📢 The workshop on Insights from negative results will be back at EMNLP'26! Your most-insightful failures can be submitted in 4 pages by June 25. It's also possible to commit short papers reviewed through ARR. insights-workshop.github.io/2026/cfp

Enormously looking forward to this event on April 21 at 6:30pm ET (online!) with the amazing @emilymbender.bsky.social and @alexhanna.bsky.social talking about their new book, The AI Con. Register here: virginia.zoom.us/webinar/regi...

New: 150 unionized ProPublica workers are on strike TODAY over AI, layoff protections, wages, and more. They're asking readers to not visit ProPublica or engage with content on other platforms. It's the first work stoppage of its kind at the newsroom. www.theverge.com/news/908401/...

something strange and horrible and somehow fitting that we should have a very real threat of madman civilizational destruction at the very moment when we also have humans on the dark side of the moon taking pictures that show how small, precious, and beautiful our world and existence is

Chinese government decides to regulate AI from the right to personal likeness, child safety and addiction angles. www.reuters.com/world/china/...

Tech makers assure parents that child safety AI keeps kids safe. The open secret? No one can say how well it works - a supply chain of ignorance that puts lives at risk. New w/ @camillefrancois.bsky.social @mmitchell.bsky.social @yjernite.bsky.social Vinay Rao www.techpolicy.press/why-ai-model...

today, @roost.tools is joining @yjernite.bsky.social, @mmitchell.bsky.social, and @natematias.bsky.social to call for more transparency in the field of online child safety, and for model cards for key child safety models specifically. www.techpolicy.press/why-ai-model...

🚨Paper on AI & Copyright Courts have credited AI companies' claims that alignment prevents reproducing copyrighted data. What if finetuning on a simple writing task breaks it. Worse: tuning on just one author (e.g., Murakami) unlocks verbatim recall of 30+ other authors' books (up to 90%) (1/n)🧵

this piece articulates some really important insights, but I have issues with the core frame presented in this paragraph

A viral dog rescue from 2016 is now immortalized with a statue in Kazakhstan. When a dog fell into the Sayran reservoir, bystanders formed a human chain to pull him to safety. The statue is a reminder of the value of unity, solidarity and collective action. 14/10 for all

Seriously what is going on with octopuses Get a job and stop bumming rides from other organisms! www.popsci.com/environment/...

🤖 Interesting study about relying on increased automation. Beyond cognitive offloading, *cognitive surrender* involves thinking system output is basically your own thought…analytical skills degrading all the while. papers.ssrn.com/sol3/papers....

🪶 Pileated woodpecker on my walk today. Chillin’ on a Doug Fir, next to a Madrone.

Even if you're pro-AI, Anthropic themselves have said that using AI when learning means you don't develop skills You need those skills later to debug the LLM's output or prompt it more effectively to fix something. These students are gonna struggle as soon as they have to do something harder.

If you worked as a woman engineer from 1975 to today—or know a woman who did—submit an essay for consideration to the Women in Engineering anthology. Call closes June 30, 2026. Details & submission form: forms.gle/Bh5AQVAXn9E8... #WomenInSTEM #EngineeringCareers #WomensHistory

I keep having to tell students that the true "product" on a homework, essay, or test is not the completed assignment, it's the brain that has gone through the thinking required to complete the assignment.

If you are interesting in having a nuanced conversation of a specific type of automation, I recommend naming the specific type of automation. >>

👨‍🎨 Spoiler alert: Banksy's identity has been uncovered. It is not the Massive Attack guy. I'm not too happy about his identity being disclosed, I really liked the mystery. www.reuters.com/investigates...

Not to toot our own horns but I feel like we packed so much in this one hour panel with @karenhao.bsky.social and John Palfrey. You can find the recording at www.youtube.com/live/I1tJnM8...

behold... "llms can produce human like text, thus llms have human-level intelligence" (disclaimer: don’t expect anything but snarky comments from me today) 3/

This is very interesting and I need to set aside time to read all the definitions and commentary. However I quite like the closing section — “When *not* to define AI” — as it points out that “conduct, outcomes, risks, or institutional roles” are often more relevant for legislation.

🤖 In policy? Thinking about the definition of "AI"? Led by @aspendigital.bsky.social, a set of us at intersection of AI/ethics/law/policy put together this resource on the lineage of policy "AI" definitions, what they're getting right, what might be improved. www.aspendigital.org/report/defin...

If someone tells you that there will always be a human in control of the AI, ask them if it will always be a good person.

“you require us — the people harmed — to put brakes on the harmful things YOU are doing” 🔥 post www.moryan.com/an-open-lett...

