[March 16 11:59pm ET] MIW 2026 Submission Deadline
The Marketplace Innovations Workshop is on May 18-20. The submission deadline is March 16. I'm not an organizer, but I really like this venue. It's virtual, and excellent. Hoping to see lots of great papers! marketplaceinnovation.net #socialcal
Still spending time over there? 🤔
A few people have asked for the syllabus from my grad seminar on Generative AI for social science -- just posted it here:
statmodeling.stat.columbia.edu/2026/03/10/n...
That's a surprise! 🤯 If there is information in law abstracts that isn't in the papers then, I agree, having AI come up with that would be quite different then if there isn't. (But yikes, that style choice!)
I do agree that the exercise of writing the abstract can be useful in one's own understanding. But, that is not what an abstract is for. I'm even happy to concede that if you can't, then you don't understand the paper. But, that's not what an abstract is for.
E.g. unconditioned on the technology. Maybe right now AI abstracts are lower quality than human abstracts. If so, I expect it to be short lived.
I'm going to disagree with this unconditional statement. E.g., on unconditional quality. An abstract does not add any new knowledge. Instead it gives a reader a concise summary of methods, results, etc. so they can ascertain its relevance. Readers want the best version of this. It could be from AI.
Economists don't have the problem of numeric citations. They use author-year which are good for identifying the paper, but per the study above, not perfect for equity.
Tl;dr: Empirically authors that appear at the top of the alphabet do better. The certified random system replaces the space between authors with an r with a circle around it.
The economists also use alphabetical, but studied its inequity and tried random, see this paper:
"Certified Random: A New Order for Coauthorship"
The system is usable. I tried it with a series of papers. I like it, however, until norms change, it's awkward.
www.aeaweb.org/articles?id=...
The economists also use alphabetical, but studied its inequity and tried random, see this paper:
"Certified Random: A New Order for Coauthorship"
The system is usable. I tried it with a series of papers. I like it, however, until norms change, it's awkward.
www.aeaweb.org/articles?id=...
So sad
#academicsky #econsky
marketdesigner.blogspot.com/2026/03/big-...
In 2005, back with @lance.fortnow.com wrote that blog post, citations in CS were all numeric. I hope he would have the good sense now to prefer author-year citations and \citet where the citation is used as a noun. Which, importantly, do not require flipping to the end.
@pkrugman.bsky.social raises a vital point.
At $1 billion/day, six weeks of the air war on Iran would cost as much as a year of US military support for Ukraine. Which would do more for US national security?
The answer may be debated, but the question is essential for America.
@poleconukinar.bsky.social returns with a series of great seminars, some of which are theory-focused 👇
Call for ACM SIGAI Autonomous Agents Research Award 2025
The award is made for research excellence in autonomous agents, to recognise researchers in autonomous agents whose current work is an important influence
Nominations Deadline: 15th Jan 2025
Details: https://buff.ly/4ir8hJR
#SIGAIAward
If you are a senior researcher (i.e., associate prof level or later), and you're interested in moving to the University of Waterloo (best CS program in Canada) for the Canada Impact+ Research Chair (biggest chair position in Canada, $$$$$), email me. Discretion guaranteed.
DOJ "is seeking to intervene in state bar assocs’ disciplinary proceedings against its lawyers." "The move comes as career lawyers in the department have repeatedly balked at carrying out orders they view as unethical or which could cause them to lose their licenses." www.nytimes.com/2026/03/04/u...
Pragmatic & actionable interpretability are buzzwords arguing for mech interp to study concrete tasks.
It's the right instinct, but still underspecified. What counts as a concrete task? What's upper bound on performance? What do users need to know? Decision theory has answers!
A good reminder for academics always to scrutinize where potential research funding comes from. www.science.org/content/arti...
Krishna Dasaratha, Anant Shah
Network Interventions: Targeting Agents or Targeting Links?
https://arxiv.org/abs/2602.12897
[Thurs, 8:30-9:45am PT] CSLaw 2026: Empirics
Empirics session at #CSLaw2026:
- “Quantifying the Benefits of Securities Regulation”
- “Mining Legal Arguments to Study Judicial Formalism”
- “AI-Assisted Moot Courts: Simulating Justice-Specific Questioning for Supreme Court Oral Arguments”
#socialcal
[Wednesday, March 11, 1-7pm CT] IDEAL Workshop on Automation in the Workforce: Ethics, AI, and the Reconfiguration of Labor
Great speaker lineup at this IDEAL-Loyola workshop. Part of the IDEAL Special Program on Data Science with Strategic Agents. www.luc.edu/quinlan/whyq... #socialcal
(I'm not on the organizing team.) Posting to get it on my calendar, and hoping to see many of you there!
[June 3-5] FORC 2026 (at Harvard)
The 7th annual Symposium on Foundations of Responsible Computing (@forcconf.bsky.social) will be held on June 3-5, 2026 at Harvard University. responsiblecomputing.org/forc-2026/ #forc2026 #socialcal
I'll be honest, while I really enjoyed #NeurIPS2025, the sheer scale made it hard to get a ton of scientific value out of it.
Looking forward to attending more small and intimate events like #ALT2026 and #FORC2026, where one can engage deeply with the pertinent communities.
I'd love to have a collection of empirical AI papers that do a great job of this. If you know any, please share!
In papers studying empirical outcomes of AI in various situations, e.g. game theory, the *first thing* I want to know is: What in the study is still relevant on completely different models with superior abilities? Research must still be relevant next week. I won't read if the paper doesn't say.
A line graph showing NSF grant awards made through 2/27/26 for fiscal year 2026 compared with grant awards for fiscal years 2021-2025.
NSF Update (Awards through 2/27/26)
Directorates to follow
1/10
[Nov 8-11] FOCS (in NYC) #socialcal
