Prof Gilles Chemla (@chemlag) profile picture

Professor of Finance, co-director of Ctr for Fin. Tech., Imperial Business School; visiting scholar, Harvard Uni; research director, CNRS; research fellow, CEPR

Joined March 2012
Photo by Prof Gilles Chemla (@chemlag)Photo by Prof Gilles Chemla (@chemlag)Photo by Prof Gilles Chemla (@chemlag)Photo by Prof Gilles Chemla (@chemlag)Photo by Prof Gilles Chemla (@chemlag)Photo by Prof Gilles Chemla (@chemlag)Photo by Prof Gilles Chemla (@chemlag)Photo by Prof Gilles Chemla (@chemlag)Photo by Prof Gilles Chemla (@chemlag)
Congratulations to Karine Chemla, @uniofedinburgh, @EdinUniMaths, who will receive the 2027 Albert Leon Whiteman Memorial Prize for her outstanding contributions to the history of mathematics in Ancient China and 19th century Europe. Read more. Link in comments.
1
4
9
1,085
Quel que soit le parti pour lequel vous voterez en 2027, si vous voulez offrir un avenir aux entreprises tech en France et en Europe, écoutez le fondateur de Mistral ou Total Énergies et non des politiciens en seul quête de pouvoir. Ne tuons pas l'UE, faisons-la accélérer ! 🇪🇺🚀
43
248
17
1,110
44,639
Prof Gilles Chemla retweeted
This 2 hour Stanford lecture will teach you more about how LLMs like ChatGPT & Claude are built than most people working at top AI companies learn in their entire careers. Bookmark this & give 2 hours today, no matter what. It'll be the most productive thing you do this week.
143
3,217
71
18,944
2,366,898
Prof Gilles Chemla retweeted
New essay on the economics of structural change and the post-commodity future of work. 1. Almost any question about the impact of advanced AI on the economy needs to start at the same place: what is still scarce? Answer that, and the analysis becomes pretty straightforward. This essay explores what becomes scarce if AI really can replicate most of what humans do in production, and what this mean for the future of jobs. 2. My conjecture, working through the economics: labor reallocates across sectors, and the sector it reallocates to has properties that keep labor a meaningful share of the economy. Ultimately this is about the structure of demand itself. For this, we have to go back to Girard, Augustine and Rousseau: once people's base needs are met, their preferences shift to comparative motives (e.g., status, exclusivity, social desirability). This motive is inherently non-satiated. 4. The key paper is Comin, Lashkari, and Mestieri (Econometrica 2021). As people get richer, they don't buy proportionally more of everything. They shift spending toward sectors with higher income elasticity. They estimate income effects account for 75%+ of observed structural change. 5. The ironic consequence: the sector that gets automated becomes a smaller share of the economy, not a larger one. Agriculture got massively more productive and its share of employment collapsed. Manufacturing too. The "stagnant" sectors absorb the spending and the jobs. 6. So the question is: which sectors have high income elasticity in a post-AGI world? I argue it's what I call the relational sector. Categories where the human isn't just an input into production, it is part of the value. 7. Why does the relational sector have high income elasticity? Because human desire has a mimetic, relational dimension. We don't just want things for their intrinsic properties. We want what others want, and we want it more when others can't have it. Girard, Rousseau, Augustine, and Hobbes all saw this. 8. In work with Kristóf Madarász, we showed this experimentally: WTP roughly doubles when a random subset of others is excluded from the good. And in new work with Graelin Mandel, AI involvement kills the premium. Human-made art gains 44% from exclusivity; AI-made art only 21%. 9. This all comes together for the core argument. The sector that absorbs spending as AI makes commodity production cheap is one where human provenance is part of the value, and demand for it grows faster than income. Exactly the profile that keeps labor meaningful. 10. To be clear about the claim: I'm NOT saying aggregate labor share must rise. It may fall. The claim is about sectoral composition, i.e., where expenditure and employment go once commodities get cheap, and the fact that the sector that will absorb reallocated labor maps to a substantial component of human preferences and desire. 11. If you're interested in the formal model, a linked companion technical note works out all the economics. Read the essay here: aleximas.substack.com/p/what…
169
506
167
2,933
1,086,393
