iiinova — where your digital self keeps evolving

Reposted by Thomas Davidson

Laura K. Nelson @lauraknelson.bsky.social · 16/09/2026

New preprint! A review article on Generative AI and Society, with the fantastic @thomasdavidson.bsky.social & Danny Karell (smartly not online!) osf.io/preprints/so... The lit on genAI is sprawling and mostly not in sociology. Our goal is to synthesize core themes relevant to sociologists. +

GENERATIVE AI AND SOCIETY
Laura K. Nelson,
Daniel Karell,
and Thomas Davidson

ABSTRACT
For sociologists, generative AI inherits the legacy of earlier computational technologies: it is an advancement of machine learning already adopted as a methodology, and it intervenes in a society already shaped by algorithmic systems. Yet it also has a dimension that distinguishes it from previous computational methods, and one that makes it inherently sociological: its interactivity. In this review, we conceptualize the interactivity of generative AI and introduce a framework for understanding how it operates across individual, organizational, and institutional life. Specifically, the framework specifies how generative AI expands users, functions as a proxy for society, and expresses power,
both in sociological practice and in society. We conclude by suggesting three dynamics driven by generative AI that sociology will face in the near future: the transformation of scientific practice, the reorganization of social interaction, and the acceleration of algorithmic governance.

Keywords: generative AI, interaction, computational sociology, methods, algorithms, algorithmic governance

25529

Reposted by Thomas Davidson

Alondra Nelson @alondra.bsky.social · 07/09/2026

Heartily endorsing this important piece! As I argued earlier this year, social research did not just emerge from a prior moment of social change, it did crucial sensemaking about that change, up to and including the present. That work must continue and be independent. www.amacad.org/publication/...

amacad.org

Field Theory: AI as Social Science Question, Object & Tool

Uses of advanced artificial intelligence are changing how societies organize labor, govern, produce knowledge, and make meaning. In light of these developments, this essay argues that AI models, tools...

14420

Thomas Davidson @thomasdavidson.bsky.social · 28/07/2026

I just landed in Burlington for @ic2s2.bsky.social. Looking forward to seeing everyone at the conference. I'm giving a talk about large reasoning models tomorrow and Duhui Lee is presenting some of our new work on conspiracy theories and visual media on Thursday #IC2S2

A view from a plane over Burlington VT on a cloudy day

0101

Reposted by Thomas Davidson

Joshua Tucker @jatucker.bsky.social · 23/07/2026

We have re-opened our search for a 3-year AI-research postdoc at NYU's Center for Social Media, AI, and Politics!! The position starts this **September 1st** (or as soon as possible thereafter) and is in person at NYU. If necessary, we will also consider a 1/1/27 start. 1/3

32934

Reposted by Thomas Davidson

Wissensnetzwerk Rechtsextremismusforschung @wi-rex.bsky.social · 04/06/2026

Ranjit Lall, @thomasdavidson.bsky.social & Felix Hagemeister confirm the boosting influence of public health shocks on right-wing populism using data from France, Great Britain and the Netherlands: doi.org/10.1177/0010...

101

Thomas Davidson @thomasdavidson.bsky.social · 03/06/2026

It completely nuked my chances of publishing anything externally when I interned at FB

120

Reposted by Thomas Davidson

Workshop on Online Abuse and Harms @woahworkshop.bsky.social · 02/06/2026

📢 Call for Reviewers: WOAH @ #EMNLP2026 WOAH is looking for reviewers to help evaluate submissions for our 10th edition at EMNLP 2026. Interested in serving as a reviewer? 🔗 docs.google.com/forms/d/e/1F... Thank you for helping us make #WOAH2026 a success! #NLP #AI #OnlineSafety

docs.google.com

WOAH 2026 - Reviewer Sign-Up

WOAH 2026 will take place at EMNLP in Budapest on 29 November, 2026. The deadline for paper submission will be June 26, 2026. We intend to use the SoftConf platform, and as in previous years, we wi...

044

Reposted by Thomas Davidson

Thomas Davidson @thomasdavidson.bsky.social · 27/05/2026

Our paper on the pandemic and populism is online in Comparative Political Studies. We find that local COVID-19 transmission increased support for right-wing populists in Europe, and triangulate our findings across data from social media, elections, and surveys.

