“In every case prior to the collapse [of an empire] there is a slow growth in inequality, oligarchy, corruption and factionalism between elites over the surplus of lootable resources.” www.thetimes.com/culture/book...

“In a heat death–type narrative of the internet’s demise, we are all infected and rendered senseless by the machine-driven decay of online content into meaninglessness. Human culture is sucked into the vortex of online collapse, and thought ceases to exist.” www.emptysetmag.com/articles/ent...

Today's decision in GEMA v. OpenAI by a German court holds that ChatGPT infringes copyright when it memorizes song lyrics. The opinion cites my paper with @afedercooper.bsky.social on memorization in generative models, and its analysis tracks ours. drive.google.com/file/d/1dUaD...

“The 21st-century algorithmic gaze dissolves both patient and doctor alike into never-ending streams of automated data.” www.theguardian.com/us-news/ng-i...

“Governments are saying 'well, we don't want to step in too soon and regulate AI'. That's exactly what they said about the internet - and look at the harm it's done to so many kids.” www.bbc.com/news/article...

First time holding my book 🥹 Thanks UCLA!

“AI maximalism across society is supercharging the ideological war on libraries, schools, government workers, and academics.” www.404media.co/ai-is-superc...

AI summaries has got to be one of the biggest concerns about professional AI use: 1) people love summaries and will assume accuracy; 2) AI summaries often look authoritative (even citing sources); 3) if you take time to check the sources, they often don’t say what AI says in the summary

“You’re a winner, baby!” 🥳😜 Honored to have won the best “Food Issues” book from the International Association of Culinary Professionals (IACP).

This from Ted is also usefully read in conjunction with the just-released statement from the MLA advocating for faculty representation on ed tech policy-making committees: www.mla.org/Resources/Ad...

“Writing is the process of discovery through language…[exploring] what we know & what we feel about what we know through language. It is the process of using language to learn about our world, to evaluate what we learn about our world, to communicate what we learn” thepalindrome.org/p/coding-on-...

New paper just out on how changing sociotechnical systems of knowledge production and access - platforms, the cloud, AI - pose profound challenges to educational practice and research doi.org/10.1080/0305...

“In…my advice to…other digital historians—I generally think that human beings are among the least promising topics of statistical analysis; I choose to visualize ships, books, and land but try to avoid visualizing the person whenever possible” dhdebates.gc.cuny.edu/read/computa...

Very proud to have worked on this statement of the Modern Language Association with @annamillsoer.bsky.social and other colleagues on the MLA’s task force on AI in Research and Teaching. It is a direct call for faculty input into Ed Tech decision-making, especially AI: www.mla.org/Resources/Ad... 🧵

There's no mystery about the likely impact of these cuts. For comparison, the Fed's post mortem on the Silicon Valley Bank failure noted that the headcount decline of 3% during the first Trump administration-a tenth of what's proposed now-strained the Fed's ability to address emerging risks.

The paper I co-authored with Hui Yun Chan has just been published in the Journal of Law and the Biosciences. In this paper, we compare Singaporean and Swiss data privacy and human research regulations, and how these may act as a barrier to cross border data sharing: academic.oup.com/jlb/article/...

On one wall: A constantly updating dashboard with students' lesson completion rates and other metrics On the other: A shop of toys (at 6x markup) they can buy with Alpha bucks, earned by hitting learning quotas In the middle: 6 yos tapping silently away at laptops www.wired.com/story/ai-tea...

“One South Carolina DoorDash driver told me he recently chauffeured a single serving of ice cream five miles; he was paid $3.50 before taxes for about 20 minutes’ work, and estimates that the customer paid about $15 for it.” www.theatlantic.com/culture/2025...

One more step toward a posthuman internet: “Winners get cloned, not repeated. Take proven content and spawn variation. Different hooks, formats, lengths. Each unique enough to avoid suppression” www.404media.co/a16z-backed-...

This is actually the thing that worries me most about AI. It's "there is no such thing as society" implemented as software solution. Needing to ask people for help, advice, or information is part of how we build and maintain social and professional bonds.

“Our findings show that once this kind of ‘brain rot’ sets in, later clean training can’t fully undo it.” www.wired.com/story/ai-mod...

Yet again, we can't afford to let LLMs become a source of epistemic grounding for society.

The deal “functions as an elaborate mechanism for socializing risk while privatizing control and returns. Intel secures capital without conditions; taxpayers inherit risk without control.” lpeproject.org/blog/intel-a...

“My poetic aspiration is that whenever you read the words “YouTube Creator” you think instead of “Audience Creation.” Social media does not create powerful Influencers but rather powerless marionettes, dancing jerkily to quantified audience tugs.” www.cambridge.org/core/element...

“A study published in The Lancet journal in June said the U.S. funding cuts could result in more than 14 million deaths, including more than 4.5 million children under age 5, by 2030.”

OpenAI’s Sora opt-out for IP is clearly OpenAI trying to do a Google ContentID pivot to turn rightsholders of key trademarked characters into price-takers not price-setters.

Last week, a Dutch court ordered Meta to give Facebook & Instagram users the right to set a chronological feed by default - one of the first EU cases applying the DSA in civil proceedings. To unpack what this means, I spoke with @bitsoffreedom.bsky.social: www.techpolicy.press/what-a-dutch...

