Causal Bandits Podcast Free Will, LLMs & Intelligence | Judea Pearl Ep 21 | CausalBanditsPodcast.com
5 snips
Aug 12, 2024 In this conversation, Judea Pearl, the godfather of modern causal inference, dives into the complexities of causal reasoning and artificial intelligence. He discusses the influence of human biases on AI's decision-making and the potential future of large language models in scientific experimentation. Pearl emphasizes the importance of bridging causal reasoning with statistical education, using malaria as a case study. He also shares insights on the challenges of implementing evidence-based frameworks in education, advocating for a unified understanding of causality across disciplines.
AI Snips
Chapters
Transcript
Episode notes
Smallpox Paradox
- Judea Pearl's high school teacher presented a smallpox vaccination paradox.
- This sparked his lifelong interest in causality and the need for a formal language to express "died from."
Causal Bayesian Networks
- Judea Pearl realized Bayesian networks' success stemmed from their causal structure.
- This led him to explore beyond probabilistic reasoning in the 1980s.
Deterministic Counterfactuals
- In 1991, Pearl and Verma formalized counterfactuals using deterministic functions.
- This shift from probabilistic reasoning was initially challenging for them.