New paper in @nature.com with @lukebeehewitt.bsky.social, @isaiasghezae.bsky.social, & @robbwiller.bsky.social: Can LLMs predict results of social science experiments? Across 70 experiments, we find a strong correspondence (r=.85) between predicted and observed effects. doi.org/10.1038/s415... 🧵
Robb Willer
@robbwiller.bsky.social
Professor of Sociology, Psychology, & Org Behavior, Stanford University Direct, the @pascl-stanford.bsky.social, the AI for Public Benefit Lab, & co-direct @Stanfordpacs Center on Philanthropy and Civil Society
It was such a pleasure working with Ashwini, Luke, and Isaias on this project over the past few years. They are a lot of fun, take the work seriously, have great values, and are seriously awesome people. 🧵👇
New paper in @nature.com with @lukebeehewitt.bsky.social, @isaiasghezae.bsky.social, & @robbwiller.bsky.social: Can LLMs predict results of social science experiments? Across 70 experiments, we find a strong correspondence (r=.85) between predicted and observed effects. doi.org/10.1038/s415... 🧵
Interesting read for sociologists given how sociologists vary in whether anthropology is to be seen as a role model or cautionary tale.
New Q&A: Is anthropology hopelessly politicized, more than any other humanistic field? That's the charge that's been flying around this summer, so I wanted to hear from Carolyn Rouse, the president of the American Anthropological Association. She had a lot to say. www.chronicle.com/article/is-a...
As always: if there's any class of people who do not have due process rights, then *nobody* has due process rights, because the government can just claim someone is in that class and there's no process to establish otherwise.
I need people to understand that this administration’s litigating position is (1) federal law currently mandates the detention, without bond, of every noncitizen who is allegedly unlawfully present and (2) constitutionally, they don’t get a day in court. they’re generally losing, but they’re trying.
I watched the Questlove documentary Earth, Wind and Fire: To Be Celestial vs. That's the Weight of the World (HBO Max) and it was just about everything I could ask from a music doc. Enthralling, revelatory, and makes the case (as if the case needed to be made) for EWF's greatness AND weirdness.
it’s hard to explain how much shit like this has radicalized people. like i’m a white middle-aged broad who’s always been liberal but i’ve gone full-on “abolish ICE/reform scotus/the next dem candidate must commit to mass prosecutions” and you will not sway me from that position.
WHOA DHS does not say that the agent feared for his life… just the vague concern for “public safety.” So he killed him just because he did not want him in America, not because of any specific threat to agents?
New public pre-print with @aarslew.bsky.social and Martin Bisgaard 🧵 We introduce a perspective on democratic erosion that focuses on quiet anticipation among political elites: what do they think happens to those who cross a strongman that challenges democracy? www.researchgate.net/publication/...
(PDF) The Imagined Strongman: How Elites Anticipate the Cost of Opposing Trump
PDF | What happens to politicians who speak up against a strongman that challenges democracy? Political scientists typically answer this question using... | Find, read and cite all the research you ne...
researchgate.net
“The most preposterous rainbow was coming out of the castle I was supposed to be going home to... It looked like something out of My Little Pony.” - Tom Holland on the rainbow that delayed shooting on Christopher Nolan's The Odyssey.
Boy this was a long road, but glad it's finally out! @dgrand.bsky.social @robbwiller.bsky.social @arechar.bsky.social (along with Michael N. Stagnaro, James N. Druckman, not tagable)
Representativeness and response validity across nine opt-in online samples
The forensic report is available here: storage.courtlistener.com/recap/gov.us... Incredible stuff
The testimony of Julian Ackert, the forensic expert in the Gino case, is fascinating for computer nerds. <raises hand> Whoever changed the metadata might have gotten away with (at least part of) it if they had remembered to include seconds in the timestamp.
New short paper w @jkalla.bsky.social ! Candidates gain from moderation, but less than many theories expect. Many conclude voters must not care about issues. This is wrong. Small *average* effects mask large effects on specific issues & are consistent with widespread issue-based voting 🧵
This is a welcome and important step to normalize and popularize the use of Registered Reports. Thanks @magdalenaskipper.bsky.social for your leadership, and for the work of the neuroscience and behavioral research editors to test and validate the model at Nature.
Excellent news! Nature is expanding Registered Reports to all the fields in which they publish! A great result by all those proposing this strong solution to publication bias and selective reporting (and getting expert feedback before data collection)! www.nature.com/articles/d41...
Glad to see this WP from our lab out today, a labor of love over last year. Kudos to @davidbroska.bsky.social and @jonnekamphorst.bsky.social for their hard work on it! Please share any comments, suggestions, or criticisms of the research. All are welcome!
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...
New in Nature Climate Change, Prof. @robbwiller.bsky.social and @pascl-stanford.bsky.social examine the most persuasive climate change messaging strategies. bsky.app/profile/robb...
Climate change impacts are here, but public support for action remains polarized. Which climate messages move people? In a ~13,500-person megastudy, we tested 10 of the most-cited messages. Six increased pro-environmental attitudes in the U.S.—but only by 1–4 percentage points. 🧵
Nina. Newport Jazz Festival, July 4, 1968. David Redfern.
Dear ChatGPT, Am I the Asshole? While Reddit users might say yes, your favorite LLM probably won’t. We present Social Sycophancy: a new way to understand and measure sycophancy as how LLMs overly preserve users' self-image.
New study from @oii.ox.ac.uk researchers finds friendlier chatbots make more mistakes. Lead author @lujain.bsky.social, second author Franziska Sofia Hafner, senior author @rocher.lc. Read more: www.bbc.co.uk/news/article...
Friendly AI chatbots more prone to inaccuracies, study finds
Researchers found adjusting AI systems to be more warm and friendly to users would result in an
bbc.co.uk
New ASR paper finds no effect of state right-to-work laws on employment or number of business establishments (1946-2019). Argues that state spillovers, such as tax breaks in non-right-to-work states, explain the non-effect of right-to-work. gated: journals.sagepub.com/doi/10.1177/...
The past decade has had a silver lining: Chuck evolving from Jordanesque business class self-interest to a moral leader on behalf of LGBT people and immigrants
The relevant portion of the Jason Collins discussion
Considerable progress has been made in the field of leadership. However, this is undermined by a commitment to ideas which have been repeatedly debunked that refuse to die--called "Zombie Leadership" This paper outlines 8 core claims of zombie leadership: leadershiphq.com.au/wp-content/u...
BREAKING: Congressman Johnny Olszewski (D-MD) is introducing a constitutional amendment to establish term limits for Supreme Court Justices. The ROBE Act would establish 18-year term limits for justices.