AI is amazing but...what should social scientists do now if humans are no longer the "gold standard" for many of our research tasks? In a new ppr we (w @JamesBisbee) show how to think abt results and their robustness given fast pace of technology change. Paper/software below!

Happy to see this published: https://t.co/Z1xleixZZk. Janet Yellen experienced sexism during her tenure as Fed Chair, from both men and women, and Dems and Reps. Lots of analysis went into demonstrating this simple and unsurprising conclusion. @atkern and @NicFraccaroli

🏦When do people report their true views on the economy? ➡️@janzilinsky & @JamesBisbee find that less-politicized survey questions better capture actual public welfare—beyond partisan bias https://t.co/M0EqxJiodO #FirstView

Political scientist in the southeast? Have cool research that you want feed back on? Want to network with other scholars? Consider applying to SoWEPS @ukpoliscidept Deadline February 17th! https://t.co/nsyGII8c6I

Today in @Salon, Stephen Skowronek, Desmond King, and I offer some thoughts on the second Trump administration's initial actions and the unitary executive theory. https://t.co/NCjNMAru7M

Of course, he's off X now, but you can find this diatribe and more on his substack:

The juxtaposition of our concern at the time of writing with the current vibes of X is a good reminder of how relevant Kevin Munger's diatribes about temporal validity are.

"This is the part of the article when we normally would 'punt' the outstanding questions to 'future research.' Except in the case of Twitter at the time of writing, our capacity to conduct the type of analyses we have described has been destroyed."

Quainter still is our hand-wringing over the loss of access to Twitter's API, in light of the loss of access to public data such as the census.

Publishing this in January of 2025 seems almost quaint, given the current assault being waged by Musk on democratic institutions.

Academics left X after Musk's take-over, especially so after he reinstated thousands of accounts that had previously been banned for violating then-Twitter's ToS, including President Trump.

AI being able to produce papers that aren't easily distinguishable from human-authored ones is potentially an extremely positive development. Even before AI, in every field I'm familiar with, the amount of published stuff exceeds the community's collective bandwidth to absorb and

Just a reminder that the internet (which made these tech bros rich) was the result of NSF research grants that built the original backbone of the web on university campuses, followed by innovative public/private partnerships. US tax dollars BUILT these billionaires.

I'm glad to see our paper with @JamesBisbee & @atkern cited by @guardian https://t.co/LVhdJPSw9V

One of the lessons from the past month or so is that people don’t understand the podcastverse, its ideological ecosystem, and/or how men/boys < 30 get their information/news these days.

Tariffs 👍🏽👍🏽👍🏽 Import taxes 👎🏽👎🏽👎🏽

📢 Episode 119 is live! 🎙️ In this 𝑳𝑩𝑺 episode, @alex_andorra & @rmkubinec discuss corruption research, the challenges of measuring it, and survey techniques like randomized response and more! 🎧 https://t.co/Io3MJzjEv4 #LearningBayesianStatistics #Bayesian #Causal

Four years ago, my coauthors and I traded notes on pro-Trump shifts we observed in Latino precincts across the U.S. There was considerable skepticism about whether these changes were real or durable. Tonight suggests the skeptics were wrong on both counts.

🚨 Just published in @Nature w @gemmadipoppa 🚨 We argue that reducing crop burning in South Asia, responsible for toxic air for 100s of millions, is not completely intractable. We show bureaucrats exert control when incentivized, w large consequences for pub health Details 👇

Hmm....