The Misrepresentation Game: How to win at negotiation while seeming like a nice guy

摘要

Recently, interest has grown in agents that negotiate with people: to teach negotiation, to negotiate on behalf of people, and as a challenge problem to advance artificial social intelligence. Humans negotiate differently from algorithmic approaches to negotiation: people are not purely self-interested but place considerable weight on norms like fairness; people exchange information about their mental state and use this to judge the fairness of a social exchange; and people lie. Here, we focus on lying. We present an analysis of how people (or agents interacting with people) might optimally lie (maximally benefit themselves) while maintaining the illusion of fairness towards the other party. In doing so, we build on concepts from game theory and the preference-elicitation literature, but apply these to human, not rational, behavior. Our findings demonstrate clear benefits to lying and provide empirical support for a heuristic - the "fixed-pie lie" - that substantially enhances the efficiency of such deceptive algorithms. We conclude with implications and potential defenses against such manipulative techniques.