Chris Bail
Duke Professor directing https://scai.duke.edu, https://sicss.io & https://polarizationlab.com. Author of Breaking the Social Media Prism.
- Reposted by Chris BailFrom concerns about social media addiction to urgent civil liberties issues, courts are asking scientists to be arbiters of alleged technology harms. How can scientists reliably inform courts and how can courts interpret our work? New article with @penney.bsky.social tsjournal.org/index.php/jo...
- Reposted by Chris Bail[This post could not be retrieved]
- Reposted by Chris BailNSF director to resign amid grant terminations, job cuts, and controversy www.science.org/content/arti...
- Reposted by Chris BailNew preprint on the global state of political polarization research 🎉 osf.io/preprints/os... We discuss the many different ways polarization can be conceptualised, how it manifests very differently across countries and domains, and the problems with the dominance of US focused research.
- New paper suggests conversations with an LLM may increase people's support for climate change interventions: osf.io/preprints/ps...
- Can LLMS reproduce the treatment effects of field experiments in silico? Interesting new paper suggests they do so about 78% of the time: arxiv.org/pdf/2504.01167
- Using Facebook advertising data to improve survey sampling: A project started at the Summer Institutes in Computational Social Science! www.cambridge.org/core/journal...
- Grateful that some Republican leaders are pushing back against efforts to defund American universities: www.washingtonpost.com/opinions/202...
- I had the great honor to award Yann leCun Duke's inaugural "Distinguished Lectureship in Society-Centered AI" last week. One of my favorite parts of his lecture was his argument that one of the greatest risks of future AI models is that they may not be built by a diverse enough group of people.
- Reposted by Chris BailBluesky is the place for science.
- Penn State faculty position in computational social science/AI: psu.wd1.myworkdayjobs.com/PSU_Academic...
- Amazing lecture on creating confidence metrics within multi-agent LLM frameworks by UNC's Mohit Bansal here at Duke's Responsible AI Symposium today. You can still join for Yann LeCun and Ronnie Chatterji later today: www.responsibleaisymposium.com
- New preprint suggests large language models use trigonometry to do addition: arxiv.org/pdf/2502.00873?
- Reposted by Chris Bail[This post could not be retrieved]
- Want to learn about computational social science *for free* and launch interdisciplinary research projects? We are so excited to announce there will be *26* Summer Institutes in Computational Social Science this year! Apply to one of them here: sicss.io/locations
- Join us this weekend for the first-ever Society-Centered AI Hackathon at Duke as part of our inaugural Responsible AI Symposium. First 75 registrants get one of these t-shirts, and the winners get to present their work before Yann LeCun's keynote talk! Register here: www.responsibleaisymposium.com
- One of the coolest jobs in social media right now IMHO
- The Local Lab team at New_ Public is hiring for a Staff Researcher! The team's at a really exciting stage of developing a new pro-socially oriented platform, focused on supporting local community & connection- with lots of opportunity for research to inform strategy! newpublic.org/jobs/staff-r...
- Want to hear dozens of amazing AI research talks, develop new interdisciplinary conversations as part of Duke's new Society-Centered AI initiative, or participate in our organization's first hackathon? Check out our www.responsibleaisymposium.com February 28-March 3rd.
- Reposted by Chris BailLast year, we published a paper showing that AI models can "debunk" conspiracy theories via personalized conversations. That paper raised a major question: WHY are the human<>AI convos so effective? In a new working paper, we have some answers. TLDR: facts osf.io/preprints/ps...
- Reposted by Chris BailBreaking news: 22 states are suing to block NIH’s cutting of indirect costs. See our updated story on the Friday night news and its aftermath. scim.ag/4hTJQ6v
- Reposted by Chris BailAttention Attention Attention If anyone knows clinicians or patients who have knowledge of pediatric patients with the very rare cancer interdigitating dendritic cell sarcoma (IDCS) or other (non-LCH) dendritic or histiocytoc sarcomas please contact Sri
- Tiktok boosted Republican messages 11.8% more than Democrat messages during the 2024 U.S. Presidential Race according to this new study that employed bots to view ~394,000 videos.
