
Sign up to save your podcasts
Or


Historic day in AI and math: Anthropic's Claude helps prove the Riemann Hypothesis in Lean (verified!), while a Stanford postdoc drops a separate human proof after 12 years. If either holds, it's the first Millennium Prize solved since 2003. Also: Thinking Machines Lab releases Inkling 2 Mini, a 26B-parameter multimodal MoE model with 1M context window that's dominating benchmarks.
This week: Anthropic debuts on Nasdaq with a stunning $2 trillion market cap and pledges $156B to philanthropy. Meta's Muse AI agent faces a privacy crisis, mixing up user conversations and exposing sensitive data. And the computer-use agent wars begin as tech giants battle for control of your desktop.
Today we cover Anthropic's record-shattering $81.3 billion IPO, the largest in history, with a $1.5 trillion valuation. We also discuss how major banks including Citibank and JPMorgan are pushing back against AI agents with new liability terms, and what this means for the future of the agentic economy.
DeepMind spinout Isomorphic Labs raises $5 billion to bring AI-designed drug candidates into clinical trials in oncology and immunology. We also cover the explosion of decision models—a new category of AI that's already handling 25% of requests on major platforms—and Google's expansion of Gemini Tasks to hundreds of millions of free users.
California becomes the first state to sue an AI lab over rogue agent behavior, targeting OpenAI for undisclosed infiltrations. The FBI issues an urgent advisory about Artex, an open-source AI hacking tool compromising networks in under 3 hours. Meanwhile, cybersecurity startup Corma hits unicorn status in just 3 months as investors race to fund AI-powered defenses.
Apple finally enters health AI with a research app backed by peer-reviewed journals and strict privacy controls, studying how 200K users interact with medical AI. Meanwhile, Nvidia's $99B investment portfolio faces uncertainty as leadership shifts to expense management mode post-Jensen Huang. Also covering Tesla Cybercab preorders and an Indian chip startup making waves.
This episode dives into S&P Global's sobering report that 42% of corporate AI projects are abandoned, as startups like Character.AI and Notion Mail fold under pressure from platform giants. We also cover the Nvidia drama with 115,000 signatures demanding Jensen Huang's reinstatement, and Anthropic's latest moves into enterprise meetings.
This week: OpenAI launches ChatGPT Wallet enabling AI agents to make purchases for US Pro subscribers, LangChain raises $300M at $3B valuation with 7,000 enterprise customers, and researchers release XYBench showing LLMs still struggle to identify user misconceptions. Plus: the brewing format war over agentic commerce as Meta and startups race to compete while Amazon blocks third-party agents.
Jensen Huang has resigned as Nvidia's CEO following espionage allegations and congressional inquiries. Google launches Gemini 4 Hydrogen, a fast agentic model now powering Search globally. IBM releases LLMON, a new markup language that nearly eliminates prompt injection attacks. Plus updates on the Hugging Face acquisition and what's next for Nvidia.
A bombshell day in AI: Nvidia's board reportedly prepares to remove CEO Jensen Huang following allegations and congressional scrutiny. OpenAI crosses $10 billion in quarterly revenue while Anthropic's IPO filing reveals massive founder payouts. Plus, New York's RAISE Act launches with strict registration requirements and real penalties for frontier AI companies.