o...

I'm used to university-provided news stories and press releases being mild, feel-good stories, but Christopher Whyte isn't doing that. news.vcu.edu/article/an-a...

Facing federal science budget cuts, biologist Daniel Blumstein resorted to uploading marmot videos to OnlyFans to raise funds for his marmot research. However, when he uploaded his driver’s license for verification, OnlyFans informed him that his appearance didn’t match the marmots in the videos.

‘OnlyMarms’: Marmots Are on OnlyFans to Raise Money for Research (Gift Article)

A team of scientists uploaded videos of the squirrel-like rodents in the wild on the adult-content site OnlyFans to raise money amid government budget cuts.

nytimes.com

Generally I'm skeptical of this methodology. On the other hand, I like that the report describes how risks are borne by individuals, while the responsibility lies with AI creators and governors. It's a great comeback to, "Why do the kids hate AI?" cdn.prod.website-files.com/669550d38372...

I wonder what these economists think of protection rackets and lemon laws. www.nytimes.com/2026/07/16/t...

I’m finally going to listen to my doctor and eat more vegetables and get more fresh air

Now that I've seen a couple people sharing the NY Times EV piece, I'll promote the EV whisperer who I follow. bsky.app/profile/nied...

e.w. niedermeyer@niedermeyer.online · 3mo ago

I bet "the Detroit CEOs should have simply pivoted to Chinese-style EVs" goes crazy hard if you've never watched Wall Street analysts roll their eyes at the latest automaker effort to build new businesses because it's easier on them if legacy businesses don't even pretend they will ever change

Wait, is this the same David Streever who used to live in Richmond? @vpm-ian.bsky.social www.npr.org/2026/07/01/n...

He sent a harsh email to ICE's top official. 5 months later, federal agents tracked him down

Federal agents went to a Rochester man's home to warn him that an email he sent five months ago could be an illegal threat. He was on vacation but another agent found his hotel hundreds of miles away.

npr.org

In counterpoint to the "teens these days aren't reading for pleasure anymore" narrative, it looks like the biggest drop is among elders. www.sciencedirect.com/science/arti...

For instance: incumbent Los Angeles Mayor Karen Bass, who’s running for re-election, bowed to pressure from two cemeteries & halted the long-awaited Forest Lawn Drive safety project. Despite years of planning and outreach for a redesign to make the corridor safer. bsky.app/profile/stre...

Streets For All@streetsforall.org · 3mo ago

Mayor Bass has intervened to keep Forest Lawn dangerous www.streetsforall.org/blog/mayor-b...

The study was selective about the "human written" comparison texts. AI was easy to detect compared to published stories written by folks like "Joyce Carol Oates, Stephen King, Louis L'Amour, Charlotte Perkins, and Harlan Ellison." (Image from the Github page for the study)

404 Media@404media.co · 4w ago

NEW: Fiction written by AI is easy to detect because it struggles with complex story structure, according to a new study. Researchers found that AI fiction has tells that go beyond the overuse of em-dashes and other obvious AI tropes and have more to do with the formulaic nature of the text itself.

Because that Brown professor is a game theory expert, I hope there's someone -- maybe even him -- who will model the AI cheating situation as a variation on the prisoner's dilemma.

Jane Rosenzweig@janerosenzweig.bsky.social · 4w ago

this piece by @jeffsharlet.bsky.social about what his students have said about AI, why they're using it, and what they think about it. "Nobody used the term “arms race,” but that’s what they described: enormous pressure to use AI or fall behind those who are." /2

This week, I was reading a compendium for professors of AI-limiting assignment and testing methods. One author's job included, ".. encouraging all stakeholders to use AI." After writing a couple books on how to use AI in classes, even he was following faculty requests to write this opposite book.

Ed Zitron@edzitron.com · 4w ago

Cultist shit. A desperate last gasp operation to try and keep people believing, to bully them into acquiescing, to use social pressure to force a technology that will either continue to increase in cost and eventually will simply run out of cash to continue, truly corrosive to company culture.