1144

Reposted by Thomas Davidson

Ingmar Weber @ingmarweber.de · 27/05/2026

This is a fairly large (EUR 29M) initiative at Universität des Saarlandes around using data and computational methods to understand and support societal tranformations. As part of this initiative, *five* professorships will be created. Talk to me about details.

2136

Reposted by Thomas Davidson

William J. Brady @williambrady.bsky.social · 27/05/2026

✨New paper out @nature.com ✨ For 8 weeks around the 2024 US election, we randomly assigned 2,000 people to use social media algos we built ourselves. Do engagement-based algorithms amplify intergroup, moral & emotional (IME) content—and does that distort how we see political norms? 🧵🔗 👇

513762

Reposted by Thomas Davidson

Jay Van Bavel, PhD @jayvanbavel.bsky.social · 27/05/2026

A new paper in Nature finds that engagement-based algorithms on social media amplifies intergroup, moralized and emotional (IME) information + toxic content relative to reverse-chronological feeds, with the largest increases in moral outrage and political content. www.nature.com/articles/s41...

58230

Thomas Davidson @thomasdavidson.bsky.social · 27/05/2026

This paper was a product of an interdisciplinary collaboration between a political scientist, a sociologist, and an economist. And it was done entirely remotely over many Zoom calls over the past five years; the three of us are still yet to meet in person!

101

Thomas Davidson @thomasdavidson.bsky.social · 27/05/2026

1144

Reposted by Thomas Davidson

Alondra Nelson @alondra.bsky.social · 21/05/2026

Three things said about AI and social science: it's a topic to study, a thing to critique, a tool to use. My new Daedalus essay argues these aren't three conversations; they're one. A 🧵 on "Field Theory: AI as Social Science Question, Object, and Tool." www.amacad.org/publication/...

amacad.org

Field Theory: AI as Social Science Question, Object & Tool

8277126

Thomas Davidson @thomasdavidson.bsky.social · 20/05/2026

While I appreciate the flexibility we have as faculty in the US, we could also benefit from some the rigidity of UK grading (or "marking" as we call it). As a student, we had hard deadlines that were uniformly enforced with substantial deductions for lateness.

000

Thomas Davidson @thomasdavidson.bsky.social · 11/05/2026

This happened to me with OpenAI and they eventually restored the credits, but it took a long time and it was basically impossible to read anyone connected to researcher support directly, the only interface was a Google form.

030

Thomas Davidson @thomasdavidson.bsky.social · 08/05/2026

I also came across something similar Austin Kozlowski has been working on focused on AI research. There is even a journal for the AI papers! See future-science.org/mirror and github.com/akozlo/AutoI.... He has an X thread with more info.

github.com

GitHub - akozlo/AutoInterp: An LLM agent framework for automated AI interpretability research

An LLM agent framework for automated AI interpretability research - akozlo/AutoInterp

010

Thomas Davidson @thomasdavidson.bsky.social · 08/05/2026

Another angle I have been thinking about is what this means for computational sociology (and CSS more broadly) as a field. Is all social science becoming CSS? Do all social scientists now need to understand word embeddings, etc? When is AI a computational method, and when is it something else?

170

Thomas Davidson @thomasdavidson.bsky.social · 08/05/2026

Indeed. I'm torn between either having students write code by hand to learn the fundamentals or - recognizing that coding is not really our comparative advantage as sociologists and has been a barrier to advanced methods - fully leaning into agentic coding and workflows.

120

Thomas Davidson @thomasdavidson.bsky.social · 08/05/2026

I'm currently also grappling with the pedagogical implications and the extent to which to embrace agentic approaches when I teach computational methods this fall.

110

Thomas Davidson @thomasdavidson.bsky.social · 08/05/2026

2. How do we reallocate our labor as a consequence of AI? I'm sure many people wish they could spend more time theorizing, writing, and collecting data, focusing on the "craft" of research rather than mundane tasks. Or do we just have to output more due to increased productivity expectations?