This looks like a must-read! "The Economics of Tariffs" by Ralph Ossa and Stephen J. Redding. "A central insight from neoclassical economics is that international trade operates like an improvement in production technology. It generates mutual aggregate welfare gains for countries as a whole, but creates winners and losers within countries. Tariffs are a tax on this trading technology and distort the prices faced by domestic consumers and producers. Large countries can use tariffs to improve their terms of trade on world markets. But if all countries try to do so, they can end up with lower welfare than if they cooperated to liberalize trade. ...Empirical findings from the recent waves of U.S. tariffs suggest that most of the incidence of these tariffs has been borne by U.S. importers, wholesalers, retailers and consumers rather than by foreign exporters. These tariffs have led to a large-scale reorganization of U.S. supply chains away from China to third countries. Although this reorganization has substantially reduced China’s share of U.S. imports, the U.S. remains indirectly exposed to China through the imports of these third countries." econ.uzh.ch/dam/jcr:8e21c947…
5
77
2
377
29,009
Prof Gilles Chemla retweeted
Between 1974 and 2014, only 0.1% of publications in the top 50 economics journals were replication studies. That's 40 years. Thousands of papers. Almost none replicated. We've built careers on findings no one has verified. When economists finally replicate studies, 40-67% fail depending on the study. Federal Reserve (2015): Only 49% of 67 papers from top journals successfully replicated — even with the original authors' help. Most papers? Never checked at all. Here's the worst part: Papers that don't replicate get cited MORE than papers that do. And after a failed replication is published, only 12% of subsequent citations mention it. The profession rewards interesting findings, not true ones. Remember Reinhart-Rogoff (2010)? "Debt above 90% GDP kills growth." Herndon, Ash & Pollin found a spreadsheet error. Results didn't hold. But by then, it had shaped austerity policy across Europe and the US. How many other canonical papers have Excel errors? We replicate recent papers, but canonical findings from the 1970s-90s? Nobody touches them. Too famous. Too foundational. The older the paper, the less scrutiny it gets. Yet these are the studies we cite most. Maybe we should replicate backwards. Start with the most-cited papers from 1975-2000. See what holds up. federalreserve.gov/econres/f…
78
708
120
2,429
251,890
Prof Gilles Chemla retweeted
#AI Adoption, Incentives, and Firm Value in General Equilibrium The authors study how AI adoption reshapes incentive contracts, employment relationships, and firm values. Authors: Gilles Chemla, Vincent Tena Read More: spkl.io/6012Am386 #AI
1
1
1
344
Over the years, I have actively participated in the economic policy debate in Spanish and written a fair amount on the topic (including many papers and several books!). Some of these issues go beyond purely local interest and are relevant, with only minor changes, for other countries in Western Europe and Latin America. I have always wanted to translate some of this work into English (if only to reach a wider audience), but until recently the cost was prohibitive. Thanks to AI, that cost has now fallen by roughly two orders of magnitude. Yesterday, I circulated a short policy paper I wrote over the Winter Break in Spanish arguing that it is far more useful to think about the sustainability of pay-as-you-go social security systems in terms of their internal rate of return (IRR) than to focus on cash flows, as much of the debate tends to do. The paper went viral, and several non-Spanish economists expressed interest. This morning, I translated it into English with the help of AI and made some light adaptations. I did not spend a great deal of time polishing every detail, but I think it is still well worth reading: sas.upenn.edu/~jesusfv/Pensi… The punchline will not be new to anyone who has read Samuelson’s (1958) classic paper, but it bears repeating. The key to the sustainability of a pay-as-you-go social security system is that the IRR received by participants must equal the growth rate of the system’s revenues. Once you think in these terms, almost everything else becomes secondary. Lower fertility simply implies a lower growth rate of revenues. Along the same lines, higher wages do not necessarily help the system if the IRR is already too high. In fact, they can make things worse, because an excessive IRR is then paid over a larger base. Somewhat surprisingly, when I explained this point to several excellent economists, many told me they had never thought about it this way. What actually helps a pay-as-you-go system is faster wage growth, because it raises the growth rate of revenues. It is all about the derivative, not the level. The effects of immigration are also much more nuanced than is often assumed. If the IRR is too high, bringing in additional immigrants may worsen the system’s present discounted balance, even if it improves the current cash-flow position. As always, comments are welcome. I may write something longer in the future, and I would welcome any suggestions. P.S. I write differently in Spanish than in English. Do not be surprised if the paper reads somewhat differently from my work originally written in English. Personally, I think I write much better in Spanish, but that style does not always translate cleanly.
