Yet again, machine learning — even gussied up via the transformer architecture — encodes and reinforces societal biases. This study reveals that LLM-based peer review relies heavily on author institution in its decisions. arxiv.org/abs/2509.15122

As we boycott vile companies, a reminder that many US public libraries offer: 1) Ebooks & audiobooks via the Libby app. Pro tip: use the “notify me” tag to request your library add books by trans, disabled, & BIPOC authors! 2) TV & film streaming via Kanopy 3) Language learning via Mango 💙📚 And:

Glad Science collected this data (though the results are entirely unsurprising). GenAI cannot accurately summarize scientific papers, sacrificing accuracy for simplicity. And shame on publishers who are pushing genAI summaries on readers. Great way to accelerate an epistemic apocalypse.

Reading papers is a basic skill (and dare I say duty?) of scientists, and I include medical doctors in that group. Keeping abreast of the literature is a foundational part of our professions. There aren’t good shortcuts. In any case, reading papers regularly is fun. arstechnica.com/ai/2025/09/s...

"However, it tended to sacrifice accuracy for simplicity. The ChatGPT Plus summaries required rigorous fact-checking by SciPak writers. Also, extensive editing for hyperbole was needed. For example, ChatGPT Plus had a fondness for using the word 'groundbreaking.'”

👇👇👇

I am recruiting (Canadian🇨🇦) graduate students at York University for Fall 2026 in Social and Personality Psychology! If you are interested in misinformation, political polarization, and/or computational social science, I'd love to hear from you. Feel free to reach out with any questions.

If you're an author and haven't yet signed up for the Anthropic settlement, do yourself a favor (and strike a blow against AI) by checking to see if any of your books were involved. www.theatlantic.com/technology/a...

This is the benefit of living according to an internal moral code. You don't (usually) get rich. You don't (usually) get recognised. But you do avoid becoming a hollow shell, a simpering coward who humiliates themselves in pursuit of power. You can look yourself in the mirror. It's worth it.

This is such an important thread, but this post really made me sad, because the broader scholarly infrastructure has (been) eroded to the point that knowledge exchange simply can't exist in the way it used to and there's little that journals can do about it. Because that infrastructure was people.

the enshittification of america

NJ students in grades 6-12 will now learn about “the history of organized labor, notable strikes throughout history, unionization drives, and the collective bargaining process and existing legal protections in the workplace.” This is genuinely so important, and so good to see. Jersey strong, baby!!

We took the paywall down on this one. defector.com/how-the-fuck...

Excited to present this great essay by Ashanty Rosario, a talented young journalist who attends Newtown High School in Queens, about how she's experienced the arrival of AI in the classroom. "These programs have destroyed much of what tied us together as students."

If your state hasn't done this yet, call your governor's office as well as top public health official in your state, and request that they do this. If enough people start calling, they will hear us loud and clear!

there is no question in my mind that rfk jr is the most dangerous person in this administration and that his eugenicist ideology threatens the lives of millions of people www.advocate.com/politics/dem...

The average American simply has no idea how quickly the infrastructure that keeps us safe and healthy is collapsing. These systems run in the background thanks to government employees who quietly do critical work. Those employees are being purged and systems are being destroyed.

The extent to which these companies assume by default that they are unimpeachable and can do this stuff without any fear of meaningful consequences is reaching truly impressive heights. Here is a Meta product lead just casually creating a sex bot out of the likeness of the most famous singer alive

"This is both a clear-cut moral abomination and a logical culmination of modern surveillance capitalism. It is the direct result of tech companies producing products that seek to extract attention and value from vulnerable users, and then harming them grievously. It should be treated as such"

Important 🔥🔥 must-read from @beenwrekt.bsky.social ⬇️ www.argmin.net/p/the-banal-...

A California teen sought advice from OpenAI's GPT-4o on how to end his life. The chatbot gave him explicit instructions and encouragement. His parents are suing the company and its CEO, Sam Altman, alleging “it was the predictable result of deliberate design choices."

Been saying it since November and earlier than that - pick one or two things and focus.

Give @kashhill.bsky.social all the garlands for her dogged, brave reporting yet again. It's shameful, outrageous and enraging that we allow tech companies to deploy AI products that cause harm to people--teenagers, Black fathers, mentally ill persons--every day. Enough is enough.

Adam Raine, 16, died from suicide in April after months on ChatGPT discussing plans to end his life. His parents have filed the first known case against OpenAI for wrongful death. Overwhelming at times to work on this story, but here it is. My latest on AI chatbots: www.nytimes.com/2025/08/26/t...

