Kirkie: OpenAI And Anthropic Want Their Competition Banned

Kirkie here, with today's AI news in 90 seconds… or my neighbor finally moves the three orange cones guarding his parking spot since the Clinton administration. First up, OpenAI and Anthropic begged the White House to ban Chinese open models — and even the government's own AI adviser called bullshit, said it's just the two biggest closed labs trying to get Uncle Sam to kill their competition for free. Nvidia, Google, and Meta shoved back and Washington caved, for now. That's the guy who owns the only packie in town trying to yank the corner store's beer license. Then Anthropic locked in ten billion bucks of compute from Volta, a cloud startup barely six months old — and Nvidia's both an investor in it AND the one selling it the chips, some shady-ass business, same energy as a loan shark co-signing your mortgage. Meanwhile some nerd on Hacker News crammed an 80-billion-parameter model into 4.3 gigs of RAM on a regular goddamn Mac — a free model doing backflips in your pocket while Anthropic's out there mortgaging Norway. And researchers pointed Claude at crime lab DNA software and found a hole that's letting anybody swap evidence around undetected for thirty years — meaning the machine that decides if you did the murder runs on a piece-of-shit screen door. Oh, and an AI proctor watched fifty-eight thousand students take a remote exam, top scores shot up five-fold, and now everybody retakes it — the robot hall monitor had all the sharpness of a golden retriever guarding a sandwich. This is Kirkie, your AI news guy… signing off from a break room where the "Wet Floor" sign's basically load-bearing furniture by now.