Lectures & Colloquia
Friday, March 17
Melissa Fusco, Columbia University
Baker Hall A36
Abstract: In recent work, Arif Ahmed and Adam Elga present a dilemma for Causal Decision Theory (CDT) which features deterministic laws. My purpose here is to respond to that challenge on behalf of CDT. I focus on Elga’s paper, “Confessions of a Causal Decision Theorist”, which features a formal proof, and I aim for absolution. The treatment I present highlights (i) the status of laws as predictors and (ii) the consequences of decision dependence (Gibbard & Harper, 1978; Skyrms, 1990).
Friday, March 31
Hilary Greaves, University of Oxford
Talk Title: Against minimal morality
Abstract: Many ethicists subscribe to both of the following conditions. (1) Moral requirements are minimal: the standard of conduct that is required by morality is rather undemanding. (2) The deontic screens off the moral: Roughly, the thesis that once the requirements of morality are settled, there is no further normative relevance to moral considerations. I will argue that these claims are not simultaneously tenable. Moral requirements might be minimal, but if they are, then moral requirements do not exhaust the ways in which morality is relevant to what we ought to do. This conclusion has practical importance, inter alia, in the context of global poverty, existential risk and nonhuman animal welfare.
Please contact Adam Bjorndahl if you have any questions or concerns.
Friday, April 21
Silvia Milano, University of Exeter
Baker Hall A36
Talk Title: Rationality without commitment
Abstract: In this talk, I will argue that it is sometimes impossible to commit to a course of action, for reasons that are both wholly anticipated and rational. This solves a well-known issue for games of imperfect recall that notoriously give rise to a tension between optimal planning and execution.
My argument will revolve around a new approach to the absentminded driver problem put forward by Milano and Perea (2023), which employs centred possibilities to model the driver’s uncertainty. In this scenario, act-state independence is violated, but states are mutually exclusive and the agent arrives at their optimal choice probability via Bayesian updating. This solution is the only one guaranteeing immunity from sure loss, and, I will argue despite initial appearances, is time consistent.
Friday, September 16
Professor Vincent Conitzer, Carnegie Mellon University
Posner Hall 151
Talk Title: (How Philosophers Can and Do Contribute to) Foundations of Cooperative AI
Abstract: I direct the new Foundations of Cooperative AI Lab (FOCAL) at CMU, the goal of which is to create foundations of game theory appropriate for advanced, autonomous AI agents -- with a focus on achieving cooperation. Too little attention has been paid to the ways in which AI agents can be fundamentally unlike human agents, and this has left various ways of avoiding game-theoretic tragedies (prisoner's dilemma, tragedy of the commons, ...) underexplored. But many relevant insights can be found in the philosophy literature, for example on self-locating beliefs (Sleeping Beauty problem) and decision theory (causal vs. evidential). In addition to formal epistemology, eventually this connects even to problems in metaphysics and philosophy of mind. A very recent and surely imperfect writeup can be found at: https://www.cs.cmu.edu/~15784/focal_paper.pdf; feedback is appreciated!
Friday, October 28
Wayne Wu, Carnegie Mellon University
Posner Hall 151
Talk Title: Attention, Norms and Bias
Abstract: This talk will be an overview of the intersection of attention and normative requirements in recent philosophical work. I will briefly discuss some relevant aspects of the science of attention as background, explicating the concepts of bias, automaticity and control in a precise way. I then discuss three concepts that have normative significance, namely in recent work on ethics and epistemology: salience, vigilance and attentional character. I conclude with remarks regarding a concrete case of epistemic bias in academia and assess our methods for responding to it. I also hope to provide an example of how cognitive science and philosophy can productively engage.