Are We Thinking Correctly About AI Intelligence?

Melanie Mitchell, Janna Levin, Steven Strogatz

Published Aug 20, 2026

52 minutes

Episode site

Guests:

Melanie Mitchellcontributor
Janna Levincontributor
Steven Strogatzcontributor

No recording yet

Episode Description

Computer scientist and cognitive scientist Melanie Mitchell discusses why artificial intelligence does not think or reason like humans, how developmental and comparative psychology can improve machine-cognition experiments, and six principles for testing AI competence, robustness, and failure modes.

Topics Discussed

AI intelligencemachine cognitionexperimental designAI benchmarks

More Episodes

How to Improve Motivation & Overcome Procrastination | Dr. Masud Husain

Aug 24, 2026

BI 244 Marco Facchin: Philosophy and Science of Biological Brains

1h 30mAug 19, 2026

Neuroscience of Emotions & Tools for Improving Emotion Regulation | Dr. Ralph Adolphs

Aug 17, 2026

Briefing Chat: Anthropic rolls out new AI watermark — will it make a difference?

Aug 14, 2026

Using AI to Increase Your Intelligence & Enrich Humanity | Dr. Fei-Fei Li

2h 6mAug 10, 2026