AI

When a Meta employee hijacked a company-wide livestream last Thursday and screamed at colleagues to tell a senior AI executive he was “a piece of shit,” it wasn’t one person having a bad day. It was the eruption of months of fury inside one of the most ambitious — and botched — corporate reorganizations in tech history. Welcome to Meta’s Applied AI unit, where roughly 6,500 engineers are openly calling their new assignment “the gulag.” ...

The AI industry just had one of its most revealing weeks — not because of any single announcement, but because three companies made moves that expose where this market is actually heading. Anthropic launched Claude Fable 5 at premium prices with built-in safety classifiers. Google slashed its AI subscription to $5/month. OpenAI acquired a cloud infrastructure company to make its coding agents run for days without stopping. Three different bets. One clear signal: the AI market is fracturing into tiers, and the gap between “good enough” and “best available” is becoming a strategic decision that actually matters. ...

Anthropic just did something no one expected: it released the most powerful AI model on the planet and simultaneously told everyone it’s terrified of what models like it can do. Welcome to the era of Claude Fable 5 — the first public Mythos-class model, and possibly the last time “safety” and “frontier AI” will share a sentence without irony. Mythos Goes Public (Sort Of) Fable 5 is a Mythos-class model wrapped in safety classifiers. Think of it as a Formula 1 engine with a speed governor — still absurdly fast, but with guardrails that prevent it from doing anything Anthropic’s safety team would lose sleep over. ...

After a decade of “Hey Siri” being the punchline of every tech joke, Apple finally did the unthinkable: it admitted Siri was broken, tore it down to the studs, and rebuilt it from scratch. The catch? The new brain running the show isn’t Apple’s. It’s Google’s. Welcome to Siri AI — the most Apple product announcement imaginable. Late, polished, and powered by someone else’s technology. A Brand New Siri (Finally) WWDC 2026 opened with what Apple should have shipped two years ago: a dedicated Siri app with persistent conversations. No more fire-and-forget voice commands that evaporate the moment Siri responds. You can now have an ongoing dialogue, reference previous questions, and build on context across sessions. ...

For nearly eight decades, the sharpest mathematical minds on the planet stared at a deceptively simple geometry puzzle and came up empty. Then an AI chatbot cracked it in two weeks. On May 20, OpenAI announced that one of its internal reasoning models had autonomously disproved a famous conjecture posed by legendary Hungarian mathematician Paul Erdős in 1946. The result has been independently verified by a panel including Fields medalist Tim Gowers — and it’s being called the first time AI has independently solved a prominent open problem in active mathematical research. ...

Anthropic said Claude Mythos was too dangerous to release publicly. They shared it with 40 handpicked organizations. They committed $100 million to help patch the vulnerabilities it found. It leaked anyway. And now Google has confirmed the first known case of an AI-developed zero-day exploit being used by actual cybercriminals in the wild. The containment era lasted about six weeks. The Model That Finds What Humans Can’t Quick recap for anyone not tracking this saga: Claude Mythos Preview is Anthropic’s unreleased frontier model that autonomously discovers zero-day vulnerabilities — unknown flaws in software that developers haven’t patched because they didn’t know they existed. ...

Your bank is being patched at emergency speed right now, and you probably have no idea why. Anthropic’s Mythos — the AI model so dangerous it was never publicly released — is ripping through Wall Street’s cybersecurity posture like a chainsaw through wet paper. US banks are uncovering thousands of vulnerabilities they never knew existed. Japan’s three largest banks just got access. And the cybersecurity community is arguing about whether any of this is actually new. ...

The Chief Data Officer took a decade to go mainstream. The Chief Digital Officer never really got there. The Chief AI Officer? Eighteen months. IBM’s latest study — surveying 2,000+ CEOs across 33 countries — drops a jaw-dropping stat: 76% of organizations now have a CAIO, up from just 26% in 2025. That’s not adoption. That’s a corporate stampede. Why This Role Exists Every company has a CTO managing infrastructure and a CIO handling systems. But who owns the question of how AI actually changes the way work gets done? ...

$4.8 Trillion and Climbing Alphabet just crossed $4.8 trillion in market cap. Nvidia sits at $5.2 trillion. The gap is closing fast. A year ago, this race wasn’t even a conversation. Google was the “search company with an AI problem.” Now it’s the integrated AI powerhouse that Wall Street can’t stop buying. The stock is up 160% in twelve months. What changed? Everything — and nothing. Google always had the pieces. It just took the market a while to notice they’d assembled them. ...

Follow the Money in a Circle Nvidia crossed $40 billion in equity investments in 2026. The biggest chunks: $30 billion into OpenAI, $3.2 billion into Corning, $2.1 billion into IREN. Here’s the pattern that has people uncomfortable: Nvidia invests in a company. That company buys Nvidia GPUs. Nvidia books the revenue. Money out, GPUs out, money back in. Rinse and repeat at scale. The Bloomberg Investigation Bloomberg’s investigative team started pulling the thread in April. What they found isn’t illegal. But it raises questions that Nvidia’s 35x forward earnings multiple probably can’t absorb. ...