Week 4·Week 4 in the Syllabus →

Predicting Outcomes

Prediction Basics

• RequiredLLMs Are Closing the Gap on Human Superforecasters (3 min)Forecasting Research Institute (2026)
• RecommendedLies, Damned Lies, and Data Mining (19 min)Cliff Asness (2017)
• OptionalA Backtesting Protocol in the Era of Machine Learning (paper)Rob Arnott, Campbell Harvey, and Harry Markowitz (2019)

Predicting Fundamentals

• RecommendedWhy Man + Machine Adds Up to Better Stock Picks (4 min)Greg Muraski (2024)
• RecommendedHow AI Is Helping to Explain Stock-Price Moves (3 min)Matt Robinson (2026)
• OptionalFrom Man vs. Machine to Man + Machine: The Art and AI of Stock Analyses (paper)Sean Cao, Wei Jiang, Junbo Wang, and Baozhong Yang (2024)
• OptionalAssessing the Benefits of Optimized Agentic AI Systems for Asset Pricing (paper)Ralph Koijen and Bradford Levy (2026)

Predicting People

• RequiredCan AI Stand In for Human Survey-Takers? Not Really (10 min)Pew Research Center (2026)
• RecommendedSimulating Human Behavior with AI Agents (6 min)Stanford HAI (2025)
• OptionalLLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individuals (paper)Joon Sung Park and others (2024)
• OptionalA Foundation Model to Predict and Capture Human Cognition (paper)Marcel Binz, Eric Schulz, and others (2025)

Contamination and Cutoffs

• RecommendedGPT-4 and Professional Benchmarks: The Wrong Answer to the Wrong Question (8 min)Arvind Narayanan and Sayash Kapoor (2023)
• OptionalAssessing Look-Ahead Bias in Stock Return Predictions Generated by GPT Sentiment Analysis (paper)Paul Glasserman and Caden Lin (2024)
• OptionalCaution Ahead: Numerical Reasoning and Look-Ahead Bias in AI Models (paper)Bradford Levy (2026)

Confidence and Instability

• RecommendedAI Chatbots Remain Overconfident — Even When They’re Wrong (5 min)Jason Bittel (2025)
• RecommendedDefeating Nondeterminism in LLM Inference (Summary) (1 min)Simon Willison (2025)
• OptionalAssessing Consistency and Reproducibility in the Outputs of Large Language Models (paper)Julian Junyan Wang and Victor Xiaoqi Wang (2025)
• OptionalDefeating Nondeterminism in LLM Inference (19 min)Horace He and Thinking Machines Lab (2025)