Reposted
Finally! 🤩 Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia:
doi.org/10.5281/zeno...
We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical
thinking, expertise, academic freedom, & scientific integrity.
1/n
doi.org/10.5281/zeno...
We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical
thinking, expertise, academic freedom, & scientific integrity.
1/n
September 6, 2025 at 8:13 AM
Finally! 🤩 Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia:
doi.org/10.5281/zeno...
We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical
thinking, expertise, academic freedom, & scientific integrity.
1/n
doi.org/10.5281/zeno...
We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical
thinking, expertise, academic freedom, & scientific integrity.
1/n
Reposted
Perhaps most stimulating paper I've read this year explores the economic consequences of attention being scarce, rivalrous, cognitive, and volitional.
Reading it while listening to the Tony Rice Unit and drinking a lovely Jester King beer gets rather meta as my attention splits among them.
Reading it while listening to the Tony Rice Unit and drinking a lovely Jester King beer gets rather meta as my attention splits among them.
August 31, 2025 at 4:49 AM
Perhaps most stimulating paper I've read this year explores the economic consequences of attention being scarce, rivalrous, cognitive, and volitional.
Reading it while listening to the Tony Rice Unit and drinking a lovely Jester King beer gets rather meta as my attention splits among them.
Reading it while listening to the Tony Rice Unit and drinking a lovely Jester King beer gets rather meta as my attention splits among them.
Reposted
|￣￣￣￣￣￣￣￣￣￣￣|
Teach
Emeddings
Not
Prompting
|＿＿＿＿＿＿＿＿＿＿＿|
\ (•◡•) /
\ /
---
| |
Teach
Emeddings
Not
Prompting
|＿＿＿＿＿＿＿＿＿＿＿|
\ (•◡•) /
\ /
---
| |
One of the things that drives me up the wall is when people say students “need to learn how to use AI” to succeed in the workplace. It is EXTREMELY EASY to use a chatbot. That’s the whole point of it! You do not need training to type in the box! What are you even TALKING about?? +
August 30, 2025 at 3:43 PM
|￣￣￣￣￣￣￣￣￣￣￣|
Teach
Emeddings
Not
Prompting
|＿＿＿＿＿＿＿＿＿＿＿|
\ (•◡•) /
\ /
---
| |
Teach
Emeddings
Not
Prompting
|＿＿＿＿＿＿＿＿＿＿＿|
\ (•◡•) /
\ /
---
| |
Reposted
Learned a ton working on this survey. Hope it's helpful for you too, check it out! Feedback welcome 🙏
kennethmarino.com/computeruse/...
kennethmarino.com/computeruse/...
Super excited that the Computer Use survey I've been working on w/ @anamarasovic.bsky.social for a while now is ready! Originally we were planning on a more traditional survey paper but as more surveys came out we decided on an interactive website survey.
August 29, 2025 at 4:47 PM
Learned a ton working on this survey. Hope it's helpful for you too, check it out! Feedback welcome 🙏
kennethmarino.com/computeruse/...
kennethmarino.com/computeruse/...
Reposted
Happy to have played a small role in this big paper, which was led by Cody Kommers!
"We offer three principles for hermeneutic evaluation—that benchmarks should be iterative, not one-off; include people, not just machines; and measure cultural context, not just model output."
"We offer three principles for hermeneutic evaluation—that benchmarks should be iterative, not one-off; include people, not just machines; and measure cultural context, not just model output."
New preprint on "Computational Hermeneutics," co-authored by too many people to list in one post. TL;DR: GenAI is a cultural technology, and needs to be evaluated in ways that recognize situatedness, plurality, and ambiguity as the conditions of meaning — not noise to be minimized.
Computational Hermeneutics: Evaluating Generative AI as a Cultural Technology
<div>
<div>
<div>
<p>Generative AI (GenAI) systems are increasingly recognized as cultural technologies, yet current evaluation frameworks often treat cul
papers.ssrn.com
August 29, 2025 at 5:09 PM
Happy to have played a small role in this big paper, which was led by Cody Kommers!
"We offer three principles for hermeneutic evaluation—that benchmarks should be iterative, not one-off; include people, not just machines; and measure cultural context, not just model output."
"We offer three principles for hermeneutic evaluation—that benchmarks should be iterative, not one-off; include people, not just machines; and measure cultural context, not just model output."
