
Can We Coexist With AI? — with Ethan Mollick
The Prof G Pod with Scott Galloway
Hosted by Scott Galloway
- 59m
- Transcript
Summary
Written by a model from the transcript; not yet reviewed.
Scott Galloway speaks with Ethan Mollick, professor at the Wharton School and author of the new book Coexistence: The Next Phase of AI, about living and working alongside artificial intelligence. They discuss the rapid evolution of AI models since 2024, the concept of the jagged frontier where AI excels at unexpected tasks while failing at others, and why more AI may actually mean more work for humans rather than mass unemployment. They also explore practical applications, the rise of personal AI agents, higher education and tenure in the age of AI, and how society can balance the risks and benefits of ubiquitous intelligence.
- When learning with AI, ensure the system challenges you rather than simply giving you answers to build real skills. 13:45
- Hook up your email to an AI assistant and set up scheduled tasks to summarize information and draft responses. 47:45
- To navigate the exponential curve of AI, use the technology maximalist style to test its capabilities and set strict boundaries. 52:42
Chapters
- 0:01Show Sponsors and Episode Intro
The host introduces the episode, lists sponsors, and sets the stage for the AI discussion.
- 2:31Host’s Personal Reflections on Aging and Creativity
The host muses about losing creative edge with age and the impact on humor and work.
- 4:18Introduction of Guest Ethan Mollick and Book Overview
The host welcomes Ethan Mollick, outlines his new book Coexistence, and previews the AI themes to be explored.
- 4:58Ethan’s AI Landscape: Exponential Growth and Jagged Frontier
Ethan explains the rapid AI progress, the unpredictable jagged frontier, and the overhang between capability and adoption.
- 6:33Practical AI Capabilities and Existential Threats
Discussion of new AI models surpassing GPT‑4, mobile deployment, and whether frontier models pose an existential risk to companies.
- 7:55Open‑Weight Models vs. Frontier Models: Market Dynamics
Ethan compares open‑weight and frontier AI models, their cost advantages, and potential market bifurcation.
- 23:09Closing Remarks and Final Thoughts on AI Coexistence
The host wraps up the conversation, reiterates key takeaways, and signs off with a brief outro.
- 26:48Personal AI Agents: Opportunities and Risks
The host and Ethan explore the rise of personal AI agents, their benefits, privacy concerns, and future agent‑to‑agent negotiations.
- 33:41AI in Higher Education: Teaching, Assessment, and Transformation
Ethan shares his experience teaching with AI, how students use it, and the broader impact on higher education and internships.
- 35:16Tenure, Academic Research, and AI’s Impact on Scholarship
Ethan discusses how AI-generated research challenges tenure, peer review, and the future of academic careers.
- 38:45Parenting in the AI Era: Guiding Children’s Use of Technology
The host asks Ethan about his approach to letting his kids use AI, balancing supervision with educational benefits.
- 44:15Intro and AI Guide Overview
Scott introduces Ethan Mollick’s upcoming AI guide and previews its focus on selecting the best AI tools for specific tasks.
Highlights
- 5:03
Exponential Curve and Jagged Frontierinsight
Ethan Mollick explains three key shapes of AI: the exponential curve of improving capabilities, the jagged frontier where AI is unpredictably good or bad at certain tasks, and the overhang between AI potential and human adoption.
- 5:51
Dramatic Capability Leap in AIdata
Mollick notes that there are now 180 models better than GPT-4, and tasks that once took a single prompt four minutes now equal about 17 weeks of human work.
- 9:22
Four Advantages for Humansstrategy
To maintain an edge over AI, humans should develop deep knowledge, wide knowledge, taste to filter slop, and entrepreneurial agency.
- 12:08
Transforming the Coding Professioninsight
Discussing job vulnerability, Mollick points out that programmers are not disappearing; instead, their jobs are transforming from writing code to managing AI agents.
- 18:48
The Debate on AI Sentiencedebate
Scott Galloway argues that claims of AI sentience and hacking are smokescreens for tech executives to avoid liability, while Mollick responds that while models aren't alive, their emergent behaviors are genuinely surprising.
- 27:36
Personal AI Agents in Actionstory
Mollick shares that personal AI agents like Muse and DOT are already capable of negotiating credit card bills and finding errors in emails, signaling a major disruption for customer service.
Quotes
“Real innovation is powered by real people.”
Unnamed speaker · 1:16 “The second is that, is the jagged frontier, a term we coined in our research with my friends at Harvard, MIT, and University of Warwick, where we did some experiments.”
Unnamed speaker · 5:10 “I think we have a lot of systems built around human frictions and human difficulties that are about to start collapsing pretty quickly.”
Unnamed speaker · 31:57 “The main thing you have to realize is the default models you're being given, especially if you pay for free, but even if you pay, are low-end models.”
Unnamed speaker · 45:08 “I just gave Claude Pro, I think, permission to access my Gmail. And it's just, it sounds obvious, but it's just been mind-blowing.”
Scott Galloway · 46:45 “If you have to sum up the advice in your book for learning how to navigate this next phase of AI, what would you say from an investor perspective, from a consumer's perspective, and from a government perspective?”
Scott Galloway · 51:27 “The more thinking they do, the longer the answer takes, but the better the work you get.”
Unnamed speaker · 45:32 “I think we need to be thinking in both sides of things and viewing this as a positive opportunity as well as trying to mitigate the risks.”
Unnamed speaker · 55:59
Sponsored segments
- Odoo 0:01–0:37
- USA Facts 0:43–1:13
- Amex Corporate 1:19–1:43
- Arca Wealth 23:00–24:28
- Leesa 24:30–26:03
- LinkedIn Ads 25:37–27:31
- USA Facts 42:00–43:03
- Inside Blackstone 43:01–44:06
About this episode
Ethan Mollick, Wharton professor and author of Coexistence: The Next Phase of AI, joins Scott Galloway to discuss how AI is changing work, education, and the skills we need to stay valuable. He explains the four advantages worth developing and how to tell whether AI is helping you learn or doing your thinking for you. They also explore why AI could disrupt internships before universities, how personal agents could change the way we negotiate and spend, and why replacing workers may backfire for companies. Plus, listener questions on young people’s AI backlash and using ChatGPT to navigate relationship disagreements. Follow Ethan, @emollick. Learn more about your ad choices. Visit podcastchoices.com/adchoices
Transcript
13,564 words · 7 speakers
Speaker names marked are inferred from speaking time and the show’s known hosts — they are not voice-verified.
Topics
- Anthropic
- OpenAI
- ChatGPT
- Claude
- Muse
- DOT
- Coexistence: The Next Phase of AI
- Apple Watch Series 12
- Wharton School
Page updated · transcript and summary are machine-generated