Joshua Gans

AI is apparently already accelerating science as a co-intelligence.

Measuring academic publications of authors: “we find that productivity among GenAI users rose by 15 percent in 2023 relative to non-users and further increased to 36 percent in 2024” and the quality of publications also went up.

Kevin Bryan has a whole suite of tools. He even reinvented PowerPoint. www.kevinbryanecon.com/tools.html

Or just have them write the paper and cut out the intermediary.

So I just want to shout out to the referee reports I received that raised issues and even though they were negative caused me to be more energised and motivated about improving my paper. This is what service to the profession is.

This is feedback you rarely get following, say, seminars which for all the complaints in economics are often more superficial than a deep reading of a paper. Although you do often get from a good conference discussion.

But … the reviewing process itself can be extraordinarily helpful. Ideally, some come to your work fresh, delve into it deeply, and determine where it is persuasive, wrong, or somewhere in between.

These days we tend to see the review process as a gatekeeping function — keep out the bad and let in the good. And many referees do believe their role is to keep the gate. And with my 1994 JEP on rejections, I have surely helped reinforce that role.

I just received a rejection from a top journal and I am pleased. Let me explain as it is a testament to our system of peer review.

The other day a student asked me about the prevalence of insider trading in prediction markets. I now have an answer.

Just over 10 days left to apply to join our area at Rotman. jobs.utoronto.ca/job/Toronto-...

Not in your interest given Peter theil and the treas sec

I don’t think this has anything to do with gay at all Rafe.

All leading bookstores in December. Also I think chapter by chapter will be open access at MIT Press.

It has arrived. Out in December. Cover by me with ChatGPT assistance.

This is one of the features we are most excited about. Took a year of development but it is the foundation for winning students back from ChatGPT to learning again. Take a look. Available to institutional subscribers right now.

So I tried it out now on Laffont, Rey and Tirole (RJE, 1998b) which has a (big) error in Prop 5. Gemini found it, doubted itself and when I pushed confirmed it. Claude missed it but found it when I asked. Gpt-5-pro failed even when asked specifically.

Back in graduate school, Paul Milgrom asked me to examine a published paper from 1984 by another person that he suspected had an incorrect proof. I found the error. I decided to see if LLMs could. Only Gemini 2.5 Pro did so. Claude Opus and GPT-5-pro found no significant errors.

Economics of Transformative AI Workshop, Fall 2025

The videos from the NBER Economics of Transformative AI conference are well worth your time. www.nber.org/conferences/...

YouTube video by NBER 2025, Economics of Transformative AI Workshop, "Genius on Demand: The Value of Transformative..."

Talk on my latest paper with Ajay Agrawal and Avi Goldfarb, "Genius on Demand" youtu.be/by7k-H7sZ8A?...

Radiologists are the canaries in a coal mine ... because they are thriving

The radiologists are doing well. My latest. open.substack.com/pub/joshuaga...

Was the War on Tylenol in Tim Snyder's book?

If you want impact, tailor your research for AI - Research Professional News The future of science will involve integrating established practices with machine intelligence, says Joshua Gans

Free-to-read opinion: Rather than replacing traditional research approaches, AI can serve as an additional pathway for knowledge translation and application, writes @joshgans.bsky.social

www.researchprofessionalnews.com/rr-news-euro...

Berry-Hausman-Pakes Should Win the Nobel Prize A prize for revolutionizing industrial organization

I am not sure about a Nobel prize list contender argument that includes Aidan Toner-Rodgers is worth taking seriously. nicholasdecker.substack.com/p/berry-haus...

What I show however is that the Rajan result can be recovered more straightforwardly. It arises in the usual property rights theory of the firm when employees have non-contractible actions and when standardisation makes employees replaceable too.

Rajan didn’t actually get anything wrong. That same effect is implicit in his address but he worked hard to get his narrative. So hard in fact that his narrative was that this effect was countered by early employees vying for succession.

The paper focusses on that standardisation choice made by the founder. It shows that founders will not choose to make themselves replaceable because the returns from higher exit value are less than what they would receive for their on-going role. That explains Meta, Oracle and co

My recent paper on “Endogenous Firm Ownership” however suggested that firms will be founded by people who had a long-term role in the firm — i.e., were not easily replaceable. How to reconcile this with Rajan is what this paper is about.

Rajan in his 2012 AFA Presidential address said that entrepreneurial firms needed to undergo two transformations. They need to differentiate to create value and they needed to standardise to limit the hold-up power of the founder themselves.

When Will Entrepreneurs Choose to Make Themselves Replaceable? Founded in 1920, the NBER is a private, non-profit, non-partisan organization dedicated to conducting economic research and to disseminating research findings among academics, public policy makers, an...

I have a new paper out this week “When Will Entrepreneurs Choose to Make Themselves Replaceable?” A thread. www.nber.org/papers/w34271

I think we will be OK ... from AI deciding to kill us that is

I think simple economics tells us that superintelligent AI is not going to be threatened by us. open.substack.com/pub/joshuaga...