Reposted
I happened to accidentally find this edited volume on quantitative approaches to literature, from 1969. Never heard of it - even though it was co-edited by one of the big figures in narratology: Lubomir Doležel (another surprise). The minimalist plots here are ⚡️
August 25, 2025 at 1:23 PM
I happened to accidentally find this edited volume on quantitative approaches to literature, from 1969. Never heard of it - even though it was co-edited by one of the big figures in narratology: Lubomir Doležel (another surprise). The minimalist plots here are ⚡️
Reposted
This analysis of how ideas spread in the early scientific revolution, from person to person and place to place, lines up so nicely with @alliemorgan.bsky.social's 2018 paper on the diffusion of scientific ideas via modern faculty hiring networks epjdatascience.springeropen.com/articles/10....
August 21, 2025 at 2:17 PM
This analysis of how ideas spread in the early scientific revolution, from person to person and place to place, lines up so nicely with @alliemorgan.bsky.social's 2018 paper on the diffusion of scientific ideas via modern faculty hiring networks epjdatascience.springeropen.com/articles/10....
Reposted
It’s not science fiction. www.vox.com/science/4226...
What if your earbuds could read your mind?
It’s not science fiction.
www.vox.com
August 7, 2025 at 5:44 PM
It’s not science fiction. www.vox.com/science/4226...
Reposted
📢 New paper, published by @knightcolumbia.org
We often talk about AI agents augmenting vs. automating work, but how exactly can different configurations of human-agent interaction look like? We introduce a 5-level framework for AI agent autonomy to unpack this.
🧵👇
We often talk about AI agents augmenting vs. automating work, but how exactly can different configurations of human-agent interaction look like? We introduce a 5-level framework for AI agent autonomy to unpack this.
🧵👇
August 1, 2025 at 3:28 PM
📢 New paper, published by @knightcolumbia.org
We often talk about AI agents augmenting vs. automating work, but how exactly can different configurations of human-agent interaction look like? We introduce a 5-level framework for AI agent autonomy to unpack this.
🧵👇
We often talk about AI agents augmenting vs. automating work, but how exactly can different configurations of human-agent interaction look like? We introduce a 5-level framework for AI agent autonomy to unpack this.
🧵👇
Reposted
👏 Upcoming workshop: Perspectivist Approaches to NLP @EMNLP 2025
Focusing on non-aggregated datasets and multi-perspective modeling, with sessions on labeling, modeling, evaluation, and applications.
nlperspectives.di.unito.it
Focusing on non-aggregated datasets and multi-perspective modeling, with sessions on labeling, modeling, evaluation, and applications.
nlperspectives.di.unito.it
NLPerspectives – Perspectivist Approaches to NLP
nlperspectives.di.unito.it
July 31, 2025 at 2:50 PM
👏 Upcoming workshop: Perspectivist Approaches to NLP @EMNLP 2025
Focusing on non-aggregated datasets and multi-perspective modeling, with sessions on labeling, modeling, evaluation, and applications.
nlperspectives.di.unito.it
Focusing on non-aggregated datasets and multi-perspective modeling, with sessions on labeling, modeling, evaluation, and applications.
nlperspectives.di.unito.it
Reposted
The CIA’s “Marshall Plan for the Mind” involved the smuggling of nearly 10 million books into the Eastern Bloc.
The Most Successful CIA Operation You’ve Never Heard of
How the agency’s program to circulate banned books helped take down the Iron Curtain.
foreignpolicy.com
August 2, 2025 at 7:30 PM
The CIA’s “Marshall Plan for the Mind” involved the smuggling of nearly 10 million books into the Eastern Bloc.
Reposted
One of the most-viewed PNAS articles in the last week is “Bridging theory and data: A computational workflow for cultural evolution.” Explore the article here: www.pnas.org/doi/10.1073/...
For more trending articles, visit ow.ly/Me2U50SkLRZ.
For more trending articles, visit ow.ly/Me2U50SkLRZ.
July 25, 2025 at 7:52 PM
One of the most-viewed PNAS articles in the last week is “Bridging theory and data: A computational workflow for cultural evolution.” Explore the article here: www.pnas.org/doi/10.1073/...
For more trending articles, visit ow.ly/Me2U50SkLRZ.
For more trending articles, visit ow.ly/Me2U50SkLRZ.
Reposted
Highlighting this thread. Based on what I'm seeing at #ic2s2 this week, this line of work is hot (if a bit crowded), but I predict will only be more widely adopted by social scientists in the future.
What are your favorite recent papers on using LMs for annotation (especially in a loop with human annotators), synthetic data for task-specific prediction, active learning, and similar?
Looking for practical methods for settings where human annotations are costly.
A few examples in thread ↴
Looking for practical methods for settings where human annotations are costly.
A few examples in thread ↴
July 23, 2025 at 1:07 PM
Highlighting this thread. Based on what I'm seeing at #ic2s2 this week, this line of work is hot (if a bit crowded), but I predict will only be more widely adopted by social scientists in the future.
