What is the definition of "pro-worker AI"? Co-director @dacemoglumit.bsky.social explains: pro-worker AI creates new, complementary tasks for workers to do, rather than purely automating tasks.

Two thoughtful essays about the impact of LLMs on graduate education:

ergosphere.blog/posts/the-ma...

economics.mit.edu/sites/defaul...

From personal experience I think the self-control problem mentioned in the first essay is very real.

Freudian lapsus.

I'm not sure you can call it "fact-checking" when a professional body says you misstated their position and your PR team says "no we didn't"

Grade inflating high-school teachers affect their students' long-term outcomes, from Jeff T. Denning Rachel L. Nesbit, Nolan G. Pope, and Merrill Warnick www.nber.org/papers/w34952

excellent paper. worth reading for anyone interested in competing AI models.

Using an Ultimatum Game Experiment to examine what happens if AI instead of humans take decisions finds that, sometimes, AI can be unusually altruistic, from Douglas K.G. Araujo and Harald Uhlig www.nber.org/papers/w34919

Happy Birthday, Wealth of Nations - Econlib Today marks the 250th anniversary of the publication of Adam Smith‘s An Inquiry into the Nature and Causes of the Wealth of Nations on March 9, 1776. Wealth of Nations remains a remarkable book, not…

"Happy Birthday, Wealth of Nations" www.econlib.org/econlog/happ...

Big drop in international students coming to US (particularly from India)

So sad
#academicsky #econsky
marketdesigner.blogspot.com/2026/03/big-...

A picture of PEW poll on global attitude survey, saying the % who rate the morality and ethics of people in their country as good vs bad, where the US has the worst rankings and Canada the best

Americans: we live in a fallen state—embroiled by sin, cheating, lying, and evil. You cannot trust anyone, not even those who claim to know you best

Canadians: I love my neighbors and my friends!

"Powerful AI can statically help human decision-makers, but can harm collective knowledge building... it can lead to what we call “knowledge collapse” whereby in the long-run all human knowledge is ultimately destroyed.”
economics.mit.edu/sites/defaul...

This language AI folks have adopted makes me think we’re in a spy novel.

In the absence of AI access, higher-education participants outperform lower-education participants by 0.548 standard deviations; with AI access, this gap falls to 0.139 standard deviations, from Cruces, Fernández Meijide, Galiani, Gálvez, and Lombardi www.nber.org/papers/w34851

So yeah, already fourth iteration of my Data Analysis with AI course. gabors-data-analysis.com/ai-course/ All open source. I have an intro overview of AI and its use in data. gabors-data-analysis.com/courses/da-w...

great thread

This is a great point and very illuminating.

I totally agree that we should care deeply about the philosophy of justice. I teach a course on economic inequality and I make a point to start with a (necessarily brief) discussion of what philosophers say about justice, which economists are typically ignoring, with their (our) utilitarian blinds.

While your exaggeration for rhetorical effect in the first sentence of your reply is not to my taste, I certainly agree that a book on justice must have a philosophical undergirding and it is informative to say what this undergirding is, and I appreciate this information.

While I am sympathetic to your statement, I fail to see how a book with “no philosophy” can possibly discuss justice seriously.

C.I.A. World Factbook Ends Publication After 6 Decades

C.I.A. World Factbook Ends Publication After 6 Decades www.nytimes.com/2026/02/05/u...

Chart of bitcoin over last year, including cliffdive over last week

Wow

Journal of Economic Perspectives Vol. 40 No. 1 Winter 2026

New JEP is out, with an excellent symposium on competition in labour markets:

www.aeaweb.org/issues/836?t...

Global Labor Mobility between Shrinking and Growing Labor Forces (Winter 2026) - Falling fertility and improved mortality create a powerful and inexorable demographic arithmetic of ageing in the coming decades around the world, with three patterns. The richest coun...

The native workforces of today's wealthiest countries are rapidly disappearing due to fertility that is low and very unlikely to rise—ever, and certainly not in time to make a difference.

The brilliant Lant Pritchett spells out the implications for migration policy:

www.aeaweb.org/articles?id=...

The CEPR Virtual IO Seminar The CEPR Virtual IO Seminar (VIOS) was created to provide an alternative to IO seminars cancelled due to travel restrictions brought on by the Coronavirus. The long-run goal, however, is to turn it in...

#EconSky Very happy to join the organizing team of the Virtual IO seminar of @cepr.org

Great speakers and discussants coming! See the program in the link. Note that you can also have access to past seminars recording!

cepr.org/events/event...

I was listening to the latest @appstories.macstories.net episode this morning and @viticci.macstories.net and @johnvoorhees.macstories.net specifically talked about this kind of risk with OpenClaw

hiring is slowing even faster than job postings

"The U.S. again underperforms, with hiring falling faster than the global average... The widening wedge between postings and hires underscores a shift toward cautious, incremental hiring rather than aggressive workforce expansion." www.reveliolabs.com/news/macro/a...

Fig 4 from Zhang, Lee, Liu "Statistical Learning Theory in Lean 4: Empirical Processes from Scratch" The dependency graph of the formalizations. Diagram shows proof of Dudley's entropy integral with preceding lemmas and Gaussian Lipschitz concentration likewise, feeding into a Gaussian Complexity inequality and an error bound for critical radius then used to prove sharp minimax error rates for linear regression.

Never have I felt more like my job will soon by taken by AI. Statistical learning theory in Lean: concentration inequalities, Dudley's entropy integral, and local Gaussian complexity bounds.
30000 lines of code, over 1000 lemmas, formalizing Wainwright and Boucheron et al arxiv.org/abs/2602.02285

"Hate gets more powerful with more hate, the only thing more powerful than hate is love. We need to be different. If we fight, we have to do it with love."

Measuring how frontier research frames what is normatively at stake along the efficiency and equity dimension, from Sebastian Galiani, Ramiro H. Gálvez, Franco Mettola La Giglia, and Raul A. Sosa www.nber.org/papers/w34714