Re-Imagining the Social Sciences in the Age of AI - March 4, 2020

Overview

Explore the intersection of artificial intelligence and social sciences in this thought-provoking conference talk featuring two expert speakers. Delve into the transformative potential and challenges arising from the dialogue between these fields. Learn about decision rules in mate selection as Elizabeth Bruch from the University of Michigan presents her research on online dating behavior. Discover how human behavior is becoming the next frontier for AI with insights from Thomas Griffiths of Princeton University. Engage in a moderated discussion and Q&A session led by Jacob Foster from UCLA. Gain valuable insights into topics such as decision-making processes, choice prediction, image classification benchmarks, and the application of AI in understanding human behavior. Examine various models and theories, including expected utility theory, prospect theory, and rational choice models, while exploring the potential of neural networks and complex data analysis in social science research.

Syllabus

Introduction
How People Make Decisions
How People Make Choices
Online Dating
The Model
Key Learning
Compensatory vs Non Compensatory Evaluation
Results
Summary
Cat or Dog
Image Classification Benchmark
Open AI
Human Behavior
Computer Science Psychology
Risky Choice Problem
Choice Prediction Competition
Data Collection
Evaluating Models
Expected Utility Theory
Prospect Theory
Validation Error
Literature
Estimates
Data
Domain
Data Set
Rational Choice Model
Utilitarian Model
Exploratory Data Analysis
Residuals
Scientific Regret minimization
Neural Network Residuals
Complex High Order Effects
Unexpected Interactions
Conclusions