Susan Athey: Tech Economists, Machine Learning, and Causation

AI Assistance

Machine learning is transforming decision-making processes in tech firms by augmenting human roles rather than replacing them. Susan Athey highlights that while AI can automate routine tasks, it often serves best as a tool to assist humans in complex decision-making scenarios 1. For example, AI can prioritize tasks by analyzing vast amounts of data, allowing humans to focus on high-value decisions. Athey explains, "There's a whole economics literature...where salespeople, gaming commissions is exhibit A of the economics literature" 2. This approach ensures that human judgment remains central, especially in nuanced situations where AI alone may not suffice.

   

Correlation vs Causation

Understanding the difference between correlation and causation is crucial when applying machine learning insights in business contexts. Susan Athey emphasizes the importance of running experiments to discern true causal relationships, as relying solely on prediction models can lead to misleading conclusions 3. She notes that machine learning can identify patterns, but without proper experimentation, these patterns might not reflect causation. Athey points out, "One of the cool things with machine learning...is that we can more quickly see interesting things in the world because we can run lots of experiments" 4. This highlights the need for rigorous testing to ensure accurate and actionable insights.