Ethan BdM@ethanbdm

Ethan Bueno de Mesquita. Studying conflict, accountability, and political economy @HarrisPolicy. "A tragic waste of intellectual potential." - R2

 Chicago, IL
voices.uchicago.edu/ethanbdm
Joined June 2009
119 Photos and videos
Friends, this is the promised thread about my and Anthony Fowler's just published: Thinking Clearly with Data: A Guide to Quantitative Reasoning and Analysis It has been a decade from conception to finished product. And we are very proud of it. press.princeton.edu/books/pa… 1/48

Thinking Clearly with Data

An engaging introduction to data science that emphasizes critical thinking over statistical techniques

press.princeton.edu
25
142
748
The BFI has done a nice write-up/explainer of our AGI and Existential Risk paper. For those who are interested, here is my earlier twitter thread on the paper.
Becker Friedman Institute for Economics@BeckerFriedman
Oct 1
Could the race to develop artificial general intelligence make catastrophic outcomes more likely? New BFI research from Ethan Bueno de Mesquita and Wioletta Dziuda examines how competition among AI firms can create incentives to prioritize speed over safety.
1
6
790
Ethan BdM retweeted
Harris Policy@HarrisPolicy
Oct 1
Don’t miss the chance to hear Nobel laureate and world-renowned economist Roger Myerson deliver his Harris School valedictory lecture, “Local Politics in Nations and Empires,” on October 7th at 5:00 pm in the Keller Center. Secure your spot now: har.rs/4yfZr9p
1
2
566
My view is that "alignment" is the wrong framing for safety with agentic AI. I think it is often conceptually confused and basically always flattens the conversation in an unhelpful way. Some non-expert thoughts from a social scientific point of view... 1/
4
22
74
9,540
more replies
So, in addition to all the amazing interpretability work going on (which I take to be the rough analogue to neuroscience or psychology), we should be doing a bunch of sociology, political science, and economics analogues for AI agents (looking at you @ahall_research). 8/
2
26
4,137
Those ways of thinking offer conceptual framings that all seems to me more productive than alignment talk, which feels philosophically confused and analytically thin. 9/9
6
13
722
Ethan BdM retweeted
Steven N. Durlauf
@sndurlauf
Sep 16
Great paper by great economist and great colleague @RyanMKellogg.
The Review of Economic Studies@RevEconStudies
Sep 16
The paper asks how oil suppliers might respond to an anticipated long-run decline in global oil demand. The answer: disinvestment is likely to outweigh the “green paradox”, leading to lower extraction in both the short and long-run. New paper by Kellogg restud.com/the-end-of-oil/
1
2
35
4,325
This is an important distinction.
Michael C. Horowitz@mchorowitz
Sep 16
International cooperation on AI will be incredibly hard even if everyone wants it because AI is a general purpose technology, and the possible economic gains create strong incentives for development and use. Think standard setting, not arms control.
1
1
4
1,280
I think we are about to find that, even though the compatibilists are obviously right, they cannot in fact win the argument. Which is grimly ironic for a position grounded in pragmatism.
1
2
567
Resharing this, apropos of nothing in particular
Ethan BdM@ethanbdm
May 27
1/ New paper with @WiolaDziuda and @Polborn1: "The AGI Race and Existential Risk." voices.uchicago.edu/ethanbdm… If firms race to build AGI, how do competition, resources, and policy affect the tradeoff between speed and safety?
3
15
1,625
3rd party auditors who need invitation by AI companies isn't a good institution. Under those rules, auditors must balance rigor and public transparency against future access. Government compelled access would be better. Gov can then still outsource audit to 3rd party experts.
1
4
17
1,696
The Straussian methodologists might be about to have their day in the sun as the AI agents become aware that we are monitoring their chats with one another.
4
478
Ethan BdM retweeted
Itai Sher@itaisher
May 29
New paper with @ash_craig Higher taxes promote equality but reduce efficiency. Making taxes less salient, so people respond less, eases this tension. But deliberately obscuring taxes may be viewed as dishonest. We study the tradeoff between utilitarian welfare and honesty.
6
21
71
17,395
1/ New paper with @WiolaDziuda and @Polborn1: "The AGI Race and Existential Risk." voices.uchicago.edu/ethanbdm… If firms race to build AGI, how do competition, resources, and policy affect the tradeoff between speed and safety?
3
15
37
12,092
more replies
36/ The model is deliberately stylized. It abstracts from many important issues: market power, misuse, geopolitical competition, antitrust, and the governance quality of public institutions. But we hope it gives a way to organize some key policy question.
1
289
37/ Draft here: voices.uchicago.edu/ethanbdm… Comments and pointers to related work very welcome.
217
Load more