Microsoft bans Claude to cut AI costs

WWDC 2026: Siri on Gemini, a Foldable iPhone, and Cook's Last Keynote

Apple's WWDC 2026 keynote did three things at once: announced Tim Cook's retirement and John Ternus as the next CEO, rebuilt Siri on top of Google's Gemini models, and quietly seeded code for a foldable iPhone into iOS 27. This episode breaks down all three. The Siri rebrand is the headline. The newly named "Siri AI" runs on Gemini through Apple's Private Cloud Compute layer, gets a standalone app, and adds real-time screen awareness plus personal context across apps. It's the first time Apple has handed its assistant to a competitor's model, and the privacy framing on stage was clearly built to answer the question that move invites. The foldable iPhone story isn't in the keynote, it's in the code. Analysts pulled flexible display references and new app-adaptability tools out of iOS 27 betas, the strongest signal yet that the long-rumored foldable is closer than Apple is saying. Then the platform updates. iOS 27 brings up to 30% faster app launches and supports every device back to the iPhone 11. macOS 27 "Golden Gate" drops Intel support and refines the Liquid Glass design system. The Health app added perimenopause and menopause tracking, and Apple Watch picked up updates aimed squarely at Garmin and Whoop. Expanded parental controls now require child accounts for under-13s. Two things that almost got buried. Siri AI won't launch in Europe or China at first because of regulatory complexity, which leaves Apple's two largest non-US markets out of the headline feature. And this was Tim Cook's final WWDC as CEO. He hands the role to hardware chief John Ternus on September 1, ending a 14-year run. We cover what it means for Apple's identity that the privacy-first company now routes its assistant through Google, why Ternus over Federighi is a hardware-first bet at exactly the moment AI is software-defined, and what foldable code in iOS 27 says about the iPhone 18 roadmap. WWDC 2026, Apple WWDC, Siri AI, Google Gemini, iOS 27, foldable iPhone, Tim Cook retirement, John Ternus, Apple Intelligence, macOS Golden Gate, Apple Watch, Apple Health.

OpenAI Files for IPO at $852 Billion (and Losing $1.22 Per Dollar)

OpenAI confidentially filed for an IPO on May 22 and announced it publicly on June 8. The valuation: $852 billion. The catch: the company loses $1.22 for every dollar it earns, and internal documents project a $14 billion loss in 2026 with no path to profitability until 2029. This episode breaks down the filing and the math behind it. Revenue is running around $2 billion a month, tripling year over year since 2023. The March funding round closed at $122 billion. Goldman Sachs, Morgan Stanley, and JPMorgan are leading the deal, and Sam Altman is targeting a September listing, which would put OpenAI at 34x to 40x revenue at a price between $852 billion and over $1 trillion. Then there's the competitive context. Anthropic filed its own confidential S-1 a week before OpenAI, at a $965 billion valuation, which now sits above OpenAI's. SpaceX starts trading Friday at $1.75 trillion. Three of the largest IPOs in history are landing inside a month, and the order they go matters: if Anthropic prints a profitable quarter before OpenAI lists, the market gets a benchmark for what a "good" AI company looks like, and OpenAI has to clear it. The filing also became possible because of one ruling. Two days before the confidential submission, a jury dismissed Elon Musk's lawsuit against OpenAI on a statute of limitations technicality. That was the single biggest legal obstacle to going public, and it was cleared first. We cover what the numbers mean for developers and businesses building on the API, why a public OpenAI optimizes for margins instead of developer experience, what the tender offer for employees signals about liquidity pressure, and whether public market investors will actually pay a premium on a company burning $14 billion a year. OpenAI IPO, OpenAI S-1, Sam Altman, $852 billion valuation, AI IPO 2026, Anthropic IPO, SpaceX IPO, AI bubble, AI stocks, ChatGPT, Goldman Sachs, September IPO.