Reposted
Paper published on the limits of classifiability with Mario Franco, Gerardo Febres, Nelson Fernández
epjdatascience.springeropen.com/articles/10....
We show that the nature of datasets imposes inherent limits on artificial intelligence and machine learning.
epjdatascience.springeropen.com/articles/10....
We show that the nature of datasets imposes inherent limits on artificial intelligence and machine learning.
The art of misclassification: too many classes, not enough points - EPJ Data Science
Classification is a ubiquitous and fundamental problem in artificial intelligence and machine learning, with extensive efforts dedicated to developing more powerful classifiers and larger datasets. Ho...
epjdatascience.springeropen.com
July 23, 2025 at 2:06 PM
Paper published on the limits of classifiability with Mario Franco, Gerardo Febres, Nelson Fernández
epjdatascience.springeropen.com/articles/10....
We show that the nature of datasets imposes inherent limits on artificial intelligence and machine learning.
epjdatascience.springeropen.com/articles/10....
We show that the nature of datasets imposes inherent limits on artificial intelligence and machine learning.
Reposted
In this Article, @oriplonsky.bsky.social et al. introduce a new model that merges behavioral science and machine learning to predict choice under risk and uncertainty. Tested on multiple large datasets, it exceeds top psychological and AI models. #BehavioralScience #XA
www.nature.com/articles/s41...
www.nature.com/articles/s41...
Predicting human decisions with behavioural theories and machine learning - Nature Human Behaviour
A new model merges behavioural science and machine learning to predict choice under risk and uncertainty. Tested on multiple large datasets, it outperforms top psychological and AI models, enabling ac...
www.nature.com
July 21, 2025 at 7:33 PM
In this Article, @oriplonsky.bsky.social et al. introduce a new model that merges behavioral science and machine learning to predict choice under risk and uncertainty. Tested on multiple large datasets, it exceeds top psychological and AI models. #BehavioralScience #XA
www.nature.com/articles/s41...
www.nature.com/articles/s41...
Reposted
Trump is picking winners and losers in all the emerging industries that will shape the global economy in the years to come.
Trump Embraces State Capitalism
The level of U.S. government economic intervention under Trump 2.0 is off the charts.
foreignpolicy.com
July 17, 2025 at 1:31 PM
Trump is picking winners and losers in all the emerging industries that will shape the global economy in the years to come.
Reposted
also this essay, which really ought to have been titled, "If you think you're a Silicon Valley classical liberal, maybe go and actually fucking read some classical liberalism why don't you" americanaffairsjournal.org/2024/08/no-e...
July 13, 2025 at 6:07 PM
also this essay, which really ought to have been titled, "If you think you're a Silicon Valley classical liberal, maybe go and actually fucking read some classical liberalism why don't you" americanaffairsjournal.org/2024/08/no-e...
Reposted
How We Misread the Social World—and Why That Matters datascience.columbia.edu/news/2025/ho...
How We Misread the Social World—and Why That Matters - The Data Science Institute at Columbia University
Most people assume they have a clear sense of the social world around them. They think that they know who spends time together, who confides in whom, and which relationships … Continued
datascience.columbia.edu
July 13, 2025 at 12:29 PM
How We Misread the Social World—and Why That Matters datascience.columbia.edu/news/2025/ho...
Reposted
The kind of comprehensive thinking we should be doing on AI, via @himself.bsky.social:
The Political Economy of AI: A Syllabus www.programmablemutter.com/p/the-politi...
The Political Economy of AI: A Syllabus www.programmablemutter.com/p/the-politi...
The Political Economy of AI: A Syllabus
As I’ve noted occasionally before, one of the most potentially useful things that academics do is preparing syllabi, and hence organizing information about the world.
www.programmablemutter.com
July 13, 2025 at 6:35 AM
The kind of comprehensive thinking we should be doing on AI, via @himself.bsky.social:
The Political Economy of AI: A Syllabus www.programmablemutter.com/p/the-politi...
The Political Economy of AI: A Syllabus www.programmablemutter.com/p/the-politi...
Reposted
Ever wanted to pan through the latent🌌 space of TikTok videos? Made using the amazing toponymy and datamapplot from @lelandmcinnes.bsky.social
and data from mine and @jurgenpfeffer.bsky.social
's first complete TikTok slice. link below
and data from mine and @jurgenpfeffer.bsky.social
's first complete TikTok slice. link below
July 11, 2025 at 4:45 PM
Ever wanted to pan through the latent🌌 space of TikTok videos? Made using the amazing toponymy and datamapplot from @lelandmcinnes.bsky.social
and data from mine and @jurgenpfeffer.bsky.social
's first complete TikTok slice. link below
and data from mine and @jurgenpfeffer.bsky.social
's first complete TikTok slice. link below
Reposted
Languages are systems. They can most certainly have biases, but they do not and cannot have goals. Exactly the same is true for the mathematical models of language that are produced by transformers ... the English language ... is never going to become conscious and decide to turn us into paperclips.
