There is a fundamental puzzle about human intelligence: How are we incredibly smart and stupid at the same time? Humans deal successfully with the world in a way that no machine can (for now), yet we routinely behave in ways that seem grossly inconsistent with normative canons of rational inference and rational choice. This course will seek to resolve the paradox by exploring the idea that while we make many mistakes, these mistakes are not haphazard; instead, they reflect a brain that is highly efficient at inference and decision making within the information, time, and energy constraints imposed by the finite resources available to it. In other words, our brains may be “resource-rational” even if they fail to conform to ideal canons of rationality. We will explore this idea by considering the structure of errors, biases and illusions in the context of perceptual judgments, more abstract cognitive judgments (perceptions of numerical magnitudes or probabilities), and economic decisions; we will see that there are many analogies between the kinds of characteristic errors that people make in all of these contexts. A potential explanatory framework, which can be applied across contexts, considers what optimal decisions should be like in the case of a decision unit that has only imprecise information about its situation. Hence statistical modeling and statistical inference are key elements in the computational models of human decision making that we wish to discuss
- Subject
- COGS GU
- Credits (min)
- 4
- Credits (max)
- 4
- Credit unit
- points
- Type
- course
- Edition
- 2026-2027
- Catalog
- 2026-2027 Bulletin
- Source
- bulletin.columbia.edu