i think inference is the thing, that also fast inference, so enterprises can just host their own and run, ig vercel do it, many more would. but zs it thinks toooo much idk how fast we can make it.
basically
disclosed public ai
undisclosed private ai
this only works, any thing public should be disclosed you can do anything when you are working locally as you know but it should be explicitly discloses as soon as it goes public.