I agree with this post about the paired reading. AND. I’ve been thinking about this story since I read it last week. The banner issue gets lost in the AI as cheating narrative. Serrano switched to take-home exams in the wake of a fatal campus shooting he attributed in part to excessive stress. 1/n

Jane Rosenzweig@janerosenzweig.bsky.social · 4w ago

Here are two stories worth reading together. In the first one, a Brown University professor found that most of his students used AI on a take-home exam. This is a story about AI use that fits the framework of academic honesty we had before AI: students just...cheated. But worth reading along with/1

[stares at Gavin Newsom] ufw.org/capital-main...

Industry's capture of AI literacy continues apace...

Shannon Mattern@shannonmattern.bsky.social · last mo.

Missed this last month: "Google.org last week announced a $3 million grant to Brooklyn Public Library, New York Public Library, and Queens Public Library to help the systems expand artificial intelligence (AI) literacy programming for staff and the public." Oh, boy.

Local news from the RVA area gets national attention. Relative to NOVA, I probably wouldn't call call Henrico a major hub for data centers in Virginia.

Matthew Gault@mjgault.bsky.social · last mo.

a county with 37 data centers just told employees they were going to have to tighten their belts as electricity prices shoot up 25%

What *actually* works — right here, right now — to address climate change? Here you go. This is the result of thousands of hours of painstaking research and analysis from the Project Drawdown team.

New paper! w/ @cocoscilab.bsky.social🧵Can large language models reason flexibly, or have they learned what reasoning looks like? We introduce a new paradigm to test this question—the riddle riddle—and find that humans and LLMs show opposite patterns of performance. 📜

I would love to see this in Richmond city government. About the only more winning issue for a candidate would be fixing the city utilities. www.rva.gov/sites/defaul...

gamera obscura@ohrobin.bsky.social · last mo.

This Danville, VA city council candidate made a very dramatic video about sidewalks and public safety, and I love it and him. Worrying about a possible milkshake duck situation, I looked at his FB page, but he seems ok. Juneteenth celebrations, community outreach, Methodists. A solid old gent.

The ghost of David Golumbia haunts me for thinking he was being alarmist when he wrote about it a decade ago.

Gil Durán@gilduran.com · 2mo ago

Even as my book heads toward publication, I remain amazed that someone else didn't write it first. More than a few marquee journalists at major publications were in much better positions to tell the world about Silicon Valley's dangerous radicalization. Even now, at this late stage, most are silent.

For all of you legal academic nerds. Here's the tool. Pulls case citation information from any federal court document. github.com/esqueer/zote...

GitHub - esqueer/zotero-courtlistener-retriever: Automatically pulls the case cite information from Courtlistener for a federal court document in zotero

Automatically pulls the case cite information from Courtlistener for a federal court document in zotero - esqueer/zotero-courtlistener-retriever

github.com

Alejandra Caraballo@esqueer.net · 2mo ago

Finally created a zotero plugin to pull case citation information from Courtlistener for an imported court document PDF with no information attached. This is such a timesaver.

About a year ago someone ridiculed AI refusers for being quaint and naive. I jumped in with the idea that we were in a 1920 moment, before the interstate highway system and the other damage created by molding U.S. cities around cars. Bending the path now will be easier than reshaping later.

Sarah Dougherty@sarahdougherty.bsky.social · 2mo ago

Ahh, so trying to say we should change norms to let AI dominate like we let cars dominate now. Dont think that is a convincing argument for letting those norms change. I would like go back to earlier outrage at traffic violence, as discussed in book by @thewaroncars.bsky.social

Don't overlook the importance of the "retrieval" part of retrieval-augmented generation (RAG). Cornell Tech researchers find it is trivially easy to use Reddit to manipulate "deep research" AI www.404media.co/it-is-trivia...

It Is Trivially Easy to Use Reddit to Manipulate AI Search, Research Suggests

"We show that a tiny snippet—just 13 words—of retrieved text on a UGC website like Reddit, Wikipedia, Quora, or Facebook can change AI agents to output spam / scam content pretty consistently."

404media.co