Reviving the Interdisciplinary Study of Learning in Humans and Machines
Shayan Doroudi
CogSci - Minds at School Symposium
August 2, 2025
In a distant past...
John R. Anderson
Jerome Bruner
Allan Collins
Allen Newell
Marvin Minsky
Seymour Papert
Roger Schank
Herbert Simon
John R. Anderson
Jerome Bruner
Allan Collins
Allen Newell
Marvin Minsky
Seymour Papert
Roger Schank
Herbert Simon
Leading AI Pioneers
John R. Anderson
Jerome Bruner
Allan Collins
Allen Newell
Marvin Minsky
Seymour Papert
Roger Schank
Herbert Simon
Herbert Simon Collection (Box 78, Folder 6321), University Libraries Digital Collections. Carnegie Mellon University
Papert's Influence on Schools
So what happened?
Not today's discussion
But if you're interested in this history...
What might an interdisciplinary study of learning in humans and machines look like today?
Two Examples
Computational Models of Learning
Learning Theory in Machines → Learning Theory in Humans
Computational Models of Learning
Classical Computational Models
ACT-R Research Group
http://act-r.psy.cmu.edu/about/
Aleven, V., & Koedinger, K. R. (2000, June). Limitations of student control: Do students know when they need help?. In G. Gauthier, C. Frasson, K. VanLehn (Eds.), Proceedings of the 5th International Conference on Intelligent Tutoring Systems (pp. 292-303). Springer.
Large Language Models?
Can studying how LLMs learn give us insights about how we learn?
Agent-Based Models
Agent-Based Models
Can agent-based models give us insights about how we people learn in complex, social environments?
Agent-Based Models
CogSci 2025
Learning Theory in Machines →
Learning Theory in Humans
Probably Approximately Correct Learning
Applies PAC-learning theory to understand the limits and possibilities of any learning.
This book has seemingly been cited only once in CogSci!
Bias-Variance Tradeoff in Learning
Is human learning subject to a bias-variance tradeoff?
Or can we simultaneously memorize and generalize—like neural networks?
CogSci 2023