Professor of Sociology, Psychology, & OB @Stanford Direct @PASCL_Stanford and the AI for Public Benefit Lab (@ai4pb_stanford)
Palo Alto, CA
- Tweets 8,388
- Following 2,794
- Followers 7,657
- Likes 22,309
Pinned Tweet
Saturday's No Kings protests were massive, inclusive, and nonviolent. It was an honor to team up with @owasow to analyze how nonviolent resistance can wins hearts, minds, and elections.
3
7
33
8,064
Robb Willer retweeted
New working paper: Can AI forecast treatment effects in economics from pre-registration documents?
Answer: Yes. 1/
21
120
434
92,423
Robb Willer retweeted
"Reviewers without training found 27% of the major errors, and training barely helped. We gave the same papers to 15 LLMs. They found 53-96%. Even the smallest models found about twice as many major errors as human reviewers"
Why we need human AND LLM review
Lots of debate this week on whether journals should ban AI in peer review.
We actually tested this in our new preprint, and the results were pretty clear to me.
For context, I've reviewed for most of the big journals and evaluated 1,000 human reviews as an editor. I love to read and evaluate science.
TLDR: From now on I won't do any peer review without AI. For clinical trials, I'd even call human-only review dangerous 🧵
1
7
39
3,179
Robb Willer retweeted
Lots of debate this week on whether journals should ban AI in peer review.
We actually tested this in our new preprint, and the results were pretty clear to me.
For context, I've reviewed for most of the big journals and evaluated 1,000 human reviews as an editor. I love to read and evaluate science.
TLDR: From now on I won't do any peer review without AI. For clinical trials, I'd even call human-only review dangerous 🧵
75
323
1,178
116,946
Robb Willer retweeted
Out today in Nature Human Behaviour: our review of large language models in computational social science, with @xu_fengli, @yongli and colleagues at @Tsinghua_Uni, @hkust & @MSFTResearch Asia. A thread on what we found, including where the field is fooling itself.
Read @ doi.org/10.1038/s41562-026-0…
6
111
450
25,447
Robb Willer retweeted
Second: privacy and discovery need not trade off. Federated learning and differential privacy block the data linkage social science depends on. The alternative: seed secure enclaves with AI social scientists that enter, analyze linked records, report robust findings and never leave.
1
1
2
539
Robb Willer retweeted
How do children learn the meanings of their first words? One idea is cross-situational learning: tracking which words go with which things across many situations. In a new preprint, we test this idea at scale on video from children's own perspective 🧵 arxiv.org/abs/2610.31842
ALT Figure 1: Training pipeline. (A) BabyView head-camera videos from 48 children are transcribed and turned into frame–utterance pairs. (B) A vision–language model learns to match words to regions of each frame, and is tested on 4-alternative picture choice trials. (C) Each pair is annotated for whether the utterance refers to something visible in the frame.
2
29
91
4,571
Robb Willer retweeted
Meta had an internal PR plan to try and discredit Jon Haidt because he was a critic of social media.
Meta was going to recruit academics, like Andrew Gelman, to write take down pieces. They were trying to turn academics against one another through a "whisper strategy".
This is a very serious problem for open academic debate. If academics who share negative evidence about a platform are targeted by campaigns to smear our credibility, we are in deep, deep trouble.
This creates a very serious conflict of interest for anyone who is paid to do this.
law360.com/articles/2542517/…
53
422
1,694
276,287
Robb Willer retweeted
Replicated this myself. Pretty wild!
mediaite.com/media/tech/goog…
44
87
764
228,681
Robb Willer retweeted
APPLY TO BE A 2027-28 CASBS PRACTITIONER FELLOW
Deadline: Dec 11, 2026
Learn about practitioner fellowships, the community practitioner fellows join, frequently asked questions & the application process, & access the practitioner fellowship application:
bit.ly/4i0vpRf
3
8
2,011
Cool quote from @carlislerainey's great piece!
I've also been thinking about the parts of social science I value that would be lost if we use AI to optimize the positivist aspects of social science, and what we can do about it.
