The Gradient: Perspectives on AI

Subbarao Kambhampati: Planning, Reasoning, and Interpretability in the Age of LLMs

127 snips
Feb 8, 2024
Subbarao Kambhampati, Professor of computer science at Arizona State University, discusses planning, reasoning, and interpretability in the age of LLMs. Topics include explanation in AI, thinking and language, scalability in planning, computational complexity in LLMs, and concerns about misinformation generated by LLMs.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
ANECDOTE

AlphaGo's 33rd Move

  • AlphaGo's 33rd move surprised Go experts, highlighting a gap in human understanding.
  • Humans lacked the vocabulary to explain the move's strategic brilliance.
INSIGHT

Symbolic Explanations

  • Symbolic explanations are compact representations of complex situations, bridging the gap between sub-symbolic reasoning and human understanding.
  • This is crucial for explaining AI decisions, as humans need symbolic models to grasp complex reasoning.
INSIGHT

Thinking vs. Language

  • While language feels central to thought, much subconscious activity occurs without it.
  • Fluency in skills like music can bypass explicit language-based thought.
Get the Snipd Podcast app to discover more snips from this episode
Get the app