For Vogue, I talked to Ramell-Correen Frederick, aka Cheeks, who embroidered Knicks gear in the streets of Brooklyn after Game 5. He uses a 104-year-old hand-crank, chainstitch embroidery machine that he named Jessica.

headed back to the isaac newton institute and im stoked to live off 5£ meal deals for the foreseeable future

Ugh www.nytimes.com/2026/06/11/u...

Best thing about Mamdani is that he's one of the only Democrats who understands you can just do things

I regret to inform you that Claude Fable is very good at writing technical tutorials. Here are 10 notebooks in both Python and R using @bsky.app @atproto.com to run common social data science tasks: search, SNA, replies, time series, NLP, images, LLMs, feeds, streams, and bots.

i don't think we should waste real expert reviewer's time reviewing mediocre AI-written papers and having real expert reviewer provide valuable feedback that, presumably, will just get sent to LLMs to complete. at that point the reviewers might as well be authoring along with the authors

has anyone written about the inherent optimism in scaling up pilots to larger-scale implementations? DM's policy is not risk-averse for beneficial social interventions that aren't expected to cause harm (but may be noisy zeroes), but risk-seeking/optimistic

I am SO EXCITED that the Journal of Stratification Economics is coming next year! This has been a labor of love to bring this to pass. Please consider submitting to this exciting new journal specializing in the best stratification economics research. @jsecon.bsky.social @ucpress.bsky.social

No special treatment for Space X in the S&P 500! Just announced. Wow. That profitability requirement means it might never get in.

I'm giving some talks in June! Please hit me up if we overlap :) June 12 - SIGMETRICS causal workshop sites.google.com/view/causal-... June 19 - INI Causality & Machine Learning workshop almost last :') but not least :') Around 6/15-25 June 26 - FAccT Tutorial programs.sigchi.org/facct/2026/p...

urbanomnibus.net/2026/06/tide... new piece I commissioned for Urban Omnibus just dropped, it's an interview with Marie Lorenz reflecting on 21 years of running an art project taking people around the NYC harbor in a rowboat and what you learn about the city by doing that

Yikes

"Starbucks Sold Its AI on a 99% Accuracy Number. Nine Months Later, 11,000 Stores Are Counting Milk by Hand" <<- aka, it worked in the lab so roll it out everywhere and let the big bang magic happen! www.thestateofbrand.com/news/starbuc...

this one is an excellent resource w/ very specific suggestions on submitting. Unique comments are worth much more than ones copied from somewhere else. They can count 100 identical comments as just 1 comment. elizabethginexi.substack.com/p/what-we-ne...

• Submit a comment opposing the destruction of 🇺🇸 science; • Pressure every org, professional body, association, uni, etc. you are affiliated with to submit the strongest possible comment opposing this calamity (highest ROI move for most of us) • Make sure your electeds see those comments 🧪

4/ 🧵 I have a history of running effective commenting campaigns. During Trump 1, I produced commenting guides for specialist, expert audiences like Hospice & Palliative Medicine Clinicians, and Pediatricians. And I've written about Notice & Comment as a tool ⤵️ www.yalejreg.com/nc/reclaimin...

I kept thinking about Boaz's blog post, and ended up responding with a fairly long comment. My response touched on a few things I wanted to say anyway, and so I decided to make it a blog post of my own: hdbennett.wordpress.com/2026/05/31/m.... (It's my second blog post ever, and first in >10 years.)

Americans say they want less inequality. Why don’t they support policies to reduce it? New paper: across 31 nationally representative survey waves (N=384,248) and a preregistered experiment (N=1,009), Americans favor PREdistributive over REdistributive policies 🧵 osf.io/preprints/so...

In the last 48h: - Jr researcher asked me wheter to use AI in making talks - Saw two talks, with AI {slop, enhanced} slides Collected my thoughts and wrote a post. Tl;dr: don't steal your own thinking, don't remove *you* from your talks. Also, give a &#@% about your talks.

yes, and — even worse: today AI systems today enforce *existing austerity* of our public benefits systems and social safety net, helping facilitate determinations of who is worthy and not of care, food, unemployment, etc. only made worse by OBBBA's Medicaid cuts + work requirements. www.btah.org

procrastinating on my projects by building this would definitely help me with my projects

need an ai agent to pm me on my multiple projects (or keep me from adding too many projects to my stack)

I wrote a blog post on my Ph. D. application experience to both reflect on the process and serve as a resource: officialadithya.github.io/blog/2026/my...! I haven’t provided too much explicit advice, but throughout I have detailed various things that I think helped my application.

one simple rule for detangling academic writing: Who is doing what to whom and why, and who should do what instead.