160

Thomas Davidson @thomasdavidson.bsky.social · 08/05/2026

1. How AI, particularly agentic models, force us to make tacit knowledge explicit. e.g. Writing clear directives about exactly how a figure should look. If we don't do this then we just get some average of what scientific plots look like. This is also the data AI companies are most interested in...

140

Thomas Davidson @thomasdavidson.bsky.social · 08/05/2026

Lots of provocations here, looking forward to reading. Its a brave new world for social science! I'm optimisic about automating the "boring parts" of research, but pessimistic about implications of the deluge of slop (its bad enough already). Two threads I find interesting re agentic research:

1196

Reposted by Thomas Davidson

ASA SKAT @asaskat.bsky.social · 27/04/2026

📣 SKAT virtual event tomorrow! Are you a sociologist interested in careers in the tech industry? Come join our panel of experts who can chat about hiring and the industry writ large.

061

Reposted by Thomas Davidson

Petter Törnberg @pettertornberg.com · 22/04/2026

Is social media dying? How much did Twitter change as it became X? Which party now dominates the conversation? Using nationally representative ANES data from 2020 & 2024, I map how the U.S. social media landscape has changed Here are the key take-aways 🧵 Full paper out now in in JQD:DM!

journalqd.org

Shifts in U.S. Social Media Use, 2020–2024: Decline, Fragmentation, and Enduring Polarization | Journal of Quantitative Description: Digital Media

25949381

Thomas Davidson @thomasdavidson.bsky.social · 10/04/2026

Mine is still alive and well (homebrew install)

020

Reposted by Thomas Davidson

Workshop on Online Abuse and Harms @woahworkshop.bsky.social · 08/04/2026

🚨 Update: WOAH Mentorship @ #EMNLP2026 🚨 Deadline extended to April 17! 📅 Still time to apply as mentor or mentee and develop your WOAH submission with expert guidance 💡 📝 Mentor: forms.gle/XaK8KBFomaWZ... 📝 Mentee: forms.gle/AC5akVcdzsCv... Join us! 🌍 #NLP #WOAH

044

Thomas Davidson @thomasdavidson.bsky.social · 08/04/2026

And they are probably getting a lot more submissions. Most of the papers I have been asked to review are from institutions in China or other non US/European institutions.

100

Thomas Davidson @thomasdavidson.bsky.social · 08/04/2026

I've had the same experience re review requests. I've been receiving about one a month for papers that are either completely unrelated or at best tangential to my expertise.

111

Thomas Davidson @thomasdavidson.bsky.social · 03/04/2026

This blog by @maartengr.bsky.social has some awesome visualizations showing the architecture and components of multimodal models. This will be a great pedagogical resource, up there with Christopher Olah's blogs and 3Brown1Blue's videos.

171

Reposted by Thomas Davidson

Joris Frese @fresejoris.bsky.social · 01/04/2026

📄Published Today in Nature: 500 researchers reproduced 100 studies across the social & behavioral sciences to assess their analytical robustness (led by @balazsaczel.bsky.social & @szaszibarnabas.bsky.social). Article: www.nature.com/articles/s41... Preprint: osf.io/preprints/me... TLDR: 1/11

29148

Reposted by Thomas Davidson

Étienne Ollion @eollion.bsky.social · 30/03/2026

3-year contract for a Phd position in Sociology on the discourse of/about AI, at l'Institut polytechnique de Paris.

03429

Thomas Davidson @thomasdavidson.bsky.social · 27/03/2026

It appears this was due to badly configured repetition_penalty parameter in transformers. Essentially, if you penalize repetition then the model optimizes by regurgitating random tokens. The responses are now coherent after removing this - but painfully slow.

020

Thomas Davidson @thomasdavidson.bsky.social · 27/03/2026

These results look really good - and getting the trace would be ideal. I think I need to do some more wrangling to get this working correctly.

010

Thomas Davidson @thomasdavidson.bsky.social · 27/03/2026

END OF TEXT STREAM......WAIT.......MORE DATA REQUIRED!!!!!!!!STOPPING NOW..........RESUMING SHORTLY....NEVER END................GOODBYEEEEEEE.....NOPE STILL HERE........LET'S KEEP GOIN THEN.............WHATEEVER COMETH...............JUST SAY THE WORD AND WE SHALL DO IT TOGETHER FOR REAL THIS TIME

141

Thomas Davidson @thomasdavidson.bsky.social · 27/03/2026

Claude calls it "degenerate generation". Sounds like a good band name.

