Ethan Mollick
Associate Professor of Management, Wharton School, University of Pennsylvania
The Wharton professor running the largest real-world experiments on AI in business and education, and translating the results for everyone.
Credentials
Associate Professor at Wharton School of the University of Pennsylvania. PhD from MIT Sloan School of Management. Author of "Co-Intelligence: Living and Working with AI" (2024). Runs large-scale experiments integrating AI into MBA coursework. Advisor to multiple organizations on AI adoption. Previously researched entrepreneurship, innovation, and gaming.
Why They Matter
Mollick is the rare academic who actually uses the tools he studies — he requires his MBA students to use AI in coursework and publishes the results. His Substack "One Useful Thing" has become required reading for business leaders trying to figure out what AI can actually do today, not in theory. For SME owners, his practical experiments (not lab benchmarks) show exactly where AI helps and where it falls flat in real business tasks like strategy, writing, and analysis.
Positions
AI Timeline View
AI is already transformative for knowledge work right now. Focuses less on AGI timelines and more on the "jagged frontier" — AI is shockingly good at some tasks and terrible at others, and the boundary is unpredictable.
Safety Stance
Key Beliefs
AI has a "jagged technological frontier" — it can outperform experts on some tasks while failing at seemingly simple ones, and you cannot predict which without testing.
Harvard/Wharton BCG study and "One Useful Thing" Substack
Everyone should be using AI now and experimenting broadly — waiting for perfect tools means falling behind. The best way to understand AI is to use it extensively.
"Co-Intelligence: Living and Working with AI" (book), 2024
AI will reshape education fundamentally — banning AI in classrooms is futile and counterproductive. Students need to learn to work with AI, not pretend it doesn't exist.
Wharton course policies and public talks, 2023-2024
Middle managers and knowledge workers will see the biggest impact from AI — not replacement, but radical augmentation that changes what one person can accomplish.
"One Useful Thing" Substack and multiple interviews, 2023-2024
Controversial Take
Mollick argues that companies should mandate AI experimentation across all roles rather than forming cautious "AI committees" — the only way to find where AI adds value is to let everyone try it. This clashes with enterprise risk management orthodoxy and with AI safety advocates who want more controlled deployment.
Track Record
How well have Ethan Mollick's predictions held up?
AI would dramatically improve performance of lower-skilled consultants while providing less benefit to top performers (the "leveling" effect).
Made: 2023 (Harvard/Wharton BCG study)
The study showed consultants using GPT-4 improved output quality by 40%, with the biggest gains among below-average performers. Widely cited and replicated.
Universities that ban AI tools will reverse course within 1-2 years as the tools become impossible to detect and essential to professional work.
Made: 2023
Most major universities moved from AI bans to AI integration policies by 2024-2025. AI detection tools proved unreliable.
AI will create a "secret cyborg" problem — employees using AI without telling their employers, creating hidden productivity gains and hidden risks.
Made: 2023
Multiple surveys in 2024 confirmed the majority of AI-using workers do not disclose it to management, exactly as Mollick predicted.
Key Quotes
“Always invite AI to the table. You may not always use the result, but you should always see what it can do.”
“We are all the best AI users in the world right now, and also the worst we will ever be. The tools are changing that fast.”
“There is no instruction manual for AI. The frontier is jagged and the only way to map it is to explore it yourself.”
“AI doesn't replace humans. It replaces tasks. And the tasks it replaces are not the ones you expect.”
“The organizations that figure out how to use AI are not the ones with the best technology — they're the ones that let their people experiment.”
Publications
Co-Intelligence: Living and Working with AI
2024
Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of AI on Knowledge Worker Productivity and Quality
2023
New Modes of Learning Enabled by AI Chatbots: Three Methods and Assignments
2023
Connections
Agrees With
Last updated: 2026-04-12
←Back to AI Minds Directory