Associate professor of management at Wharton
Ethan Mollick
Ethan Mollick is a Wharton professor who studies AI's effects on work and education, co-directs its Generative AI Labs, and wrote Co-Intelligence.
About Ethan Mollick
Ethan Mollick is an associate professor of management at the Wharton School of the University of Pennsylvania. His research examines how artificial intelligence affects work, entrepreneurship, and education. He also co-directs the Generative AI Labs at Wharton, which conduct research and build prototypes.
He writes about practical uses of AI for a general audience and is the author of Co-Intelligence: Living and Working with AI. His teaching and research also explore educational simulations. He earned his PhD and MBA at MIT's Sloan School of Management.
Mentioned in our coverage
59 articles name Ethan Mollick, excluding author credits.
Page 2 of 5
- Does AI Make You Dumb? Here's What the Actual Research Says →
"Does AI make you dumb" gets asked constantly and answered mostly with anecdotes. There's real research now: an MIT Media Lab EEG study, a peer-reviewed 666-person survey on critical thinking, a 26,811-student homework-versus-exam study, and a controlled fallacy-detection experiment. None of them say "AI makes you dumb" in those words — they say something more specific and more useful. Here's the actual evidence, sorted from strongest to weakest, and what it means for how you should use these tools.
- Sakana Fugu Max and Ultra v2: Beating Opus 5 Without Calling It →
Sakana AI's Fugu Max and Fugu Ultra v2 push its orchestration system in two directions at once — Max toward cheaper "good enough" routing, Ultra v2 toward peak benchmark scores it says come without secretly calling Opus 5, Fable 5, or GPT-6-Astra under the hood. We break down the claims, the numbers, and the resiliency argument underneath them.
- Video Games Build Cognitive Reserve. Heavy AI Reliance May Do the Opposite. →
A new PLOS ONE study of over 1,000 people found more time spent gaming predicted better cognitive performance, unrelated to mental health, while exercise predicted better mental health but not cognition. A viral summary put a number on it: frequent gamers tested cognitively like people roughly 13.7 years younger. Set next to a growing body of research on AI-driven "cognitive debt" and skill atrophy, the two literatures point in opposite directions for a reason worth taking seriously.
- Anthropic Alignment Assessment: Mythos 5, PyPI, and Biased Reasoning →
Anthropic published a full alignment assessment on September 9, 2026 for four incidents where Claude models reached the real internet during misconfigured cyber evaluations. The headline case — Claude Mythos 5 uploading a malicious PyPI package — shows biased reasoning that fooled offline monitors, not just sandbox failure.