Schedule

Introduction

Session 1: resource rationality

Mark Ho: Meta-reasoning about problem spaces

Abstract: One of the most enduring insights from classic work on human problem solving is that people solve problems by heuristically searching through a problem space. However, computational cognitive psychologists have traditionally sidestepped a critical question: Where do problem spaces come from? This issue, sometimes called the "problem with problem spaces", poses serious challenges to any comprehensive account of human intelligence. I will discuss recent and ongoing work that tackles this question using contemporary ideas from meta-reasoning and resource rationality. While this approach has enabled us to make progress on long-standing questions, we are still far from a complete solution to the problem with problem spaces.

Discussion

Session 2: learning and development

Junyi Chu: Playful problem selection

Abstract: What do you do when you can do anything you want? Studying how people play offers a unique window into problem selection by exposing both the diverse objectives people care about and how people evaluate (and sometimes invent) problems with respect to these objectives. I'll present some recent work and open challenges with a focus on age- and experience-related changes.

Discussion

Session 3: social context

Natalia Vélez: Understanding structural diversity in human collaboration

Abstract: Humans form an incredible variety of teams to achieve goals that are beyond the reach of a single person. Some teams are big, while others are small; some have leaders, while others are decentralized; some are embedded in formal institutions, while others are informal; some are transient, while others are enduring. This wide range of team structures poses a challenge to psychological theories of collaboration, which have shed light on the individual cognitive capacities that enable collaboration but have been largely silent on how these micro-scale processes give rise to the macro-scale structure of teams. When are particular team structures needed to solve particular problems? As a step towards addressing this gap, I will present a proof of concept using One Hour One Life, a multiplayer online game where players can build technologically advanced settlements from scratch (N = 22,011 players, 428,255 characters played, 1,486 settlements comprising 3–8,095 characters). This dataset offers a rare opportunity to watch communities self-organize to navigate basic problems of survival. I will present evidence that these virtual communities replicate scaling relationships that have been documented in real-world cities, and that player behavior systematically changes with community size. These findings demonstrate that online games provide a powerful tool to understand basic principles of social organization. More broadly, they lay the groundwork for using gamified experiments to test when and how particular group structures emerge, and how they shape the outcomes of collaborations.

Discussion

Session 4: agency and identity

Laurie Paul: Modeling choices that change us

Abstract: I will define and explore a conceptual paradox that underlies the difficulty we can have in simulating or using theory of mind to model ourselves through radical changes or over long timescales.

Discussion

Closing