The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) Social Commonsense Reasoning with Yejin Choi - #518
Sep 13, 2021
Yejin Choi, a professor at the University of Washington, dives into the fascinating world of common sense reasoning and its role in natural language processing. She shares insights from her recent keynote at the Stanford HAI workshop, discussing how AI can enhance creative storytelling and the challenges in achieving human-like reasoning. Topics include the evolution of knowledge graphs, the intersection of common sense and narrative, and the importance of integrating knowledge in AI. Yejin also challenges conventional views on AI scalability in the quest for Artificial General Intelligence.
AI Snips
Chapters
Transcript
Episode notes
Common Sense Reasoning Research
- Yejin Choi focuses on common sense knowledge and reasoning, an area previously deemed too difficult for AI.
- Recent advancements in neural models allow for processing large datasets, reviving this research area.
Defining Common Sense
- Common sense is defined as everyday, practical knowledge for safe and reasonable functioning, like knowing it's okay to leave a closet door open, but not a freezer door.
- This knowledge is contextual, cultural, and often expressed as defeasible rules of thumb.
Comet's Reasoning Abilities
- Comet, a common sense transformer model, makes inferences about everyday situations.
- It generalizes from a symbolic knowledge graph, demonstrating surprising reasoning abilities, like analogizing repelling an attack in a physical fight vs. a chess game.