Excited to share our paper! Due Process on Hold: A Queueing Framework for Improving Access in SNAP arxiv.org/abs/2605.15165 Millions of Americans interface with the social safety net via call centers that are too congested. In Holmes v. Knodell, bad operations = procedural due process violation.

🚨SNAP tracker update: Data through February is now available for all states. In the 7 months since H.R. 1 enacted the deepest SNAP cuts in history last summer, we estimate that SNAP participation fell by more than 3.5 million people nationwide. cbpp.org/research/foo...

arXiv📈🤖 Designing Persuasive Experiments By Adusumilli, Vemulapati

An auditor for the Ontario, Canada government found that AI agents tasked with turning doctor/patient conversations into structured notes routinely hallucinated false treatments, replaced drug names with entirely different drugs, and missed crucial information

every restructuring/revision starts on paper these days

Look, Eminem already covered this: “’Cause, man, these goddamn food stamps don’t buy diapers / And there’s no movie, there’s no Mekhi Phifer, this is my life” Diapers are EXPENSIVE

Absolutely deranged website. Private developers have proposed putting a (controversial!) gondola in LA State Historic Park. This requires a General Plan Amendment, on which comments must be solicited. The "virtual open house" is BONKERS lastatehistoricparkgpa.com/voh/

new piece i co-wrote with the propel team: we spent a day with 40 SNAP beneficiaries watching how folks actually use AI chat. the biggest finding had nothing to do with prompts, accuracy, or hallucinations. it was about modality. www.propel.app/insights/voi...

We are implementing a similar policy at @arxiv.bsky.social. If there is incontrovertible evidence of LLM slop in a paper, this means the authors did not take the time to read the LLM output and we can't trust anything else in the paper. Penalty is 1 year ban from arXiv followed by...

"What fixing the FAFSA shows, even if it took much too long, is that administrative burdens are not inevitable. Barriers can be removed or lowered, when consistent advocacy and political will lead to needed legislative changes, and smart hard working public servants are given...support."

And that played a big role in the redesign - moving FAFSA from a form where you answered 100 questions to one where you answer 38 means the process of applying is much simpler - and the "nudge" project told us that this would *matter*.

The nudge programs were good, but limited. But the point of a successful nudge project shouldn't be "we need to do more nudges". It's "Wow, if this nudge is having a big effect, it means something is terribly wrong with the aggregate process!".

I filed for unemployment for the first time when I was fired last year, and never got past MD's broken identity-verification system. I was the most famous fired person in America at the time, and a 20-year career bureaucrat, but I found another job before I ever successfully filed a UI claim.

Learned a lot from this! "UI, not UBI" Something concerning about a lot of "existential AI risk discourse" is its futuristic orientation about hypothetical harms, not about deeply engaging with existing systems to improve social protection.

Part 2 of my series on AI and job loss. What do you do if you lose your job and never find another one? The US needs an ambitious, creative long-term unemployment program to foster second careers with dignity. www.bloomberg.com/opinion/arti...

My exhibition Taking Over: Subvertising in NYC, a decades long photographic record of interventions by artists & activists, opens at @deutscheshausnyu.bsky.social on June 17. More info & RSVP: as.nyu.edu/research-cen...

hivemind what are your recommended reads for getting out of the chicken-egg cycle of co-design ?

We are very excited to announce our first workshop on From Theory to Practice: behind the scenes on research deployments at EC’26 (July 6 in Rome)! Call for posters and submissions now open! Organized by myself, @ericachiang.bsky.social , Bailey Flanigan, @brwilder.bsky.social

Deploying algorithmic research in practice is an opaque process. We're organizing a workshop at EC to share behind-the-scenes stories and move the field foward. Call for submissions open! With @nkgarg.bsky.social @ericachiang.bsky.social, Bailey Flanigan sites.google.com/cornell.edu/...

This story by @spettypi.bsky.social is extremely important because she documents in Arizona - SNAP participants losing access to the program who should be exempt from changing eligibility criteria - The state admitting that they need to deny eligible participants access to reduce their error rate.

Check our new work on OpenCourier — an open protocol for interoperable, worker-owned delivery platforms. By: Yuhan Liu, Varun Rao, @s0hw.bsky.social @cyberlyra.bsky.social and @andresmh.hci.social.ap.brid.gy Thanks to @pcc-global.bsky.social for amplifying it! platform.coop/blog/toward-...

why "ai for social impact/good" (however you want to call it) should get better at engaging with organizations and institutions that deliver social impact

new project for "trapping rogue AI web crawlers in an infinite self-generating labyrinth of babble" by @julian0liver.bsky.social - including "zip bombs" unpacking from 127mb > 128gb, designed to fill RAM and crash crawlers: julianoliver.com/projects/sci...