Search for a command to run...

Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Medium CEO Tony Stubblebine joins the podcast to discuss the evolving landscape of writing and publishing in the AI era. (02:32) The conversation covers how Medium is fighting AI-generated spam, which increased by 10x but was manageable with existing filtering tools. (06:54) Stubblebine explains that while AI "slop" flooded the platform, most readers never see it due to Medium's recommendation systems and anti-spam measures. The discussion explores how AI is becoming genuinely useful for writers as research and editing assistants, while examining the fundamental tension between AI companies training on content without compensation and publishers' need for sustainable business models. (32:45)
Tony Stubblebine is the CEO of Medium, the popular online publishing platform. He took over leadership during a challenging period and successfully guided the company back to profitability while repositioning it to serve "real people" rather than competing in the professional creator economy. Before becoming CEO, Stubblebine was a partner to Medium and has extensive experience in the productivity and writing tools space, bringing a deep understanding of how people think and write.
When Wired reported that 47% of Medium content was AI-generated, Stubblebine pushed back on the framing. (02:32) The key insight is that platforms should be measured by what reaches users, not what gets submitted. Medium saw a 10x increase in AI slop submissions but maintained quality for readers through filtering systems. This approach recognizes that most platforms already delete the majority of submitted content as spam. The lesson for any content platform or business is to focus metrics on user experience rather than raw submission volumes, and invest in systems that filter signal from noise before it reaches your audience.
Stubblebine explains that societies run on exchange of value, and AI companies training on content without offering anything in return violates this fundamental principle. (17:17) Medium discovered its content heavily influenced ChatGPT's writing style (particularly em dashes), proving how significant their corpus was to AI training. This creates leverage for content creators and platforms. The takeaway is that if you create valuable content, you have more power than you might realize in negotiations with AI companies, and collective action can force better terms than individual efforts.
Despite AI advances, Stubblebine argues that writing will never disappear because "writing is thinking, and smart people like to think." (11:41) The process of putting thoughts into words reveals gaps in understanding and forces deeper thinking. He gives the example of believing you understand something until you try to write it down and realize there are huge holes in your knowledge. This insight applies beyond professional writing - anyone seeking to advance their career or expertise should continue writing to clarify their thinking, regardless of AI capabilities.
The most promising use of AI for writing is as an assistant that helps you stay in your flow of thinking rather than replacing the thinking itself. (23:10) Stubblebine describes AI that can search and organize your own writing in natural language, and provide editing suggestions during revision passes. This "lets you be more human" by handling organizational tasks while you focus on ideas. For professionals, this means embracing AI tools that augment your core skills rather than trying to replace your expertise entirely.
Unlike other platforms that pocket AI licensing deals, Medium plans to pass any AI company payments directly back to creators. (40:03) Stubblebine argues that platforms don't own the content and shouldn't profit from selling data they didn't create. This represents a fundamental choice about platform business models - whether to extract value from user content or to genuinely serve as an intermediary. The broader lesson is that sustainable creator economies require platforms to align their incentives with creators rather than extracting value at creators' expense.