End of an era: As of this morning, crowdtangle, the Meta platform many researchers and journalists used to study public accounts on Facebook and Instagram, is officially dead. #RIPCrowdtangle In this 🧵I'll highlight some of the spectacular #polcomm work done with CT. 1/n

How easy it is to publish😁Email w/ no punctuation👇 "Greeting from the Journal Since it has been long time to be published your article We would like to publish your another article which is going to be publish in this issue Anticipating for your another article Have a nice day!

Wrong time to be an MBA, good time to be an AI builder, perfect time to be a plumber or an electrician?

Here are results from a sentiment analysis of television news coverage of Harris and Trump, up to today. Methodology is described in more detail on the website, and I'm hoping to update this page semi-regularly during the campaign. snsoroka.com/2024

📢New in PNASNews Our 28 bots responded to 28K+ users posting about sports, entertainment & lifestyle to encourage engagement w/ verified balanced news. Did it work? Read the paper & thread👇 W/ Hadi Askari Michael Heseltine Bernhard Clemm Anshuman Chhabra ✈️ ICML 2025 academic.oup.com/pnasnexus/arti…

Does this happen to other female academics? An article (by male scholar) citing my work: "Wojcieszak and *his* collaborators (2020) showed that..." #AcademicTwitter #Gender #assumptions

Interested in #harm detection & mitigation? See preprint w/ the *wonderful* Wonjeong Claire Jo👇 2-in-1: taxonomy of online harm & large-scale annotation using MLLMs, crowdworkers & experts. Core result: GPT outperforms crowdworkers but still a 'silver standard' osf.io/2dn8s

"To give yourself the best chance of winning a Nobel, you should ideally be born in North America and stay there."

In my mailbox, I have a folder 'Adulting'. And it is by far the least favorite folder in my mailbox.

LOL 😆

When will be the moment when AI will start being trained on data generated by AI?

#Reviewers, be *constructively* critical & avoid unnecessary attacks on an entire project. #Editors, check the tone of the reviews you receive before sending them out. Many authors are graduate & early career scholars who may be badly affected by senior jerks. #AcademicChatter

This is a *really* great opportunity - Princeton is hiring AI Policy Fellows to work directly with federal agencies to develop policy puwebp.princeton.edu/AcadHire/apply…

This paper shows key shortcomings in current LLMs regarding mathematical reasoning—especially under complex or altered conditions. We’re still far from true reasoning models, as LLMs rely heavily on pattern-matching. #AI #LLM #MathematicalReasoning arxiv.org/abs/2410.05229

#AcademicChatter Spending Saturday writing recommendation letters for undergrads applying to grad school & for grads applying for jobs. Easy to say no when asked as it takes time. But remember friends, u r where u r because someone wrote a rec letter for u #AcademicSuccess

In our Science Advances paper, we also find that the majority of Twitter users do not follow any politicians, journalists/pundits, or news media outlets 👇 science.org/doi/full/10.11…

I am recruiting 1-2 PhD students for Fall 2025 at Princeton to work on Comp Social Science/Societal Impact of GenAI/GenAI for SocSci I wrote a bit about my research flavor & interests here: manoelhortaribeiro.github.io/advising Deadline: December 15th (cs.princeton.edu/grad#prospecti…) Please boost!

The one and only Cristian Vaccari closing the 10th annual #IJPP conference in Edinburgh 🏴󠁧󠁢󠁳󠁣󠁴󠁿 And…closing his phenomenal 6-year tenure as the EIC of IJPP 🙏 for your leadership to you and the team @sgonzalezbailon Sophie Lecheler CW Anderson D Brown

200K+ people cancelled The Washington Post subscription. Now cancel Amazon prime👇 "Readers angered over the WP decision not to endorse a candidate for president shouldn’t cancel their subscriptions—they should cancel Amazon Prime instead" Ellen Cushing writes: theatlantic.com/culture/archiv…

the prompt "you are an expert peer reviewer. critique this paper." will give you a review that is easily top-tier in quality and comprehensiveness across all the frontier models (just tested in o3, 2.5 Pro, and Opus 4)