Big Technology Podcast

Anthropic’s leaked 2026 numbers anchor a wide-ranging tech news episode about AI market anxiety, IPO expectations, and whether the trade has gotten too chaotic. It also detours into a suspicious recording device at OpenAI, Apple’s HomeView, and the surprisingly heated Food Slop debate.

AI safety takes center stage as Kevin Roose examines whether feuding industry leaders can responsibly build systems more powerful than humans. The episode explores why he grew more concerned after reporting inside OpenAI, Anthropic, and Google DeepMind, and why he thinks the industry should slow down.

This episode surveys a busy week in AI and tech: Meta and Microsoft easing back on Claude, signs of slower momentum around Claude Code, and what that could mean for Anthropic. It also gets into agent mishaps, a banking-info leak tied to Grok Bot, and the surprising question of whether being rude to AI can get you banned.

AI infrastructure spending is the focus here, with SemiAnalysis’s Dylan Patel and Jordan Nanos examining whether exponential growth can continue. The conversation looks at where future revenue might come from, who benefits, and what happens if the boom keeps accelerating.

The episode examines the race to build the dominant personal AI assistant, with M.G. Siegler arguing that Meta’s Muse is stronger than many expect. It also looks at Apple’s opening, Google’s stumbles, Anthropic’s limited consumer presence, and how AI agents could change daily digital life.