Daily Tech News 18 June 2026

• Mandy Rice-Davies applies: Anthropic employees accuse Trump Administration of targeting them. (New York Times)
  
  The article is unreadable. Literally so if you don't maintain a paid NYT subscription for their puzzles page. A free login traps you in Dark Pattern Hell, an endless series of inescapable popup ads offering free paid trials cancel any time by submitting your cancellation request carved into the side of a living member of an extinct species of goat through first-class double-registered triple-certified mail. The usual archive sites don't work either.
  
  Even if you have a subscription or you know of another workaround not that I would ever endorse such a thing, you will simply find that it is not worth reading:

  "What are you telling your clients?" one employee asked in a chat viewed by The New York Times. Another said, "Does anyone know what to believe?" In another message, a worker said, "I don't understand what the issue is."

  Playing dumb or actually dumb? What difference, at this point, does it make?
  
  
  
• Anthropic asked for regulation. Washington went much further. (CNBC)
  
  An actually informed and balanced take on the situation.
  
  Quick precis: Anthropic announced its new Mythos AI tool with supernaturally dangerous hacking powers restricted to approved researchers, and its defanged Fable AI which is exactly the same thing only not. The Trump Administration ruled that if they're as dangerous as Anthropic claims, they need to be restricted for reasons of national security to American citizens on American soil. Anthropic threw its toys out of the pram and blocked access globally, and is now seeking to lay all the blame at the feet of that mean old Mister Trump.
  
  What CNBC highlights and The New York Times buries is a blog post written by Anthropic CEO Dario Amodei a fucking week ago:

  Frontier AI models, like airplanes, should be required to go through technical testing and auditing, and their release should be blocked or reversed as a threat to public safety if they do not meet high standards of safety.

  Amodei is every bit as much of a weasel as Sam Altman. And while Altman is a sociopath and a compulsive liar, Amodei is something much worse: He's whiny.

• World leaders want American AI. They just don't want America. (Tech Crunch)
  
  Substitute almost anything in place of AI and the story is the same.
  
  
  
• The Orange Pi 6 is a nonplussed and hopefully cheaper version of the Orange Pi 6 Plus. (Liliputing)
  
  This is a single-board computer along the lines of a Raspberry Pi, but more powerful than even the latest Raspberry Pi 5. Just one drawback: That power means the cheapest model of the Orange Pi 6 Plus costs $367.
  
  
  
• Hewlett Packard is offering a year of free access to its new virtualisation platform, targeting customers drained of blood and left to die in a ditch following Broadcom's acquisition of VMWare. (Ars Technica)
  
  After the free year is over, HP is suggesting prices around $600 per CPU socket.
  
  VMWare charges around $400 per CPU core.
  
  And a single CPU socket can run to 288 cores.
  
  
  
• OpenAI increased its revenue from $3.7 billion in 2024 to $13 billion in 2025. (Where's Your Ed At)
  
  Thanks to careful cost-control, losses over the same period dropped from $5 billion in 2024 to $38.5 billion in 2025.
  
  Oh.
  
  Actually, before fudging adjustments the company's net loss for 2025 reached $60 billion.
  
  
  
• US science is in chaos. (Scientific American) (archive site)
  
  Thanks to... Scientists no longer having access to an infinite money glitch.

posted by Pixy Misa at 04:30 AM

| Access Comments