Virtual event starting soon! My lab @communitydata.science is hosting a discussion about new research on organizing software communities. Sign up & join us! wiki.communitydata.science/The_Impacts_...
Hey how's it going here? been a minute.
LLMs are now widely used in social science as stand-ins for humans—assuming they can produce realistic, human-like text
But... can they? We don’t actually know.
In our new study, we develop a Computational Turing Test.
And our findings are striking:
LLMs may be far less human-like than we think.🧵
kudos to @zarine.net whose research and expertise really drove this. thanks also to @gowder.io for inspiring conversation in the event that precipitated the writeup.
#2 distinct categories of threats require different response strategies, which must also be localized depending on context, resources, etc.
#1 public information goods created in online communities face distinct threats over different phases of their lifecycle (e.g., a new community software project that hasn't built much yet is vulnerable in different ways than a more mature project)...
(at least) two valuable ideas here...
Defending public information goods and online communities under attack. zenodo.org/records/1560... A @communitydata.science strategy brief led by @zarine.net (with @mako.cc and me riding shotgun)
a new short research brief with @mako.cc and @aaronshaw.bsky.social synthesizes some of the work I presented at @communitydata.science Science of Community Dialogues back in April! zenodo.org/records/1560...
Important message from @ethanvporter.bsky.social @aaronshaw.bsky.social @emmaspiro.bsky.social
“NSF has accepted a dangerous myth that misinformation research (and research more generally) is really just partisan activism […] Nothing could be further from the truth”
www.science.org/doi/10.1126/...
Some of this stuff is obvious, of course, but we think it's important to say certain obvious things out loud, in print, etc. when the attacks are overt, misinformed and/or made in bad faith.
Misinformation research continues to be urgent science. By @ethanvporter.bsky.social, @emmaspiro.bsky.social, and me. www.science.org/doi/10.1126/...
A policy brief on what generative AI simulations of people might be good for! (based on work with the amazing team of @joon-s-pk.bsky.social @cqzou.bsky.social @mako.cc @robbwiller.bsky.social @mbernst.bsky.social Merrie Morris, Carrie Cai, and Percy Liang)
🧵🧵🧵 effective commons governance doesn't mean you get the equilibrium you want...
That part of the timeline where I'm watching the White Sox game and realize maybe I'm in solidarity with the pope 🧦
anyhow, just awesome. thank you @rahaeli.bsky.social
And also also, author says they'd like to write a dissertation. Hermeneutical skills ✅➕
Also, while I'm waiting, has anyone got a word count on this monster?
Find me a better Bluesky thread. Go ahead, I'll wait.
University leaders are looking for capital. They have political capital. As with their endowments, they are inclined to sit on it. The thing about political capital though, is that if you spend it wisely it grows.
New interview up at the Protocol Oral History Project—Camille Acey on organizational closures: protocol.ecologies.info/interviews/a...
(what have we been waiting for? idk, but grateful that were here)
Northwestern faculty Senate has passed a version of the Rutgers mutual defense compact tonight, making it (I think) a dozen faculty governance bodies around the country to do so. Go cats!
AI-generated from a model prompted with someone's truth social feed?
👀👀👀
Person trailing after a car shouting "Uber STOP" as if the driver would respond to it like their name
welcome back!
the fascist Trump admin is coming after Wikipedia now. the letter reads like a crowd-sourced listicle.
"With Rogers’ arrest, there have been nine firearm-related busts in or near the Naperville Topgolf parking lot this year, Krakow confirmed."
