Is there a way to cut the archive down to a manageable yet relevant subset based on one's interests? Say, "I have 500GB free space and I care about mathematics, what do I take from the archive"? Because 50000TB is too much for crowdsourcing, particularly if it's supposed to be crowd-served rather than just crowd-hoarded. I would guess a high percentage of that is video, which perhaps should never have been part of the archive to begin with due to its signal-to-noise rate.