Hope you're right. The best possible outcome out of this is higher investments into open weight model development. I'm already looking for local inference options. Claude was good while it lasted.