Some of us former 538 folks are reviving a project we started in 2018 tracking key endorsements in congressional primaries. Just finished checking all @emilyslist.bsky.social endorsees through 6/23 primaries: they're winning 81% of their races. In past primaries their win rate was 81-85%.

My new theory paper (early draft) A new reason why concentrated wealth destroys democracy: common ownership A standalone media owner maximizes profit. A common owner like Musk (who owns X + SpaceX) will run the media firm at a loss to increase the $ of his total portfolio tinyurl.com/mcekr7cd

I am hiring a qualitative postdoc to start this fall for the @datasociety.bsky.social project AI-Enabled Scams as a Systemic Security Challenge. Fully remote but must be US-based. Interested in youth, scams, hustle/financial/consumer culture? More here: job-boards.greenhouse.io/datasocietyr...

Last chance to register for polnet!

“The entire National Science Foundation budget is less than what we spent in the first week of the war with Iran. The National Institutes of Health,… accounts for a fraction of 1 percent of all federal spending” www.nytimes.com/2026/06/20/o...

Midterms are always referendums on incumbents, but the president’s party typically puts distance between themselves and the incumbent. That pattern has never held for Trump and in fact seems to be even more pronounced this cycle. This and more in our latest report ⬇️⬇️

Join us for a virtual workshop discussing AI tools that political scientists are using and the challenges and considerations of incorporating AI in your research, presented by the APSA Status Committee on Graduate Students in the Profession. Using AI Tools in Graduate School 📅 Friday, June 28,…

Happening in less than a week! June 22, 9 AM in Daytona Beach! View details and sign up now: forms.gle/rA1XxibZ2ydX...

No one owns ProPublica. As a nonprofit newsroom, our journalism is free from corporate interests, free from political pressure and free to read. Join us. Become a member by midnight: https://propub.li/49vQRJ9

Excited that the final version of my article "TikTok v. Garland and the First Amendment Anticanon" has been published. It argues that the case is a shameful example of the all-too-common tradition betraying 1A principles in the name of national security www.journals.uchicago.edu/doi/10.1086/...

NEW: Days after WIRED revealed that Meta had embedded an unreleased facial-recognition system in software distributed to more than 50 million users, the company appears to have removed it. www.wired.com/story/meta-r...

If you will be at Sunbelt, please come check out our workshop on the National Internet Observatory, a large, NSF-funded infrastructure, for studying people's experiences online.

Too often we emphasize the role of the individual (prepping, insurance, etc.) or the state (warning systems, seawalls etc). But we overlook the critical nature of #socialcapital - the ties that bind us - to help us survive and thrive during shocks www.youtube.com/watch?v=p4oY...

1. How common is LLM use in scientific publishing, and how does it vary across field, publisher, journal prestige, author demographics etc.? @kylesiler.bsky.social has new paper in PNAS that addresses this question on a massive scale: 7.3 million papers from Elsevier, PLOS, MDPI, and Frontiers.

Check out this Associate or Senior Editor opportunity @science.org recruiting.ultipro.com/AME1123ASEM/...

Social media is no longer social. Most of it is passive viewing of videos and pictures from people we've never met. But we're still studying social media like it's 2010. We've entered the post-social media era — and research needs to catch up. osf.io/preprints/so... preprint w/ Richard Rogers

danah boyd recently argued that social media (as we knew it) is mostly dead; what we have now is parasocial media: journals.sagepub.com/doi/10.1177/... I agree. Networked social media is disappearing, replaced by a new logic/structure of content creator, creator-centered community, & algorithm.

Very pleased to announce Who Follows Whom, a bargain, free to download for the next 2 weeks, looking at patterns of following on Twitter. www.cambridge.org/core/element... The dominant driver of patterns of online attention, with no spatial constraints on who you can connect to, is...

Of all the harm this administration has done to US science, today's proposed changes to the way federal grants are awarded and administered is the most damaging yet. The text of the federal register document is impenetrably tedious by design, so here's an excellent summary of what it says.

Very pleased to announce Who Follows Whom, a bargain, free to download for the next 2 weeks, looking at patterns of following on Twitter. www.cambridge.org/core/element... The dominant driver of patterns of online attention, with no spatial constraints on who you can connect to, is...

He's right. "Nothing short of devastating" By @holdenthorp.bsky.social @science.org www.science.org/doi/10.1126/...

“Scientific knowledge is, fortunately, a public good, and as such, its benefits transcend international boundaries. But relinquishing its prominence in creating the future is nothing short of devastating for the United States.”

New, from me: The self-proclaimed "Most Transparent Administration in History" wants to compel all federal employees to sign Non Disclosure Agreements. The purpose is to vastly expand the culture of secrecy in government, undermining accountability. 🧵 donmoynihan.substack.com/p/trump-want...

