I could imagine a safe the size of a gun safe completely containing a nano-fiche machine and all of the information on Wikipedia today. This isn’t a hard problem.
The hard problem is meting through the information. Fast, contextual search is probably the most important development we have today and you can’t really replace it with anything but tons of compute and digitization. Big ass physical libraries don’t solve this either.
Wikipedia doesn't nearly store as much information as you might think. It's basically the intro 101 on all subjects, but it barely scratches the surface of serious scientific fields. They have a million articles. A large science publisher has more than that for its journals, and there are many large publishers.
> The hard problem is meting through the information.
It doesn’t have to be an instanteous thing. Just think of how much time schollars spend on deciphering papyrus fragments. If we would have a nanofiche of the contents of the library of Alexandria multiple departments would spend multiple lifetimes reading and contextualising it.
The hard problem is meting through the information. Fast, contextual search is probably the most important development we have today and you can’t really replace it with anything but tons of compute and digitization. Big ass physical libraries don’t solve this either.