- Reposted by Chris Bail[This post could not be retrieved]
- The National Science Foundation has now instructed all researchers to pause new purchases now too :(
- Reposted by Chris BailMore from inside NIH: Per a source with knowledge, for all internal research (of which there is like $10 billion worth or so), ALL purchasing shut down as of yesterday. That means gloves, reagents, anything involved with lab work, which means a lot of that work will stop.
- Cool new paper by Duke computational social scientist @jongreen.bsky.social on "curation bubbles": www.cambridge.org/core/journal...
- Reposted by Chris Bail[This post could not be retrieved]
- The National Institutes of Health has been asked to pause all communications, some are reporting study groups are "on pause" www.cnn.com/2025/01/21/h...
- A new $350k grant for assistant professors working on "diverse intelligences" from the Templeton World Charity Foundation: mailchi.mp/392f25043250...
- What Trump 2.0 means for science: www.nature.com/articles/d41...
- 51% of Americans now identify as independent of the two major parties, according to Gallup: ivn.us/posts/gallup...
- Cool post-doc with @janalasser.bsky.social : jobs.uni-graz.at/en/jobs/a122...
- New computational social science position at Konstanz: stellen.uni-konstanz.de/jobposting/2...
- Peak polarization discourse?
- Come join our international team studying whether we can simulate social media using LLMs to increase civility on our platforms.
- Job ad is now live! 18-month postdoc. with me @chrisbail.bsky.social @pettertornberg.bsky.social and Anikó Hannák my.corehr.com/pls/uoxrecru... Come work with a great team. Deadline is Jan 31 📆
- Welcome to Bluesky @jurafsky.bsky.social... instant follow for all things NLP and computational social science!
- Reposted by Chris Bail
I am recruiting PhD students at Duke! Please apply to Duke CS or CBB if you are interested in developing new methods and paradigms for NLP/LLMs in healthcare. For details, see here: monicaagrawal.com/home/researc.... - "Large language models surpass human experts in predicting neuroscience results" www.nature.com/articles/s41...
- The attempted assassination of president Trump decreased support for partisan violence and did not increase polarization according to a new study in the Proceedings of the National Academy of Sciences: www.pnas.org/doi/full/10....
- Our Polarization Lab at Duke hopes to hire another post-doc this year-- apply below and/or please share with people who might be interested: academicjobsonline.org/ajo/jobs/29305
- Applications to host a 2025 Summer Institute in Computational Social Science are due December 1st. Feel free to email me with questions and/or help connecting with others (especially since grant support will be limited): sicss.io/host
- Cool new post-doc at Northwestern/Kellogg: facultyrecruiting.northwestern.edu/apply/MjI3Nw==
- Consider submitting research on misinformation to this special issue of Behavioral Science and Policy: behavioralpolicy.org/wp-content/u...
- Come join our team!
- Really excited to be able to announce that I, along w @chrisbail.bsky.social @pettertornberg.bsky.social and Aniko Hannah won a T-AP grant. Our project will simulate SM platform algorithms to improve quality of political discussion. Job oppos. in 🧵 www.transatlanticplatform.com/activities/d...
- Wow: Boston University suspends PhD admissions in the humanities and some social sciences: www.insidehighered.com/news/admissi...
- Do podcasts play a decisive role in politics? This fascinating new paper maps the ecosystem using a large database of podcast transcripts from 2020. My first take away: politics is dwarfed by entertainment (not unlike social media): arxiv.org/pdf/2411.07892
- Cool job at Princeton open to Computational Social Scientists: puwebp.princeton.edu/AcadHire/app...
- LLMs can seem smart, but are they *wise*? Interesting new paper on the need for better metacognition among models: arxiv.org/abs/2411.024...
- Latest open-source LLM to rise to the top of the leaderboard is from China. Claims in paper need to be verified but pretty impressive if true (e.g. compares favorably to Llama 3.1), especially since they had less compute to work with: arxiv.org/abs/2411.022...
- Several days before the election, a paper was posted that employed GPT 4-o to predict the election outcome via data from the American National Election Study. It produced this electoral map (below). See arxiv.org/pdf/2411.01582
