It does seem like a lot of the valuations and infrastructure investments for these companies only make sense if each one assumes they will be the first and only one to invent superintelligence and that it will largely replace all knowledge work.

In which case the global economy is dead and there’s no one to buy their work?