
Machine Learning Street Talk (MLST) Prof. Nick Chater - The Language Game (Part 1)
10 snips
Mar 1, 2024 In this conversation with Nick Chater, Professor of Behavioural Science, he unpacks the intricacies of human motivation using Tolstoy's insights and reflects on the brain as a prediction machine. The discussion navigates the complex relationship between AI learning and human cognition, emphasizing the limitations of certainty. Chater explores the evolution of language and collective intelligence, highlighting how communication adapts through cultural interactions. He also touches on the challenges of universalism in AI and the vital role of diverse perspectives in innovation.
AI Snips
Chapters
Books
Transcript
Episode notes
The Shoggoth Brain
- The brain, like a "shoggoth," is a mysterious entity whose workings are not fully understood, even by itself.
- Our rationalizations are merely attempts to make sense of this complex system, not accurate representations of its processes.
Fragmented Simulations
- Human simulations of the world are inconsistent and fragmented, demonstrated by thought experiments with simple shapes like cubes.
- Our understanding is based on local experiences and extrapolations, not a coherent model of everything.
The Limits of Consistency
- Consistent theories have limited scope, only applying to small slices of reality, like thermodynamics.
- The world's complexity requires multiple models and analogies, not a single universal framework.













