I am sharing a set of slides from my lecture on agentic AI tools for Applied Economists at @penn_state. Materials are available here: github.com/ballesterogh/A-Pr… This builds on this thread and great work by Paul Goldsmith-Pinkham, @aniketapanjwani, and @VelikovMihail

Closing the second meeting of the AI and the Economy Initiative. Here is a thread with the highlights @PSULiberalArts @PennStateEcon @SmealCollege @ISTatPENNSTATE @PennStateCSRAI 1/n

When multiple companies deploy AI systems that learn from overlapping market data, optimize similar objectives, and operate at machine speed, a pattern emerges: Those systems tend to arrive at the same conclusions independently. Is your company deploying AI systems trained on the same market data as your competitors? Researchers call this the Agentic Convergence Trap. Understanding it requires understanding not just how AI systems behave, but how executives have enabled the behavior. Here’s how to avoid this trap. s.hbr.org/4uRi6WZ

Glad to see this paper forthcoming in IJIO. A meaningful milestone for me personally!

Excited to share that my paper on Algorithmic Collusion is now published in the International Journal of Industrial Organization. Many thanks to the referees and editor. Special thanks to @RanShorrer and @luliquesadalcs for their guidance along the way sciencedirect.com/science/ar…

@ballesterogh told us about strategic algorithmic monoculture. His work (w/ @RanShorrer @hadi_hoss and @samarthkhanna98) shows that LLMs adjust coordination to incentives and are effective convergence coordinators, but lag behind humans in diverging 6/n

Join us next week!👇

Terence Tao spent a year at the Institute for Advanced Study - no teaching, no random events of committees, just unlimited time to think. But after a few months, he ran out of ideas. Terence thinks that mathematicians and scientists need a certain level of randomness and inefficiency to come up with new ideas.

This winter I co-taught a grad seminar on when & how LLMs can stand in for human subjects in social science research, to a mix of CS & social science students. It was fun (albeit w/occasional friction) to talk methods w/such a diverse group. Reading list: statmodeling.stat.columbia.e…

📢 New policy brief: AI agents that can simulate human behaviors and attitudes can help test ideas in social science. Our latest brief introduces a generative AI agent architecture that simulates the attitudes of 1,000 real people. Learn more: hai.stanford.edu/policy/simu…

An excellent scholar and an even better advisor

When admissions are correlated, diversifying your college applications is the optimal strategy. We spoke with @SNageebAli and @RanShorrer of @PennStateEcon about why the standard advice to apply to reach, match, and safety schools is sound. aeaweb.org/research/applicat…

This is a useful paper on open-ended survey data. Open-ended questions can help reveal individuals’ motives, mental models and decision-making processes. Open-ended survey data have been used to study what’s on top of people’s minds w.r.t. important economic policy questions.

Excited to talk about AI at the CSRAI Rump Sessions next week. Hope to see you there!

Thrilled that my Minimalist Market Design monograph is forthcoming in the Econometric Society Monograph Series. The revised and much expanded version is now available on arXiv: arxiv.org/abs/2401.00307 This thread unpacks the framework — scholarly, reflective, personal. (1/20)

Early career mismatch matters: Overqualified are less motivated yet outperform others in the same job; underqualified exert more effort but still lag, from Julie Berry Cullen, Gordon B. Dahl, and Richard De Thorpe nber.org/papers/w34215

Come join us @PennStateEcon for the Midwest Trade & Theory Conference in mid-November. Submission deadline is September 6. Details here: sites.google.com/view/midwes…

Read more at: econ.la.psu.edu/news/ @PennStateEcon

La Nación citó un informe que escribmos sobre el Examen Nacional de Residencias Médicas en Argentina. El trabajo busca aportar evidencia al debate público sobre los sesgos del sistema de admisión