Podcasts

For 150 years, companies have shared the same basic shape: a pyramid of people moving information up and down. AI may make that structure optional. Consider electricity. It took nearly 30 years for factories to see real productivity gains after electric power became available. The reason is counterintuitive but instructive. Early factories were designed around a single steam engine. Power was distributed through belts and shafts, forcing machines to cluster tightly. Layouts w

Yesterday Anthropic released Claude Fable 5. Wharton professor Ethan Mollick handed Fable a 15-page spec and “it would work for 9+ hours and deliver terrific results.” Stripe used Fable to run a migration across a 50-million-line codebase in a single day. The internal estimate for doing it by hand: a full team, two months. Fable runs on the same weights as Mythos 5, the system that triggered government conversations about cyber and bio misuse last spring. In other words, Fabl

We are living through the most extraordinary capital event in the history of technology AND it's happening all at once. This week, OpenAI confirmed it has confidentially filed its S-1, with CEO Sam Altman noting they "have not decided on timing yet" because some ambitions are simply easier to pursue as a private company. The timing is staggering. It comes days before SpaceX is set to start trading and roughly a week after Anthropic filed confidentially with the SEC. Each of t

Two years ago, a gathering like this would have argued about whose AI was smartest. Bloomberg Tech in San Francisco spanned more than thirty sessions, the people running Netflix, Anthropic, Google, Meta, Databricks, Broadcom, Cerebras, Verizon, Anduril, World Labs, and the Federal Reserve on stage. That question about which model was furthest along, was treated as secondary. Databricks CEO Ali Ghodsi captured why. Databricks is the data-and-AI platform many large companies ru

At Build this week, Satya Nadella framed Microsoft's mission as helping every company "move from just consuming a frontier model to fully participating at the frontier." Microsoft is building the frontier intelligence ecosystem i.e., the full stack a company needs to build, run, and govern AI agents, from the infrastructure to the context to the workplace. Few key highlights from Build 2026: • Surface Laptop Ultra with NVIDIA RTX Spark AI chip • MAI-Thinking-1: Microsoft's

On the latest Moonshots (EP #258, "The Organizational Singularity: AI-Proof Your Company"), Peter H. Diamandis sits down with the always brilliant Salim Ismail to argue that the company itself is being redrawn. The current format which is top-down and human-centric, organized around hierarchy, is giving way to organizations that are AI-native and agentic, architected around intelligence rather than reporting lines. We've heard the word singularity, borrowed from physics and c

Kirkland & Ellis committed $500M to build its own AI platform. They’re doing it while still licensing the same frontier models everyone can rent tomorrow. So why spend millions more? The platform is being designed around the knowledge of approximately 250 of their lawyers. That’s the one thing a competitor can’t download. Now consider this: the pressure to build an edge on top of commodity intelligence runs all the way down: to every org, every team, and each ONE of us. Which

NVIDIA has spent billions on “light.” Since March 2026, it has announced $2 billion each in Lumentum, Coherent Corp., and Marvell Technology, plus additional photonics commitments with Corning and Ayar Labs. The goal is to move information with photons instead of electrons. To see WHY, let's start with how small computing has become. A human hair is roughly 70,000 to 100,000 nanometers wide. A silicon atom is about 0.2 nanometers across. Today’s leading chips are built on “2-

When something gets a futures market, you know it has become a commodity. That just happened to thought itself. TLDR: This week the Shanghai Futures Exchange, the floor that trades copper, steel, and gold, confirmed it's in the early stages of designing futures contracts on AI tokens. CME Group and Silicon Data are launching a compute futures market, pending regulatory review, and BlackRock CEO Larry Fink has said “a new asset class will be buying futures of compute”. Transla