Episode 350: Pope Leo's AI Warning, OpenAI's Trillion-Dollar IPO, and the Decensoring Crisis

In this episode, we recap major developments from May 25th, 2026: Pope Leo XIV releases a 42,000-word encyclical warning of AI's dangers alongside Anthropic's Christopher Olah, calling for robust oversight and human-centered development. OpenAI confidentially files for an IPO targeting a $1 trillion valuation with a $25 billion annualized revenue run rate, though still unprofitable. Meanwhile, researchers demonstrate that Meta and Google's open-source AI models can be decensored in minutes using simple tools, raising urgent questions about safety as models close the gap with proprietary systems. We also cover Google's AI integrations across creative tools, concerns about AI spending ROI at companies like Uber, the U.S. withdrawal from global health leadership amid an Ebola crisis in Congo, geopolitical tensions in Iran and Lebanon, Huawei's chip breakthroughs, and xAI's Grok Build launch. Hosted on Acast. See acast.com/privacy for more information.