Nicholas Thompson on Open Source, China, and AI Power Games | RegulatingAI Podcast

In this episode of the RegulatingAI Podcast, host Sanjay Puri is joined by Nicholas Thompson, CEO of The Atlantic, to talk about one of the most pressing issues in AI today: the scraping of content and the future of journalism in an AI-first world.  

✅ Topics covered: 

• The “original sin” of AI companies and scraped content 
• How The Atlantic is navigating AI disruption in publishing 
• Legal and ethical paths forward: lawsuits, licensing, and collaboration 
• Why Thompson thinks AI companies should drive traffic to journalism 
• A peek into The Atlantic's deal with OpenAI 

🔍 Nicholas shares candid takes on balancing innovation and fairness and what the future might look like if we don’t course correct. 

Resources Mentioned: 

https://www.linkedin.com/in/nicholasxthompson/ 

⏱️ Timestamps: 

00:00 - Podcast Highlights 

02:56 - The "Original Sin" of AI Companies - Data Scraping & Compensation 

05:04 - Media vs AI Companies: Finding Fair Value Exchange 

08:19 - Recent Court Rulings: Anthropic & Meta Cases Analysis 

12:19 - The Future of Search & Web Architecture 

16:45 - Generational Impact: How Young People Consume Information 

17:47 - Federal vs State AI Regulation Debate 

21:47 - What AI Developers Actually Want from Regulation 

22:05 - EU AI Act: Over-regulation Concerns 

23:45 - Open Source vs Closed Source AI Models 

24:43 - China's Open Source AI Strategy & US-China Relations 

26:19 - Chip Export Restrictions: Effectiveness & Consequences 

28:41 - US-China AI Cooperation Needs 

29:08 - The "Job Apocalypse" Debate: Dario vs Jensen 

32:48 - Government Role in AI Transition & Retraining 

34:43 - The "First Rung" Problem: Entry-Level Jobs at Risk 

35:51 - AI Medical Diagnosis: Outperforming Human Doctors 

39:07 - AI Companionship: Solution or Danger for Loneliness? 

41:10 - The "Westworld" Risk: AI-Powered Social Media Dystopia 

42:46 - Key AI Thinkers: Audrey Tang & The Vatican's AI Paper 

44:59 - Lightning Round: Quick Takes on AI's Future