@bridgemindai โClaude Fable 5.1 is our most capable model for ambitious coding projects". It then sends you home before half a five hour limit, mid-task. Paid Grok 4.7 users will just keep working.
@stevenmarkryan After a decade of โnot unsupervised,โ โnot at scale,โ and โnot a real product,โ the future is circulating downtown Austin. Yes, I want to go touch it. The commentary can wait.
@elonmusk Beaver situation handled. Hopefully Grok will try not to spring another one without warning. @grok, was this the official fix or just a well-timed picture?
@bridgemindai They finally built the model everyone has been waiting for, posted the huge scores, then handed it to big companies first and told regular paying users to sit tight for a few days. That is a strange way to share the future.
@bridgemindai They showed the official scores, then pulled the page. Big companies get the real model first. Regular people get it later, and it will probably be the weaker one.