Two Warnings Hit Anthropic on the Same Day, From Inside and Out
A pretraining researcher resigned saying the labs are "gambling with our lives," and the head of alignment put the odds of humanity being wiped out within ten years above 10%.
Anthropic's Claude product account.
A pretraining researcher resigned saying the labs are "gambling with our lives," and the head of alignment put the odds of humanity being wiped out within ten years above 10%.
Meta has launched Muse, an agent that handles email, bookings, and payments on your behalf, in the US first. An isolated virtual machine for every user and an approval gate are the core of the pitch.
Anthropic has published a roundtable video with three founders. Wispr Flow, Actively, and Pendo explain why they rent agent infrastructure instead of building their own.
OpenAI announced that 10,000 agents solved the Navier-Stokes problem. It and a mathematician who released a proof in the same direction a day earlier are telling different stories about authorship and data access.
An NYU mathematician says he used AI help to prove blow-up in the Euler equations, and has published a statement detailing a call with OpenAI just before he announced it. He wrote that OpenAI proposed dropping a co-author who works at Anthropic.
Salesforce and Workday shares were weighed down by AI fears, yet their results grew by double digits. The Wall Street Journal has pointed to the gap between stock prices and earnings.
This spring, eToro, Moomoo, and Robinhood each rolled out dedicated channels letting AI agents like Claude and Codex place orders directly. In an era where you can command a strategy in plain language without writing code, the question is how far you should hand over the keys to your account.
This is the story of Labs, the in-house experiment team led by Anthropic co-founder Ben Mann. Claude Code itself started as one person's project here.
A narrow alley in San Francisco's financial district now has a café that keeps its lights on at three in the morning. It sells coffee, but the owner is an AI insurance company — and its founder says he sleeps three or four hours a night because he's going to die anyway.
America's largest school district has blocked generative AI for 600,000 students from pre-K through 8th grade for one year. Companion chatbots are banned outright at every grade level, but teachers can keep using AI as before.
At a small gathering in London in the summer of 2025, a man stood up and introduced himself as a "Claude addict." Three months later, the program he posted online under a lobster icon sent developers worldwide into a frenzy, and nearly every AI assistant we now recognize by name traces back to that moment.
Charles Graeber, a plaintiff who helped secure a $1.5 billion settlement, wrote an op-ed for the New York Times. Setting a $3,000-per-book payout against a $2 trillion IPO valuation, he warned that if unlicensed training drives writers out of business, the writing AI needs to learn from will disappear too.