Post

How to teach human conflict to AI? I'm sitting on both "computational game theory" in the CS dept and "commercial negotiation and mediation" in the law school this term, because neither tradition has a complete answer.

Game theory assumes preferences are static, whereas good mediation works by actively shifting what people think they want. Formalizing that drift seems like where the math breaks down.

I agree! Do you know of a reference which discusses this?

Yes. Two useful starting points: Bush & Folger's *The Promise of Mediation* (1994), and Henner Gimpel's *Loss Aversion and Reference-Dependent Preferences in Multi-Attribute Negotiations* (2007), which models preferences changing during negotiation.