There is valuable writing & thinking done even by positivist social scientists that would be lost in a world where AI helps us streamline papers to be extremely technically correct and faster to produce, writing and thinking that is aesthetic, exploratory, speculative, uncertain, subjective, and/or vague. That writing can be interesting, generative, and frankly fun. I don't want to lose it. And of course lots of academics work primarily in this register.
Important: "I believe that human research has an aesthetic quality in addition to its technical correctness. I've come to understand myself and my own research as interpretivist, even as it draws deeply and faithfully from the canon of positivist methodology in many respects."
1
3
11
2,055
Preserving ample space for speculation, exploration, and aesthetic writing can support even the strictly positivist work in many ways, including by exploring the vast space of ideas with little constraint.
But at least in my fields, it has caused big problems that our unfalsifiable content is so closely intertwined with falsifiable, positivist work. The former work would be much more scientifically useful if categorized clearly, rather than bundled with work focused on testable theory.models, rigorous data collection and analysis, etc.
1
79
Beyond just enjoying beautiful writing on interesting topics, I also expect new findings and insights will be coming at us faster than ever. It will be helpful to preserve space for us humans to process it all - sharing our work with one another - and allowing for imperfect, incomplete, not yet testable thought that we freely share with one another as we collectively work through it all.
76
Robb Willer retweeted
🗓️ 𝐃𝐨 𝐲𝐨𝐮 𝐨𝐫 𝐲𝐨𝐮𝐫 𝐬𝐭𝐮𝐝𝐞𝐧𝐭𝐬 𝐝𝐨 𝐬𝐮𝐫𝐯𝐞𝐲 𝐞𝐱𝐩𝐞𝐫𝐢𝐦𝐞𝐧𝐭𝐬?
On 12/3, I'm hosting a free, 1 hr. online seminar called, "Doing Better Survey Experiments" from 1pm - 2pm (ET). Please help spread the word if you can! 🙏
instats.org/seminar/doing-be…
1
18
44
3,981
Robb Willer retweeted
Now published at @apsrjournal. Check out "The Symbolic Politics of Housing" doi.org/10.1017/S00030554261…
New results!
In "The Symbolic Politics of Housing," @dbroockman @j_kalla & I showed that public opinion about housing policies correlates w/ affect towards the groups that the policies make salient (via framing or criteria in the policy itself).
🧵/19
1
45
140
9,305
New paper: "AI Article Reviews Dominate Human Peer Reviews at Their Best: Evidence and Policy Choices" See: GaryKing.org/peerless. with Steven Worthington. Comments welcome!
22
168
502
154,020
Robb Willer retweeted
varying user model varies p(eval). comparing with our sycophancy measure, we see that all models are more sycophantic the more they think they are deployed (vs being evaluated).
1
1
14
641
Robb Willer retweeted
A one-time 5% wealth tax would merely bring Silicon Valley billionaires' fortunes back to where they stood just a few weeks ago, at the beginning of September.
Back when their wealth was "only" 178% higher than on January 1, 2023.
They would rather see 3 million people lose their health insurance than allow that to happen.
10
81
320
35,666
Robb Willer retweeted
This is insane. We cannot allow AI agents to go into your bank account and act independently without your authorization. But that is what’s happening right now.
Left unchecked, AI could endanger the entire global financial system — including your bank account.
Pause AI NOW!
SCOOP: Major U.S. banks are detecting unauthorized third-party AI agents inside customer accounts, including some attempting to take actions that appear to be independent of user instructions, sources told Axios.
The findings have alarmed banking executives, who are racing to restrict access to their systems as AI moves at a velocity faster than financial institutions can respond, multiple industry and bank executives told Axios.
Several banks are now preparing for worst-case and even “day after” scenarios, including how to rapidly secure or restore customer accounts if things go wrong, the people said.
Several bank executives told Axios they're struggling to distinguish between human users and AI agents. One executive said banks also struggle to determine whether agents are authorized by customers or potentially operating without their permission, more akin to rogue agent behavior.
Read my latest for Axios here: axios.com/2026/10/22/ai-agen…
Community note
Per Axios, detected AI agents access bank accounts with the customer’s permission. This violates most banks’ terms requiring human logins, which is why banks call it “unauthorized.”
Source: Axios (Oct. 22, 2026) — axios.com/2026/10/22/ai-…
1,275
2,618
10,939
752,536