Episode Details

In this episode of Model Behavior, Nina Park and Thatcher Collins examine a critical data exposure incident affecting Anthropic users and the strategic shifts in model deployment following the release of Claude Opus 5. We dive into the technicalities of why shared chatbot conversations keep appearing in public search results and what this means for enterprise security. The show also covers the 'psychosis' surrounding Chinese AI advancements, specifically the launch of Moonshot AI's Kimi, and the resulting policy debates in Silicon Valley and Washington D.C. Finally, we look at the infrastructure side of AI with Gigatoken, a tool aiming to resolve the preprocessing bottlenecks that slow down large-scale model training.

Topics Covered

• 🔐 Data Privacy: Anthropic's Claude conversations indexed by Google search
• 📊 Model Economics: Routing strategies for Opus 5, Sonnet 5, and Haiku 4.5
• 🌐 Geopolitics: The debate over Chinese open-weight models and U.S. protectionism
• 💻 Infrastructure: Gigatoken's 1000x speed increase for text preprocessing

Neural Newscast is AI-assisted, human reviewed. View our AI Transparency Policy at NeuralNewscast.com.

• (00:12) - Introduction
• (00:27) - Claude Data Leak
• (00:27) - Chinese AI and Policy
• (00:27) - Anthropic Model Strategy