This was accepted to #NeurIPS 🎉🎊

TL;DR Impoverished notions of rigor can have a formative impact on AI work. We argue for a broader conception of what rigorous work should entail & go beyond methodological issues to include epistemic, normative, conceptual, reporting & interpretative considerations

I can't stress enough how close U.S. science is to the cliff.

"Numbers released in May by the National Science Foundation (NSF) indicate that if Congress approves the cuts to the agency proposed by the White House, the number of early-career researchers it supports could fall by 78%" (@science.org)

Very true. In our causal tutorial slides we point at the edges in a DAG and say the edges we included aren’t the assumptions, the edges we deleted are the assumptions. (Never mind nodes!)

Having said that, LLMs are a useful tool for brainstorming what might be missing, complementing domain experts

Congratulations @kous2v.bsky.social!

Want to learn more? Check out our paper! written with a fantastic group of coauthors Su Lin Blodgett @amabalayn.bsky.social @angelinawang.bsky.social @841io.bsky.social @fcalmon.bsky.social @mmitchell.bsky.social @md.ekstrandom.net @rdbinns.bsky.social @s010n.bsky.social arxiv.org/pdf/2506.14652 7/7

We have to talk about rigor in AI work and what it should entail. The reality is that impoverished notions of rigor do not only lead to some one-off undesirable outcomes but can have a deeply formative impact on the scientific integrity and quality of both AI research and practice 1/

HAPPY 75th, NSF!

We’re celebrating this milestone by highlighting some of NSF’s most transformative accomplishments—innovations that have shaped our world and continue to drive progress in health, technology, the environment, and beyond.

Read on 🧵(1/11):

In this issue: our CHI 2025 & ICLR 2025 contributions, plus research on causal reasoning & LLMs; countering LLM jailbreak attacks; and how people use AI vs. AI-alone. Also, SVP of Microsoft Health Jim Weinstein talks rural healthcare innovation: msft.it/6013SHuu1

Carl Sagan when asked, "Are you a Socialist?"

Even more so when compared with the books left on the shelves.

Gift idea: what do you get for the AI researcher who has everything? 👇

Tariffs xkcd.com/3073

Seattle right now

Looking fwd to discussing career paths next week at the Society of Causal Inference & Online Causal Inference Seminar's joint webinar:

Exploring Career Paths in Pharma, Government, and Technology
w/Gabriel Loewinger and Natalie Levy
Tue Apr 15, 11:30am-12:45pm ET
sci-info.org/online-events/

Excellent statement from Nature’s editorial board.

The last section on “How to respond” is the most important.

www.nature.com/articles/d41...

My friends and colleagues at IU are being intimidated for their work on social media’s harmful impacts on civic discourse. Here is why their work is vital.

osome.iu.edu/research/blo...

Sharing an announcement from the Society for Causal Inference annual meeting --- they are looking for student volunteers, in exchange for complimentary registration

More info: sci-info.org/annual-meeti...

The abstract deadline for SIGIR's Industry Track is coming up in just a few days, with a final deadline of Feb 27.

So I'm a big fan of Claude, and anyone can goof up, but this did make me laugh.

Since its launch two decades ago, Microsoft Research India produced extraordinary innovation in areas from health and education to agriculture and accessibility. Learn about the lab’s track record of technological advances: www.microsoft.com/en-us/resear...

Contrastive focus reduplication is such a fun construction.

languagelog.ldc.upenn.edu/myl/llog/Sal...

Happy International Day of Women in Science. The National Science Foundation’s list of flagged words includes both “Women” and “Female.”

We should not be allowing non-government employees to waltz in to government offices, and illegally get access to sensitive government and personal data, and get read and write access to critical software systems in multiple government agencies.

www.npr.org/2025/02/08/g...