BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//UM//UM*Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/Detroit
TZURL:http://tzurl.org/zoneinfo/America/Detroit
X-LIC-LOCATION:America/Detroit
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20070311T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20071104T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260526T134332
DTSTART;TZID=America/Detroit:20260603T100000
DTEND;TZID=America/Detroit:20260603T120000
SUMMARY:Other:Dissertation Defense: The Pursuit of Value: Essays on the Cognitive Science of Motivation
DESCRIPTION:ABSTRACT:\n\nHumans have motivations to do things. The same is true for animals\, and perhaps for artificial agents as well. How does this work? This dissertation is an attempt to answer that question\, and explore the normative and descriptive consequences of the answer. To do so\, I draw on formal and computational models from philosophy\, economics\, and artificial intelligence research\, and empirical evidence from cognitive neuroscience and psychology.\n\n\nAt the heart of the dissertation is a new view of motivation that I call externalism about expected value maximization\, or externalism for short. Externalism agrees with all standard accounts of rationality that a rational agent performs actions that maximize the achievement of some metric of value. However\, it relaxes an additional assumption that is typically made implicitly\, namely that what has value is a function of the agent themselves\, for example their preferences or desires. According to externalism\, rational agents do not determine the value of an option any more than they determine its weight. Instead\, they represent value as they would other quantities: as existing independent of themselves and able to be estimated in light of evidence. Unlike estimations of other quantities\, however\, the estimations of value govern their behavior: motivation is the direct pursuit of expected value.\n\nI present the idea of externalism in a new introductory essay written for the dissertation\, titled ‘Rationality Externalized’. The next three papers elaborate on the externalist picture\, although implicitly so. (1) In Chapter 2\, titled ‘What Stands to Desire as Perception Stands to Belief?’\, I consider how our motivational attitudes can be updated from experience\, and whether any mental capacity allows for this in the way that perception does for belief. I propose a novel functional capacity that I call conaception\, and argue that it provides signals that update representations of value and thereby change our desires. I also suggest that evolution has plausibly tuned our minds to interpret these signals so as to steer us towards what is good for us in a normatively substantive sense.\n\nIn Chapter 3\, ‘Reward is Evidence of Value’\, I develop this idea in more formal and empirical detail\, suggesting that reward signals should be seen as one crucial source of evidence of value. This resolves an ongoing debate about how to interpret computational models in Reinforcement Learning as applying to humans and other animals. In Appendix A\, I demonstrate the formal details of this account\, and show how standard RL algorithms converge on Bayesian solutions even assuming an evidential interpretation of reward. In Chapter 4\, ‘A Conflation of Valences’\, I consider the relation between reward learning and hedonic valence. A common view is to identify hedonic experience as the biological instantiation of reward signals. I argue that this is empirically implausible\, and articulate the implications of this separation for the distribution of sentience and other philosophical questions.\n\nChapter 5 and 6 deal with two quite different but related issues. In Chapter 5\, ‘Do Expected Utility Maximizers Have Commitment Issues?’\, I consider a long-standing criticism of standard expected utility theory\, namely that it implausibly recommends us to tie ourselves to the proverbial mast whenever we expect our future preferences to change in a way that is contrary to our current goals\, even at great cost. Drawing on Bayesian reputation games\, I propose a novel solution by arguing that expected value maximizers typically have an incentive to follow-through on prior commitments after all\, because by doing so we foster a reputation with ourselves that permits rationally pursuing better outcomes in the future. In Appendix B\, I prove the main result of this chapter.\n\nFinally\, in Chapter 6\, ‘The AI Epistemic Deference Index: A Continuous Measure of Sycophancy’ (co-authored with Alejandro Botas and Luke Hewitt)\, I approach a contemporary practical problem related to human motivation: it is well-known that current artificial agents trained on human feedback provide excessively sycophantic interactions. This provides inaccurate evidence for users\, shaping both our beliefs and values in detrimental ways. We present new techniques for providing a scalar metric to assess the sycophantic behavior of current models\, and demonstrate substantial differences between them.\n\n\n(1) These papers were written prior to the introduction and do not reference externalism as such\, but can naturally be interpreted as contributing to that overarching program.\n\nThose interested in reading the full dissertation beforehand can contact the candidate to request a draft.\n\nCOMMITTEE:\nSripada\, Chandra  (chair)\nRailton\, Peter\nJoyce\, Jim\nLewis\, Rick (cognate\, Cognitive Science)\nChalmers\, Dave (special member\, NYU)
UID:148420-21904230@events.umich.edu
URL:https://events.umich.edu/event/148420
CLASS:PUBLIC
STATUS:CONFIRMED
CATEGORIES:Dissertation
LOCATION:Angell Hall - 2271
CONTACT:
END:VEVENT
END:VCALENDAR