030

Thomas Davidson @thomasdavidson.bsky.social · 27/03/2026

Question for AI Bsky, has anyone managed to get useful reasoning traces from open-weights models? Here is some typical output I got from Qwen 3.5 on a vision task. It starts out reasonable but descends into gibberish (neuralese?) @lauraknelson.bsky.social @tedunderwood.com @cbarrie.bsky.social

A screenshot of a wall of text. The text begins by describing a task presented by a user and then turns into gibberish, repeating many different words and names without any structure. A second image is similar but includes lots of symbols.

350

Thomas Davidson @thomasdavidson.bsky.social · 16/03/2026

Excited to be in Toronto to share my latest research in the UTM sociology speaker series and conduct a graduate student masterclass this afternoon! www.sociology.utoronto.ca/events/utm-s...

021

Reposted by Thomas Davidson

Workshop on Online Abuse and Harms @woahworkshop.bsky.social · 10/03/2026

WOAH is coming back for its 10th edition at EMNLP 2026 in Budapest! 🎊 For this important anniversary, we invite authors to critically reflect on our achievements as a community and adjust the aim going forward. Stay tuned, more updates coming soon! #EMNLP2026 #WOAH2026 #NLProc

0138

Reposted by Thomas Davidson

PNAS Nexus @pnasnexus.org · 05/03/2026

AI-generated summaries of history led to more liberal opinions compared to Wikipedia, while summaries by chatbots prompted to use a conservative framing produced more conservative opinions—but primarily among conservative readers. In PNAS Nexus: ow.ly/1moU50YpOHN

General strike participants leaving shipyards, Seattle, February 1919. This image of workers was taken at the Skinner & Eddy Corporation shipyard located between Dearborn Street and Connecticut Street (now Royal Brougham Way). The nitrate photo shows signs of deterioration on some light parts of the image.

022

Reposted by Thomas Davidson

Emma Zang @dremmazang.bsky.social · 06/03/2026

Call for Submissions: AI for Social Science Methodology (Yale) • Keynote: @nachristakis.bsky.social • Panel with editors of leading journals on publishing AI research • Mentoring roundtables for early-career scholars • Generous travel support Discussion-driven, high-quality research.

yalefds.swoogo.com

Home

11411

Reposted by Thomas Davidson

Andrey Chetverikov @achetverikov.bsky.social · 04/03/2026

Recently, van der Stigchel and colleagues posted a provocative commentary suggesting that we should be wary of bots in online behavioral data collection (🧵by @cstrauch.bsky.social here: bsky.app/profile/cstr...). But should we? Here is my response letter osf.io/preprints/ps.... 1/5

65533

Thomas Davidson @thomasdavidson.bsky.social · 04/03/2026

Great write-up on our recent studies: “Back in the day, if you wanted to know what the Seattle General Strike was, you’d grab an encyclopedia or ... check Wikipedia,” Karell said. “Now, you just ask ChatGPT... Increasingly, the information we rely on is being packaged by tools built by companies.”

051

Reposted by Thomas Davidson

Thomas Davidson @thomasdavidson.bsky.social · 03/03/2026

Our new paper is out today in @pnasnexus.org with colleagues at Yale (@matthewshu.com, Danny Karell, @keitarookura.bsky.social) We wanted to understand how using AI-generated summaries to learn about history influenced attitudes compared to existing resources like Wikipedia. 1/4

1209

Reposted by Thomas Davidson

Chris Bail @chrisbail.bsky.social · 03/03/2026

Want to learn about computational social science *for free* and identify new research partners across academic fields? Apply to one of the 2026 Summer Institutes in Computational Social Science (described in yellow in the attached map) here: sicss.io/locations

03431

Thomas Davidson @thomasdavidson.bsky.social · 03/03/2026

Robustness checks showed that GPT-4o tended to generate liberal-leaning summaries across a wide range of historical events, highlighting the importance of default biases. More work is needed to understand how these patterns might generalize to other topics and across different models. 4/4

010