Is this @outlook outage a sign of things to come.... this just shows @AlboMP if your gunna regulate the shit out of AI, hackers will still attack us with the "Latest SOTA" models...
Dear @OpenAI,
For two years the whole industry waited on you.
Today it waited on you again, and Anthropic answered instead.
Your flagship, GPT-6 Astra, now costs more than the model that beats it on every row of Anthropic's launch table. Claude Fable 5.2: 20% cheaper, ahead on all 10 benchmarks listed.
Worse: in the same thread, Anthropic tells people to start with Opus 5.5 for most work. That's their cheaper model. It beats Astra on 7 of those 10 rows, at 60% less.
Your flagship is losing to your rival's default.
Meanwhile, the GPT-6.2 story is still rumors. Sol "routed to some users." Astra "in the logs." No price, no model card, no table.
I use Claude every day. I still want you to win the next round, because a 20% price cut on the top model doesn't happen when nobody is chasing.
So chase.
Ship Sol. Ship Astra. Make Anthropic answer you for once.