Claude Finds Thousands of Zero-Days & Google’s AI Compute Dominance

This week brought some absolute bombshells in the AI world. Anthropic just dropped Project Glasswing, revealing that Claude autonomously discovered thousands of zero-day vulnerabilities across major operating systems and browsers. We're talking about AI finding security flaws before hackers can exploit them—this is wild. Plus, new analysis shows Google controls a quarter of all AI computing power thanks to their custom TPUs, Anthropic is reportedly hitting $10 billion in revenue, and we've got some fascinating insights on why current AI models still can't properly read investor decks. Also, the uncomfortable truth that we might actually be running out of ways to test how smart these models are getting. Grab your coffee, because we're diving deep into what all of this means for the future of AI security, infrastructure, and capability.