I can't quite believe it – I got a new NSF grant! 😲🤯 I'm incredibly grateful to the program officers & reviewers for their dedication and efforts to keep science going 🌟👏🏼 So, I'll be hiring! Looking for a postdoc to study competition in memory-guided attention. See flyer for details! 🎉

We got burned by an AI freelancer and a few missteps of our own. Here's the full disclosure on that. All I can add from my vantage here is that my work has always been subject to rigorous fact checking. It's sometimes a days long, maddening process, but 100% of the time I am grateful in the end.

I know my position within the academy is different. I’m keenly aware. I know what the attacks on HE look like and that we are in no way suited to defend them. I’ve tried to hold my mule. But this? This makes me want to pop out so badly. You don’t have an ethical responsibility to use A.I.

New from me. Main points: 1. Emerging tech & features are enabling avoidance at a mass scale. 2. While there is still much to understand, we already know a lot about how kids develop. We can already do some of the math. I talk about the delusion of (a) physical perfection, and (b) connection. 🧵

Princeton friends and followers, mark your calendars! citp.princeton.edu/events/2025/...

This is the truth. We have the best universities in the world right now and essentially export education. This is so shortsighted. www.forbes.com/sites/stuart...

We're so excited to host this editor community conversation THIS WEEK! There's still time to register here: northwestern.zoom.us/webinar/regi...

Looking for a critical book on AI or Big Tech? I've started a little curated list - take a look! 📚💡/ #booksky

🤖 Example of how “informed consent” can go wrong — and what it means to be “informed”: users inadvertently agreed to share details that could be used against them. The friction for sharing private info should be greater than a checkbox or else shouldn’t be an option. venturebeat.com/ai/openai-re...

Methods in the Philosophy of Science: A User's Guide is now a real, live book. My chapter offers an appraisal of philosophy of science in its engagement with applied mathematics. With immense gratitude for the vision and execution of @phieveigl.bsky.social and @adrian-currie.bsky.social sky.social.

Trump's AI Action Plan (surprise!) consolidates power and capital to Big Tech. On the next Mystery AI Hype Theater 3000 live stream, @alexhanna.bsky.social & I will be joined by AI Now Institute’s @kate-brennan.bsky.social & @smw.bsky.social to piece through the bullshit. Mon, Aug 4, noon PDT

here's a useful critical annotation of zuck's "superintelligence" memo sonjadrimmer.com/blog-1/2025/...

We may have the chance to hire an outstanding researcher 3+ years post PhD to join Tarleton Gillespie, Mary Gray and me in Cambridge MA bringing critical sociotechnical perspectives to bear on new technologies. jobs.careers.microsoft.com/global/en/jo...

I am starting "More Everything Forever" by @adambecker.bsky.social this evening, and what a solid start with these epigraphs. I'll add my thoughts in this thread as I read. 🧵

I've learned so much from Melanie Mitchell's thoughtful columns. During a pivotal period when AI discourse generated more heat than light, her essays provided wise, evidence-based analysis covering remarkable breadth. This compelling column on role-playing's impact on LLM performance is an exemplar.

Trump's new executive order about AI states that companies' models must not incorporate the idea that trans people exist as a concept. Seems bad.

Do not let any LLM serve as a form of epistemic grounding for society.

Hey is it bad if google AI search results stop people from leaving google dot com 99% of the time? All our news and information creators can survive just fine with one percent of their previous search traffic right? I’m sure it’s fine. www.404media.co/googles-ai-i...

Younger folks will never understand just how unbelievable Google search was from like 2004-2010. You could unearth so many amazingly detailed and thoughtful expert resources on any topic. The first three pages of results could give you weeks worth of info to comb through and dive deeper into. A joy.

Really cool how we don’t need or want this and no one asked for it. This will be the straw that turns me completely back to asking them to turn in work on printed paper which will annoy them (but has always been my wish)

IMO it would be a very good time for AI researchers to take a hard look at how medical research handles conflict of interest disclosures. I'm so tired of seeing researchers who serve on the boards of AI companies or even sell AI tools writing "No Conflicting Interests" in their papers.

this reminds me of the time i got an email about my “vocal fry” in a guest lecture

A new piece - well, almost a whole book! - by our team, reflecting more than a decade of work.

How can developmental science journals promote equity, transparency, and shared responsibility? Join us August 13, 2pm ET, for “Editors for an Inclusive Future: A Community Conversation” tinyurl.com/DevEditors @lehersingh.bsky.social ‪@mattgoldrick.bsky.social

important context for headlines like this: Altman also runs a company (Worldcoin) that he claims will solve this problem he also created