Cultural theory was right about the death of the author. It was just a few decades early
How old theories explain the new technology of LLMs
www.programmablemutter.com
July 10, 2025 at 2:37 PM
Languages are systems. They can most certainly have biases, but they do not and cannot have goals. Exactly the same is true for the mathematical models of language that are produced by transformers ... the English language ... is never going to become conscious and decide to turn us into paperclips.
Reposted
What does modern occupation look like online? 🤖
Join the DFRLab, OpenMinds & Eurasia Center on July 16 at 10:00a.m. for a live discussion on how pro-Russian Telegram bots flooded Ukrainian channels w/ messages simulating support for occupation: bit.ly/4lDuLrQ
Join the DFRLab, OpenMinds & Eurasia Center on July 16 at 10:00a.m. for a live discussion on how pro-Russian Telegram bots flooded Ukrainian channels w/ messages simulating support for occupation: bit.ly/4lDuLrQ
Digital occupation: Inside Russia’s Telegram battle in Ukraine
A new report—launched in partnership with the Digital Forensic Research Lab, Eurasia Center, and OpenMinds—uncovers how thousands of fake Telegram accounts were used in a covert Russian influence camp...
bit.ly
July 9, 2025 at 7:47 PM
What does modern occupation look like online? 🤖
Join the DFRLab, OpenMinds & Eurasia Center on July 16 at 10:00a.m. for a live discussion on how pro-Russian Telegram bots flooded Ukrainian channels w/ messages simulating support for occupation: bit.ly/4lDuLrQ
Join the DFRLab, OpenMinds & Eurasia Center on July 16 at 10:00a.m. for a live discussion on how pro-Russian Telegram bots flooded Ukrainian channels w/ messages simulating support for occupation: bit.ly/4lDuLrQ
Reposted
🚨New preprint on a verboten topic: social correlates of views on abortion.
Analyzing millions of geolocated tweets, we find a gender gap in abortion stance, show it widens in conservative states and identify the structural factors involved.
w/ @ashwinrao.bsky.social
📄 arxiv.org/abs/2507.05443
Analyzing millions of geolocated tweets, we find a gender gap in abortion stance, show it widens in conservative states and identify the structural factors involved.
w/ @ashwinrao.bsky.social
📄 arxiv.org/abs/2507.05443
July 9, 2025 at 6:56 PM
🚨New preprint on a verboten topic: social correlates of views on abortion.
Analyzing millions of geolocated tweets, we find a gender gap in abortion stance, show it widens in conservative states and identify the structural factors involved.
w/ @ashwinrao.bsky.social
📄 arxiv.org/abs/2507.05443
Analyzing millions of geolocated tweets, we find a gender gap in abortion stance, show it widens in conservative states and identify the structural factors involved.
w/ @ashwinrao.bsky.social
📄 arxiv.org/abs/2507.05443
Reposted
Hey, if you support CDT but can't give cash- give a retweet.
Maybe someone with more funds will see it.
They are almost at their modest goal of $10,000, but there's no reason 10,000 RTs shouldn't have been reached ages ago.
Maybe someone with more funds will see it.
They are almost at their modest goal of $10,000, but there's no reason 10,000 RTs shouldn't have been reached ages ago.
In recent months, we have lost a significant portion of our funding. Yet we remain committed to helping the world better understand China at this critical juncture in global politics. We are asking for your help to keep our websites up and running.
giving.chinadigitaltimes.net/page/serverc...
giving.chinadigitaltimes.net/page/serverc...
Keep CDT Online: Donate for Server Expenses
China Digital Times (CDT) is a non-profit organization that relies on institutional and individual donations to sustain our operations. In recent months, we have lost a significant portion of our inst...
giving.chinadigitaltimes.net
July 8, 2025 at 9:57 AM
Hey, if you support CDT but can't give cash- give a retweet.
Maybe someone with more funds will see it.
They are almost at their modest goal of $10,000, but there's no reason 10,000 RTs shouldn't have been reached ages ago.
Maybe someone with more funds will see it.
They are almost at their modest goal of $10,000, but there's no reason 10,000 RTs shouldn't have been reached ages ago.
Reposted
I’m not sure I’m happy with our new robot overlords. www.nytimes.com/2025/07/01/d...
The Map Rating Restaurants Based on How Hot the Customers Are
www.nytimes.com
July 8, 2025 at 9:02 PM
I’m not sure I’m happy with our new robot overlords. www.nytimes.com/2025/07/01/d...