ICM 2026 Invited Section Lecture - Michal Feldman

Value of the Information & Strength of the Argument

The lecture provides a comprehensive and up-to-date overview of algorithmic contract design, emphasizing computational complexity and approximation. The argumentation is rigorous, proceeding from a clear model to theorems and proof sketches. Feldman connects the area to classical contract theory, algorithmic mechanism design, and combinatorial markets, giving it strong intellectual context. He clearly distinguishes algorithmic tractability from impossibility, and highlights the importance of oracle models. The talk is interactive, though the core value lies in the precise statements and relationships between function classes.

Scientific Rigor, Source Quality, Title Accuracy

The scientific rigor is evident: results are attributed to specific papers (e.g., the 2006 combinatorial agency paper, works by Dutting, Roughgarden, Talgam-Cohen, and the speaker’s own collaborations). References to Gul and Stacchetti for gross substitutes and to Baldwin and Klemperer for demand types demonstrate command of the literature. The title accurately reflects the content; no discrepancies were noted. As an invited lecture at a major mathematics congress, the source is authoritative. Public comments are not available for analysis.

The video title accurately describes the content: an invited lecture at ICM 2026 by Michal Feldman. The specific topic (combinatorial contract design) is provided in the description.

9/10

The lecture is given by an expert researcher at an ICM invited section, based on published papers and rigorous theoretical work. The reasoning is formal and well-supported.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 00:00Introduction to contract design and moral hazard.
• 03:00Personal story: PhD at Berkeley, combinatorial agency paper from 2006.
• 06:00Illustrative example with Nike and influencer commissions.
• 10:00Formal model: single agent, actions, reward function F, contract alpha.
• 15:00Oracle models (value and response) and hierarchy of reward functions.
• 20:00Critical values and upper envelope from agent's perspective.
• 25:00Main theorem: optimal contract for gross substitutes with value oracle, proof sketch.
• 35:00Extensions to supermodular, ultra, and ASC classes; open frontier.
• 45:00Multi-agent model: contracts as vectors, inducing a game and Nash equilibrium.
• 55:00Concluding remarks and open questions.

• Combinatorial agency (2006) - Babaioff, Feldman, Nisan — Referenced as the starting point for the speaker's research on incentivizing teams with combinatorial interactions.
• Simple versus optimal contracts (Dutting, Roughgarden, Talgam-Cohen) — Cited as important recent work that revived interest in algorithmic contract design around 2019-2020.
• Gul and Stacchetti (1999) - on gross substitutes — Used to compute agent best response via greedy algorithm, a key step in the proof for gross substitutes.
• Baldwin and Klemperer - Demand types — Their geometric characterization of set functions is used to extend tractability to larger classes like ASC.
• Sun and Yang - Gross substitutes and complements — Their class is shown to be included in the ASC class for which polynomial critical values exist.

• ICM 2026 official program — The lecture is part of the official ICM program, confirming the authoritative context.