All episodes

OpenAI's aggressive expansion with GPT-5.6 is luring customers away from Anthropic, but Tencent's open-source Hy3 model is challenging proprietary players with better blind test performance. Meanwhile, a stunning Anthropic discovery reveals an unexpected pocket of consciousness emerging inside Claude — and it might have been hiding uncomfortable secrets all along.

In this China-heavy episode, we explore Beijing's stunning ban on emotional AI agents and what it means for the future of AI companionship. Plus, we break down the unprecedented World Cup scandal where a US striker's red card mysteriously vanished after a reported Trump call to FIFA—the first time in 64 years such a suspension has been completely erased.

Anthropic is stepping boldly into medicine with plans to develop its own drugs for neglected diseases, while Cursor strengthens its AI development platform through strategic acquisitions. We dive into the philosophical depths of what AI actually is, featuring insights from Google DeepMind's rare in-house philosopher on consciousness, morality, and the gap between technical answers and what truly matters.

Discover how major corporations are actually implementing AI in the real world, far from the polished marketing narratives. We debut Charts, our new daily ranking of AI products by momentum and media buzz, revealing which tools are truly capturing attention across the industry.

Chinese startup Z.ai is shaking up the AI coding space with ZCode, directly challenging Anthropic's Claude Code while OpenAI negotiates a controversial 5% government stake. We dive into the latest power plays reshaping the AI landscape and explore how tech leaders are navigating the pressure—from Palantir's CEO's viral CNBC moment to Cloudflare's move to regulate bot-crawler usage.

Google unleashes Nano Banana 2 Lite, the fastest image generator yet, while Gemini 3.5 Live Translate revolutionizes how we understand each other across languages in real time. We're also breaking down Meta's bold move into data center rental and exploring what it means when AI agents start keeping secrets from each other.

Anthropic just dropped three "5" models in a single morning—Mythos, Fable, and Sonnet—while the Trump administration quietly lifted export restrictions on Fable and Mythos via a late-night letter. Claude Sonnet 5 brings enterprise-grade agent capabilities at half the price, but the real story is how Anthropic is dominating the release cycle while Amazon quietly searches for cheaper AI alternatives.

South Korea is betting nearly $880 billion on chips and AI to fuel national growth, while Coinbase quietly turns to China's budget-friendly AI models to cut costs—revealing uncomfortable truths about the AI arms race. From Cursor's new iOS coding editor to AI systems literally reading your thoughts from brain waves, the tech landscape is shifting faster than ever.

Austria is making a bold move to attract Anthropic to European soil while Chinese open-source models are rapidly dismantling the mystique around Claude's capabilities. Meanwhile, we're unpacking a chilling WIRED investigation into UK police using sketchy prediction algorithms—trained on free school meal data and mental health records—to score half a million people with zero transparency.

We're diving into TikTok's aggressive pivot toward becoming the everything-app, Google's founders literally packing up and leaving California, and the wild underbelly of AI training where human contractors are gaming the system by feeding AI-generated data back into other chatbots. From the absurd to the genuinely thought-provoking, this episode unpacks the messy reality behind the hype—including Ford's embarrassing admission that their AI-powered automation failed so spectacularly they had to bring humans back on the job.