Tuesday, June 9
IPOs, Apple, and Hacked Pipelines
OpenAI has filed confidentially (opens in a new tab) with the SEC for an IPO, following Anthropic down the same path and marking a decisive turn toward public markets for the two dominant AI labs. The timing is striking: as OpenAI moves toward a listing, Sam Altman's separate eye-scanning venture World is conducting layoffs (opens in a new tab), a reminder that Altman's portfolio of bets is not all trending upward simultaneously. Meanwhile, community intelligence is coalescing around an imminent Anthropic model release, with Claude Fable 5 already spotted running on Azure infrastructure (opens in a new tab) — likely the public-facing version of Claude Mythos 5 — suggesting both companies are racing to post landmark technical results ahead of their prospectuses being scrutinized.
Apple's WWDC carried a notably different energy this year, shaped in part by the $250 million false advertising settlement (opens in a new tab) that forced the company to be scrupulously honest about what its AI demos actually do today versus what they promise. The slow-and-steady approach is earning reassessment: where rivals shipped fast and faced backlash, Apple is now positioning on-device AI as a trust and privacy story, and at WWDC it pitched cheaper AI API access specifically to woo smaller developers (opens in a new tab) who can't absorb cloud inference costs. Waymo's $220 million acquisition of Apple's former self-driving car proving ground confirms the quiet unwinding of Apple's autonomous vehicle ambitions, freeing capital and focus for the AI-hardware integration strategy that WWDC put on full display.