Invited Speakers

Aviad Heifetz (Open University of Israel)

Rationality amounts to reason when others are ends, not means

Abstract: Kant's categorical imperative "Act only according to that maxim whereby you can at the same time will that it should become a universal law" is not compatible with the dominant strategy choice "defect" in the prisoner's dilemma, because "cooperate" is the act that each player would like to become the universal law. 

For Kant, the categorical imperative follows from Reason (Vernunft). The prisoner's dilemma therefore demonstrates that Reason is a more demanding concept than mere Rationality: one can be rational and yet unreasonable.  In other words, adhering to the characterization of an individual by the preferences over outcomes that she strives to maximize, not all rational (i.e. complete and transitive) preference relations characterize reasonable individuals.

Kant's second formulation of the categorical imperative is "Act in such a way that you treat humanity, whether in your own person or in the person of any other, never merely as a means to an end, but always at the same time as an end." In other words, when you act so as to maximize your own wellbeing, care about others' wellbeing (i.e. consider them as ends) rather than about their choices (as means that are conductive, more or less so, to your own wellbeing).

We formally prove that this second formulation of the categorical imperative indeed implies the first: in purely affective interactions, where each individual's utility is a function of her choice and others' utilities (rather than of others' choices, as in standard game theory), the equilibrium choice profiles are both efficient and locally dominant - whether choices are simultaneous or sequential. 

Reasonable preferences thus form a knife-edge subset of the rational preference relations usually considered in game theory. Yet, these are precisely the innate preference relations developed by well-held children according to prominent views in developmental psychology (e.g. of Donald Winnicott). At the same time, and unlike in economics and game theory, a long philosophical tradition views reasonable preferences not as an immutable characterization of the individual, but rather as a continuous achievement of negative work, of unchaining (as in Plato's allegory of the cave) our attachments to means that provide only superficial gratifications. 

Short Bio: Professor Aviad Heifetz of the Open University of Israel is a mathematician specializing in game theory. In recent years his research focuses on behavioral ecology and on political theory.

Willemien Kets (Utrecht University)

Incorporating Culture into Game Theory

Abstract: Game theory is commonly used to model human behavior. But while there is growing empirical evidence that culture influences behavior in games, there are still few formal models that can capture this. This talk discusses how insights from psychology and sociology can be incorporated into game-theoretic models to account for the impact of culture. The key idea is that culture can be viewed as shared cognition: what people pay attention to, how they respond to contextual cues, and what is salient to them. Rationality cannot always override such culturally-driven impulses: Because players understand others are also influenced by their sociocultural environment, this will affect how they reason about others. Building on research in psychology on theory of mind, I develop a new model of how culture influences perspective-taking and behavior in games. The model can help better understand when having a strong culture is beneficial, and when it is harmful.

Short Bio:  Willemien Kets is a professor of game theory at the Mathematical Institute at Utrecht University (Netherlands). Her work aims to develop better models of strategic behavior. For example, her recent work develops a new mathematical formalism to model the effects of culture, and earlier work uses epistemic game theory to model bounded rationality. Before joining Utrecht University in 2022, Kets was an Associate Professor at the University of Oxford and an Assistant Professor at Northwestern University. Her work aims to develop better models of strategic behavior. For example, her recent work develops a new mathematical formalism to model the effects of culture, and earlier work uses epistemic game theory to model bounded rationality. Before joining Utrecht University in 2022, Kets was an Associate Professor at the University of Oxford and an Assistant Professor at Northwestern University

Jon Kleinberg (Cornell University)

Modeling Conflict in Social Media 

Abstract:  Social media platforms provide a means for people to connect with information and with each other, and active lines of research have studied both the opportunities for engagement and the risks of conflict that arise on these platforms. We consider models for two forms of conflict in social media. One consists of conflict between people, and leads to phenomena including polarization. A second is sometimes harder to perceive, and consists of conflicts that take place within a single person, as an individual social media user is guided by preferences that may be internally inconsistent. We consider the implications of these models for the design of online platforms, in settings where even large amounts of observed data may not fully represent the internal preferences and goals of the users. 

This talk is based on joint work with Jason Gaitonde, Sendhil Mullainathan, Manish Raghavan, and Eva Tardos. 

Short Bio:  Jon Kleinberg is the Tisch University Professor in the Departments of Computer Science and Information Science at Cornell University. His research focuses on the interaction of algorithms and networks, the roles they play in large-scale social and information systems, and their broader societal implications. He is a member of the US National Academy of Sciences and National Academy of Engineering, and serves on the US National AI Advisory Committee. He has received MacArthur, Packard, Simons, Sloan, and Vannevar Bush research fellowships, as well awards including the Harvey Prize, the Nevanlinna Prize, the Allen Newell Award, and the ACM Prize in Computing. 

Anna Mahtani (London School of Economics)

Contextualism and Awareness Growth

Abstract: There has been recent interest in the phenomenon of 'awareness growth' - where an agent starts off unaware of a proposition and so assigns it no credence, and then becomes aware of it and assigns it a credence, redistributing credences in other propositions accordingly. This phenomenon creates a problem for Bayesian epistemologists who standardly claim that a rational agent's credences change only by conditionalization. I show that we can handle these cases if we adopt a contextualist account of credences. I describe this contextualist account of credences, and show how cases of apparent awareness growth can be understood in light of this account. I argue that this approach dissolves a number of problems that philosophers working on awareness growth have faced.

Short Bio: Anna Mahtani is Associate Professor in philosophy at the London School of Economics. She did her PhD on vagueness at Sheffield, and then worked at Oxford and the Open University, before arriving at the LSE. She studies decision theory, formal epistemology, and the philosophy of language, and works at the intersection of these different disciplines. She is currently working on several projects: tracing the implications of Frege's puzzle for various principles of welfare economics; analysing the phenomenon of 'awareness growth'; and writing a book called 'The Objects of Credence'.