Saturday, April 18
OpenAI Sheds Skin, World Spreads
OpenAI is cutting deep into its executive layer, with chief product officer Kevin Weil (opens in a new tab) and Sora lead Bill Peebles both departing as Sam Altman publicly frames the exits as a purge of 'side quests' — a signal that the company is narrowing hard around its core AGI mission. The losses follow a broader pattern of senior talent churn at a company simultaneously managing a nonprofit-to-for-profit conversion, a $40 billion raise, and intensifying competition. Meanwhile, Cursor is in talks to raise over $2 billion (opens in a new tab) at a $50 billion valuation, a figure that would have seemed absurd two years ago, built almost entirely on enterprise developers who are defecting from incumbents. The AI coding boom is minting new infrastructure giants even as the lab that sparked the wave reorganizes around what Altman believes actually matters.
Sam Altman's World is executing a rapid land-grab in human verification, announcing integrations with both Tinder and Zoom (opens in a new tab) in a single news cycle — requiring users to scan their iris with World's orb to prove they are human before dating or joining a meeting. The strategy is straightforward: as AI-generated personas become indistinguishable from real users, whoever owns the authentication layer owns the trust infrastructure of the internet. The irony is sharp — the same AI wave that Altman has done more than almost anyone to accelerate is now the threat his other company is monetizing. Meanwhile, the EU's own age verification app was publicly hacked (opens in a new tab) with minimal effort at a live demo, demonstrating that the demand for credible identity solutions far outpaces what governments can build.
Anti-AI sentiment (opens in a new tab) is crossing from rhetoric into physical confrontation, with unions amplifying the message that automation threatens not just jobs but human dignity, while reports of violence against AI-associated targets are increasing. Amazon is drawing fresh fire by shipping new 4K Fire Sticks that block sideloading entirely, locking users inside its content ecosystem in a move critics read as the slow erasure of user ownership over personal devices. Anthropic is taking a different bet on public legitimacy, previewing a dedicated cybersecurity model called Mythos aimed squarely at winning back favor with the Trump administration after earlier friction. Across all three stories, the axis is the same: who controls the machine, who benefits from it, and who gets to push back.