The Computational Gauntlet of Human-Like Learning

Explore a thought-provoking seminar on developing AI systems that learn in a human-like manner. Delve into the challenge posed by Pat Langley from Stanford University, as he illustrates this concept using mathematics and driving domains. Examine the history of machine learning and its evolving relationship with cognitive psychology. Discover the characteristics of human behavior that form a 'computational gauntlet' for AI systems to overcome. Learn about five AI systems that have successfully navigated most of these obstacles, serving as role models for future research. Gain insights into encouraging more studies on human-like learning in AI. Understand the speaker's extensive background in AI and cognitive science, including his contributions to scientific knowledge discovery and experimental studies of machine learning.

Introduction
Problem
Mathematics
Driving
Machine Learning History
Machine Learning Evolution
Core Features of Learning
The Gauntlet
Modular Cognitive Structures
Incremental Learning
Prior Knowledge
Innate Knowledge
Why should we change paradigms
Are there any questions
What is cobweb
Learning
Prodigy
Sage
What can we do
Questions
Critiques