✨New paper out @nature.com ✨ For 8 weeks around the 2024 US election, we randomly assigned 2,000 people to use social media algos we built ourselves. Do engagement-based algorithms amplify intergroup, moral & emotional (IME) content—and does that distort how we see political norms? 🧵🔗 👇

One glimmer of hope in the Ebola response: Within hours of the outbreak in DRC being declared, researchers convened to prepare drug trials. A lot has changed on that front since the Ebola outbreak in West Africa a decade ago. But huge challenges remain 🧪 My story in @science.org (🧵 to come)

Three things said about AI and social science: it's a topic to study, a thing to critique, a tool to use. My new Daedalus essay argues these aren't three conversations; they're one. A 🧵 on "Field Theory: AI as Social Science Question, Object, and Tool." www.amacad.org/publication/...

What made David and Gerald's work so powerful and so threatening to the powerful is that they didn't just write rigorous history. They also leveraged it. Their research supported litigation across the US holding firms accountable in the courts and building archives for others to use for the same.

Congratulations to the IIID team who produced this latest research on how Americans spend their time online. Read it here: www.nuinfoinstitute.org/publications...

As a scholar, it's often impossible to control how your research is used in the real world. My favorite example of this is Edward Luttway's handbook on Coup D'Etat. His book was intended as a detached academic political science book, but was framed an an accessible "how to" manual.

Many academics think we are doing virtuous work by addressing critical issues. But this example reveals that we cannot control how others will use our research--for good or evil. This is why I try to make a point of communicating my research in ways that can help them put it to good use.

Web history disappears when it can’t be preserved. Today, many publishers are blocking the Wayback Machine from archiving parts of the public web, putting decades of digital history at risk. Tell publishers: don’t block the #WaybackMachine. Sign the petition ➡️ www.savethearchive.com/newsleaders/

wow — the preprint host, arxiv, is banning authors for a year if they submit papers with hallucinated citations 🤖

State media is a component of LLM training data. A new Nature study says this material is laundered into what presents as objective chatbot text. LLMs also give more pro-regime answers when prompted in the official language of low-press-freedom countries than in English.

New paper in Nature. The more a government controls its domestic media, the more it dominates AI training data, the more pro-regime outputs we get from AI. By scraping the open web, LLMs are unwittingly laundering state-coordinated narratives into seemingly objective answers.

Half of online time goes to two activities: social media (25%) and email (25%). LLM tools (ChatGPT, Gemini) now account for 2.9% of all online time. News sites get 2.5%. AI is already a bigger share of attention than news. For adults 18-29 the gap widens: LLMs 5.1%, news 1.1%.

Our new report is out: How Americans Spend Their Time Online. www.nuinfoinstitute.org/publications... Five companies (Alphabet, Meta, Microsoft, Yahoo, Amazon) capture more than half of all online time. Alphabet alone takes 35%.

🚨 Excited to share our new paper to appear in the Findings of ACL 2026! "JudgeMeNot" asks: Can LLMs learn the specific reasoning signature of an individual judge? We tackle this in the low-resource setting of Hebrew case law. 🔗 arxiv.org/abs/2604.18041 🧵👇

“It’s a massive data pipeline: phone metadata, location pings, SIM card swaps, app usage, social media behavior, sometimes even banking or facial recognition inputs. A lot is ‘scraped’ from commercial platforms, mobile networks, partner intelligence agencies, or spies on the ground..."

The we're-paying-too-much-attention-to-Twitter-and-not-enough-to-YouTube problem, quantified New pre-print from @dfreelon.bsky.social @davidlazer.bsky.social @pranavgoel.bsky.social @meredithpruden.bsky.social et al osf.io/preprints/so...

Here’s a presentation for students on data brokers & digital surveillance, with a lot of links that might be of interest. (Got a little enthusiastic about animations!) docs.google.com/presentation...

But science is a social practice, one that depends on cooperation and trust. I've studied cooperation and trust for 25 years, often with computational models, including AI. The longer I've done this work, the more I've seen the limitations of these tools. >>

Last weekend I received the Troland Research Award from the @nationalacademies.org. I’m so grateful to the communities who made this work possible. It's strange to receive this award at a time when much of the work being recognized is not eligible for federal funding. My remarks 👇 and a 🧵>>

LLMs have been widely reported as left-wing biased. The finding has shaped policy and debate — with Trump banning "Woke AI". Our new paper challenges this story. It's not that the models are biased. It's that they think the auditor is. 🧵 w/ michelleschimmel.bsky.social‬arxiv.org/pdf/2604.27633

Science shapes our lives — but how often do we talk about how it connects to society? We spoke with 2024 SciComm Excellence award winner C. Brandon Ogbunu on why creativity and communication are essential to science itself. Read: https://ow.ly/X51R50YQLhy

lol [ It wasn’t like Twitter was best characterized by well-informed, carefully reasoned democratic discourse that constituted a Habermasian ideal speech situation. ] David Nickerson via Edsall

Michael A. Santoro highlights the denominator problem: counting harms without measuring the opportunities for harm, and attempting to draw conclusions from data that cannot support them. He argues that this measurement gap is undermining AI governance across numerous domains.

Political campaigns have no idea what's about to hit them www.nytimes.com/2026/04/28/o...

Journal of Online Trust and Safety published its Spring 2026 Issue and here is the list of authors and editors for the issue. Read the issue at tsjournal.org/index.php/jo... and follow the authors here!