Making Sense with Sam Harris #164 — Cause & Effect
Aug 5, 2019
In this engaging discussion, Judea Pearl, a UCLA professor and Turing Award winner, dives deep into the mathematics of causality and its implications for artificial intelligence. They explore the complexities of understanding causation versus correlation, emphasizing how AI needs to grasp these concepts for true intelligence. The conversation critiques historical views on causation and touches on provocative ideas about free will and consciousness, challenging listeners to rethink their understanding of decision-making and the nature of reality.
AI Snips
Chapters
Transcript
Episode notes
Science's Struggle with Causation
- Science, contrary to popular belief, has historically struggled to understand causation.
- People often confuse correlation with causation, hindering scientific progress.
Causation's Asymmetry
- Physics equations are symmetric, implying that X causes Y to the same extent Y causes X.
- Humans intuitively understand causality's asymmetry, but robots need explicit mathematical representation.
Rooster Crow and Sunrise
- A rooster crowing before sunrise doesn't cause the sunrise, despite the temporal order.
- Time is an indicator but not the sole determinant of causal relationships.