As part of @kuleuvenuniversity.bsky.social 600th anniversary, I’m proud to contribute to 𝐀𝐈 𝐃𝐢𝐚𝐥𝐨𝐠𝐮𝐞𝐬 with my essay 𝐆𝐨𝐯𝐞𝐫𝐧𝐢𝐧𝐠 𝐀𝐈 𝐚𝐭 𝐖𝐨𝐫𝐤. I argue for anticipatory foresight and real-time reporting mechanisms so all actors can adopt a preventive oversight approach to AI.

Yay! @sahinck.bsky.social from our SCiDA-team at @hhu.de won the EuCML Best Paper Award 🏆. Sarah works on rule-setting power of Big Tech through technology (building on @lessig.bsky.social’s groundbreaking work). Pic w/ @frankpasquale.bsky.social (speaker of jury) & Alberto de Franceschi (EuCML).

“.. Notably, the poll shows U.S. adults expressing strong, bipartisan support for vaccines .. @nbcnews.com www.nbcnews.com/politics/tru...

This undersells what insurance exchange policyholders will feel if Congress doesn't act. The 18 percent increase is on the raw cost of insurance. But many will also lose enhanced subsidies that expire at the end of December. Those people will feel an average increase of *more than 75%.*

'The Australian Ad Observatory: A web of invisible influence' 🎥 admscentre.org/ad-observato... Produced by ADM+S in collaboration with project researchers and industry partners, this short-film explores key questions, challenges, outcomes & future directions in one of our major research projects.

“States experiencing recessions are spread across the country. This includes the midwest and parts of the rustbelt. California, Texas and New York, which together account for close to a third of US GDP, are holding their own, and their stability is crucial.” www.ft.com/content/e9be...

The AI authorship issue is real. We've caught papers that looked OK but had references to works by the *journal editors* that *do not exist*. Like, don't send us papers citing papers we know we didn't write, right?

“This week’s syndication of UK 10-year debt was enormously oversubscribed. The government wanted to issue £14bn. It received £140bn in orders. The IMF can stand down.” www.ft.com/content/4edb...

"We identify structural failures in scientific data curation & [thus call for] a forward-looking data development agenda that supports advanced scientific intelligence, advocating for a closed-loop feedback between autonomous...discovery & scientific data infrastructure." arxiv.org/html/2508.21...

“So, the rise of these industrial-scale computing facilities that are needed for A.I. and other Internet applications is certainly one of the factors driving up prices.” “And I think there’s a chance that A.I. is going to drive our bills even higher in the future.”

“We need to grasp our current moment as a giant non-linear outcome generator with wicked convexities. In plain English, there is no mean, there is no average, there is no return to normal. It’s one way traffic into the unknown.” adamtooze.substack.com/p/chartbook-...

“Camouflage extends beyond visual disruption into multispectral and algorithmic dimensions: adaptive textiles that shift chromatically with light or temperature, nanocoatings that suppress infrared signatures, and nets to dissolve radar profiles.” diffractionscollective.com/2025/08/30/i...

“Online discussions are dominated by a surprisingly small, extremely vocal, and non-representative minority…while only 3% of active accounts are toxic, they produce 33% of all content” www.sciencedirect.com/science/arti... @jburnmurdoch.ft.com perhaps these are the new gatekeepers?

“Social platforms’ inbuilt tendency to reward indignant and hostile content creates incentives that systematically reward the production of simplistic messages and extreme positions, while rendering moderate views less visible.” www.ft.com/content/9251...

“One of the things that distinguishes totalitarianism from other kinds of state tyranny is the way in which it invades every nook and cranny of human existence…[it] makes apolitical life impossible, thereby rendering impossible any apolitical treatment of art.” libertiesjournal.com/online-artic...

Platforms “reward attention metrics like comments, shares, and retention of viewers who watch the whole video. Influencers like myself carefully script our videos for retention, include “ragebait” to generate comments, and deliberately play into trends.” www.ft.com/content/8501...

One “chatbot, when asked to name an inspiring AI character from science fiction, chose AM, from the short story “I Have No Mouth, and I Must Scream”. AM is a malevolent AI who sets out to torture a handful of humans left on a destroyed Earth.” www.ft.com/content/7f14...

“Any attempt to encourage people to treat AI systems with respect should be understood as an attempt to make people defer to corporate interests. It might have value to corporations, but there is no value for you.” lareviewofbooks.org/article/life...

“If we each had a dedicated writer—our own court scribe—then perhaps we could follow the trajectory of the selfie further into the territory of self-based narrative” 😬 www.newyorker.com/magazine/202...

“To see the world ‘at scale’ is to regard humans as specks of data, a vantage that becomes even more dangerous when it belongs to machines, which construe individuals as mere numbers and are insensible to human pain.” harpers.org/archive/2025...

Replacing human judgement with mathematical models to make things more "fair" and "consistent" is an idea that's been around for over a century, but we keep discovering (again and again) that these models fail due to their design and input while providing cover for systemic cruelty.

My article "Protecting Consumers in a Post-Consent World," about how we can broaden antitrust and consumer protection to deal with the fact that we have abandoned notice and consent in contract law, is now published in the Stanford Law Review Online. www.stanfordlawreview.org/online/prote...