AI Inequality
May 17, 2026
Summary
This episode explores emerging AI inequality, driven by compute scarcity, security concerns, and geopolitical restrictions limiting access to frontier models. The host analyzes Anton Licht’s essay “Cut Off,” which argues that access to top-tier AI will soon become scarce and selective, favoring US-based entities. The discussion highlights how business model shifts, like Anthropic’s pricing changes, and potential policy missteps, such as data center moratoriums, could exacerbate this divide between the “haves” and “have-nots.”
Topics discussed
Intro, sponsors, and show updates
Defining AI inequality: Access to frontier models
Anton Licht's 'Cut Off': Security constraints on AI
Distillation risks and model theft concerns
Compute crunches and the zero-sum nature of AI
US government influence and geopolitical leverage
Geopolitical rifts and unevenly distributed futures
Sponsor ads: KPMG, Robots & Pencils, Superintelligent, Blitzy
Proposed solutions: Infrastructure and global cooperation
Summary of thesis and current AI usage disparities
Economic impact: Pricing changes and developer access
Policy critique: Data center moratoriums and supply/demand
Conclusion: The political significance of AI access
Listen ad-free on Castria