Why do you need huge economies of scale though? For 90%? of queries a local model will be good enough?
You'll still need a cloud AI subscription for the remaining 10%, so you might as well enjoy the superior performance. If you're cost sensitive, models like Deepseek have you covered.
For inference alone they get far better utilization of the hardware because they can keep it fully loaded with maximum concurrency. Other than that they get to collect a lot of data for training.
Well beside for roleplaying, local models are a bit useless, especially the ones running on end user class hardware.
Maybe when chip would infer at 10-20x the speed of a 5090 RTX, otherwise it's really..bad.