The Trump admin dismissed AI risk assessment, evaluation and bias testing as burdensome overreach. Now their own DoD is requesting exactly that to ensure systems actually work--in war. Basic, foundational safety and effectiveness testing is essential for *ALL* AI use, as are our rights to privacy.

really appreciated. for the consideration/discussion of how people are perceived by their language, esp DHH folks and signers, suggest the work of @deafkilljoy.bsky.social and Dr. Jon Henner on modality chauvinism criticalstudycommunicationdisability.org/index.php/jc...

I'm trying to understand what's going on at Grammarly and what's changed in recent years. For those not in the know, Grammarly was an NLP darling for a long time, very present at our conferences, and it was a Ukrainian company based in Kyiv that now has new CEOs and is based in SF. What's the story?

Interested in working at DAIR? We're hiring a communications lead, who we needed yesterday. Is that you? Please apply. Is that people you know? Please share the job ad with them so that they apply.

"AI" is not a stochastic parrot.🦜 I wrote this piece a couple weeks ago, but it was hard for me to finish up given AI's role in society and war over the past few weeks. I should share it at some point though. Not perfect, but here it is. medium.com/@margarmitch...

Amazon has fully embraced AI pair programming: the bot codes the outage and a senior engineer explains it, during postmortem. futurism.com/artificial-i...

Good words of warning here. AI models are generally designed to read and incorporate (and, in a sense, believe) virtually everything they find online, with very few guardrails. The information can get divorced from its original source, context, & rebuttals or retractions. That's a really big problem

Hey @officialgrammarly.bsky.social we're gonna need to know the full list of identities you have stolen here as well as clear info on how to opt-out techcrunch.com/2026/03/07/g...

I guess we are going to have to find a way to copyright ourselves so that we have the right to demand automatic takedowns of bots that steal our identities…

Wait, what?! Grammarly made an AI persona of me and all these other journalists?!!!!! Thank you @caseynewton.bsky.social for shaming them into at least offering a paltry optout— although obviously that is not enough. www.platformer.news/grammarly-ex...

I'm no artist, but hopefully this gets the point across. A comic in three panels.

No discussion of tech media can get past this basic traffic fact: in the AI world, Google and social no longer refer traffic, which means that the vast majority of readers just never find you in the first place. Analysis: growtika.com/blog/tech-m...

U.S. Military Announces Plan To Consolidate All Wars Into Final, Epic Battle https://theonion.com/u-s-military-announces-plan-to-consolidate-all-wars-in-1824018300/

Demand for Grok within some fed agencies not that high, unless you're trying to act like a bad actor for defensive testing: www.wsj.com/politics/nat...

NEW! Mystery AI Hype Theater 3000: How the War Department Learned to Stop Worrying and Love AI @naomiaklein.bsky.social joins @emilymbender.bsky.social & @alexhanna.bsky.social to discuss how AI boosters & the US military are engaged in a lethal love affair. www.buzzsprout.com/2126417/epis...

I talk more about a positive vision for recommendation, and why I think it is a necessary tool for the Internet to truly live up to its potential as an engine for equity, most explicitly here: md.ekstrandom.net/talks/2025/v...

I dunno how many times someone has told me, “the stochastic parrots paper was wrong,” and i’ve had to stutter, regain my composure, and put in real work to salvage the conversation. Like lol, wrong about what? That coherence is in the eye of the beholder? We see things as we are, not as they are.

AWESOME piece deconstructing some (crummy) ethical AI arguments. tante.cc/2026/02/20/a...

This is really important/relevant from the perspective of how power is centralized in AI. @ainowinstitute.bsky.social in particular has had some great scholarship on the role of Cloud infrastructure in AI: ainowinstitute.org/publications...

Come work with me! We are recruiting for a Head of Comms and Marketing to help us communicate the work of @britishacademy.bsky.social and the power of the humanities and social sciences in making sense of our changing world app.loxo.co/job/MzM2MzQt...

Annnnd now I can never use a coffee maker in a hotel room again.

A new paper shows that less than 2 months of exposure to Twitter’s algorithmic feed significantly shifts people’s political views to the right. Moving from chronological feed to the algorithmic feed also increases engagement. This is one of the most concerning papers I’ve read in awhile.

The framing of this episode is extremely silly ("Is she a prophet... or is she just wrong?"): That is the oddest dichotomy I've encountered in a while. I'm a scholar, carefully observing what's going on now through the lens of my expertise, and sharing what I see. >>

This piece is coming out at a time when AI dissatisfaction is rising, not as a partisan issue, but as a social one. LLMs specifically are pissing everyone off because of how they're being rolled out. I'm saying this as someone who is in apolitical tech spaces like gaming and home networking.

"if human beings are nothing more than machines — stochastic parrots — then it's easy to justify replacing us with other machines that can do certain things better and faster." I'm not religious, but do appreciate thoughtful insights. Nice article in this vein: www.ncronline.org/opinion/lent...