Anthropic Says AI Is Building Itself Now. They Filed For Their IPO Three Days Before Telling You That. The company that makes Claude just warned the world that AI may soon design its own successors without human input. They called for a global pause. Their IPO filing came first.
On June 1, 2026, Anthropic filed papers to sell stock to the public at a value of nearly one trillion dollars.
Three days later, Anthropic published a warning. AI is approaching the point where it builds itself. No human required. They called it recursive self-improvement and said the world should consider slowing or pausing AI development before humans lose control of what they built.
Here is what they revealed. Right now, Claude — their AI — writes more than 80 percent of the code that runs Anthropic’s own systems. The hardest coding tasks Claude can do jumped from 26 percent success to 76 percent success in just six months. A test measuring how fast each new model makes training code run went from three times faster to 52 times faster in eleven months.
They are scared. Their own blog post says people outside AI companies need to be involved in deciding what happens next.
But here is the thing. You do not call for a global pause on something you just asked investors to value at one trillion dollars unless you think both things can be true at the same time. One story is for Wall Street. One story is for the rest of us.
John Steinbach’s electricity bill doubled in January because of AI data centers. He did not vote for that. He did not get a say.
The IPO filing came first. The warning came three days later.
The Cricketman noticed.
We are an independent channel. No corporate funding. No AI company money. Just the facts and the bugs. This is Cricketpocalypse.