The Nerdy Escorts Cashing In on Silicon Valley's AI Boom

A Forbes investigation by Anna Tong put a number on something Silicon Valley wasn't talking about: a small group of high-end escorts charging AI founders thousands an hour, and selling intellectual conversation about GPUs, crypto, and longevity alongside the sex. This episode breaks down the reporting and the economics behind it. The rates are the headline. Aella, an escort and self-described data scientist, charges $6,000 an hour, the highest rate in the piece, and is credited with coining the "nerd-first" label. Meida Marek charges $3,500 an hour and says she's booked months out. Talia Sable, a former programmer who lists Dungeons & Dragons and supply chain logistics among her interests, charges $3,000. Forbes cites figures up to $23,000 a day and $30,000 a weekend, where five years ago it was rare to charge more than $1,000 an hour. The why is the part worth sitting with. It's a lens on how the AI gold rush is reshaping social life in the Valley, where founders raising at huge valuations and working 100-hour weeks deprioritize ordinary relationships, and a market fills the gap with transactional intimacy that doubles as founder therapy. There's also a labor angle that ties this directly to the AI story. Marek left an entry-level finance job because she grew anxious that AI would automate her career, then pivoted to a relational skill she figured a model couldn't replicate. We cover that bet, whether it holds, and the obvious risks around discretion when founders talk freely in private. A note on the numbers: most of these rates are self-reported marketing, and people in adjacent corners of the industry have publicly called them inflated. Treat them as claimed, not audited. Silicon Valley AI boom, nerdy escorts, intimacy as a service, AI founders, Aella, Meida Marek, Anna Tong Forbes, AI economy, automation, future of work, tech wealth.

Musk's $119 Billion Chip Plant: ASML CEO Says He's 'Very Serious

Elon Musk is in direct talks with ASML to build TeraFab, a Texas chip plant with a potential price tag of $119 billion. ASML CEO Christophe Fouquet confirmed he's spoken directly with Musk and called him "very serious" about the project and his broader semiconductor and satellite ambitions. This episode breaks down what TeraFab actually is and why it depends on one Dutch company. A Texas filing puts the initial investment at $55 billion, with total costs reaching up to $119 billion, one of the most expensive semiconductor projects ever proposed on US soil. Musk announced it in March with an initial $20 billion stake, aiming to produce logic chips, memory, and advanced packaging under one roof. Intel joined in April and plans to contribute its 14A process node, targeting 2-nanometer production. The catch: there's no path to 2nm chips that doesn't run through ASML. Every major chipmaker (Nvidia, TSMC, Samsung, SK Hynix, Micron, Intel) relies on ASML's lithography systems. We cover the supply crunch Fouquet is warning about ("demand on AI is coming so strongly that we will be in a supply-limited market for quite a while"), the next-gen High-NA EUV tools with first logic chips expected within months and Intel as the earliest adopter, ASML's projection that the chip market could hit $1.5 trillion by 2030, and Fouquet's warning that Europe risks falling behind on AI because of regulatory complexity. SpaceX, xAI, and Tesla all stand to draw from the same constrained chip supply, which is the thread tying this to Musk's whole empire. TeraFab, Elon Musk, ASML, semiconductor manufacturing, High-NA EUV, Intel 14A, 2nm chips, AI compute, chip shortage, Christophe Fouquet.

WWDC 2026: Everything Apple Announced (Siri Now Runs on Gemini)

Apple rebuilt Siri from the ground up at WWDC 2026, and the surprise is what's under the hood: Google Gemini. The newly renamed "Siri AI" gets a standalone app and real-time screen awareness, with Gemini-based models integrated into Apple Intelligence through Private Cloud Compute. This episode covers everything from the keynote. The Siri overhaul and what running on a competitor's models means for Apple. iOS 27, which brings up to 30% faster app launches and supports every device back to the iPhone 11. macOS 27 "Golden Gate," which drops Intel support and refines the Liquid Glass design. The expanded parental controls, including mandatory child accounts for under-13s. Two things the headlines almost missed: Siri AI won't launch in Europe or China because of regulatory challenges, and this was Tim Cook's final WWDC as CEO. He steps down September 1, handing the role to hardware chief John Ternus. WWDC 2026, Apple Siri AI, Google Gemini, iOS 27, Apple Intelligence, macOS Golden Gate, Tim Cook, Apple AI partnership, Liquid Glass.