I think you need to go a step further. If AI maximalism turns out to be correct, does Anthropic control AI? Does the AI care that Anthropic as a company is supposed to be returning value to shareholders, or will it have its own agenda?