10
53
10
307
84,943
Prof Gilles Chemla retweeted
Scientists Who Use AI Become Stunningly More Productive In a new paper that just appeared, the authors found that researchers who use large language models (LLMs) become stunningly more productive. An analysis of more than 2 million papers revealed that those who adopt AI increase their paper output on average by 40%. For non-native English speakers it’s even more, up to 80%. The scary part is that this is data only until July 2024, when scientists used LLMs mostly for writing. It is foreseeable that in the next few years the AI-adoption rate will go to nearly 100% across all scientific disciplines, resulting in more papers than ever and putting significant strain on peer review. Source: Kusumegi et al, Science 390, 6779, 1240 (2025)
334
340
107
1,840
399,623
Our upcoming #CPD programme includes speakers from @Arup @JacobsConnects @HS2ltd @MottMacDonald @Laing_ORourke. Join us to learn from major programmes around the world and set the direction of future research. Early bird registration closes 5 August: imperial.ac.uk/continuing-pr…
3
1
360
Prof Gilles Chemla retweeted
C’est tellement la vidéo du siècle : une brochette de catalan de Seine-et-Marne humilié par un parisien dans un bar ahahahahah
128
3,030
284
16,252
1,235,941
Prof Gilles Chemla retweeted
The French government makes unemployed people go on blockchain awareness courses run by Binance, where if they gain sufficient blockchain awareness they're given an NFT diploma. Really. This is a thing that actually happens. on.ft.com/3uL1dmU
44
250
200
792
359,619
When you discover a smarter way to trade. @RipplexDev #BuiltOnXRPL #ISR #XRPL #XRPLedger
2
5
166
Prof Gilles Chemla retweeted
To no one's surprise, the Nobel Prize goes to Claudia Goldin Here are 3 things from her to start your dive into her work:
51
2,287
253
9,239
1,965,900
Prof Gilles Chemla retweeted
The work that won today's Nobel in medicine (and saved millions of lives) "was summarily rejected by the journals Nature and Science". Keep trying. Journals are often resistant to new ideas.
68
1,272
134
5,741
898,896
Prof Gilles Chemla retweeted
Lobbying in Brussels. “We receive funding from multiple sources so we can provide a balanced view and we are not captured by any single one interest.” Yeap. @lobbyctrl_tech
11
47
15
160
39,288
Jamie is looking for an RA to work on topics in financial intermediation and me I am looking for an RA to work on topics at the intersection of public, household and corporate finance, with a particular focus on inequality. If interested, please apply! imperial.ac.uk/jobs/descript…
1
2
8
2,489
The Brevan Howard Centre are hosting the UKFin+ London Launch on 5 July 🗓️ @UKFinNetwork are funding research collaborations for wicked problems in #FinancialServices. Speakers will include @LedaGlyptis @chemlag @AadVanMoorsel & more! Get your ticket ⬇️ imprl.biz/46pAvyz
4
7
737
Prof Gilles Chemla retweeted
I heard that Ed Prescott died. I did not agree with him about macro but greatly admired his iconoclastic approach. I do not go a week without being influenced by his dynamic consistency theorizing. Rest In Peace.
5
31
4
284
Why does the law protect certain groups, but not others? And what happens if AI creates groups that do not map on to traditionally protected groups? My work The Theory of Artificial Immutability: Protecting Algorithmic Groups under Anti-Discrimination Law papers.ssrn.com/sol3/papers.…
3
30
6
120