Eric Olav Chen and I wrote a research agenda on Micro theory and AI while we were at GPI.
Working Papers
"Robust Regulation of Labour Contracts" (with Théo Durandard), Sept 25, Submitted
Extended Abstract in Proceedings of the 26th ACM Conference on Economics and Computation (EC'25)
[PDF] [Supplemental Appendices] [arXiv]
Which contracts should a regulator authorise in moral hazard problems to minimise their worst-case regret? Simply impose a minimum piece-rate compensation.
"AI Testing Should Account for Sophisticated Strategic Behaviour" (with Vojtěch Kovařík, Eric Olav Chen, Sami Petersen, Vincent Conitzer), Aug 25
Forthcoming in Proceedings of the 39th Conference on Neural Information Processing Systems (NeurIPS-25)
This position paper argues that AI testing needs to account for AIs' strategic reactions and that game theory should thus inform testing design.
"Grabbing the Forbidden Fruit: Restriction-Sensitive Choice" (with N. Boissonnet), June 25 (important revisions), Submitted
The forbidden fruit effect is when restrictions steer people’s desire toward the forbidden options. We axiomatise a choice model which rationalises it.
"Redistribution Through Market Segmentation" (with D. Barreto and V. Augias), Mar 25, new version soon
Extended Abstract in Proceedings of the 25th ACM Conference on Economics and Computation (EC'24)
We study how to segment monopolistic markets with a redistributive objective. These segmentations induce the seller to price progressively and may not maximise consumer surplus.
"Imperfect Recall and AI Delegation" (with Eric Olav Chen and Sami Petersen), Nov 24
[PDF] [arXiv]
Simulating real tasks in testing and deleting AI's memory allows an evaluator to both screen misaligned AIs during testing and discipline their behaviour in deployment.
"Note on the Indeterminacy of Deliberate Preference Changes" (with N. Boissonnet)
[PDF]
Publications in Economics
"Revealed Deliberate Preference Change" (with N. Boissonnet and S. Gleyze), Nov 2023, Games & Economic Behavior
[PDF] [Journal version] [Older version with additional results]
Agents' preferences stem from their 'values', which they can change when becoming aware of new ones. We axiomatise deliberate preference changes rationalised by a metapreference.
Other Publications
Solo climbing: a descriptive and normative approach (with Marion Pollaert and Jean-Baptiste Duez), in Sports: Theory vs. Praxis? (Ed. by Constantino Martins), 2021