The internet has had some interesting repercussions for the idea that all knowledge is worth saving. We're producing data many orders of magnitude than ever bef

The internet has had some interesting repercussions for the idea that all knowledge is worth saving. We're producing data many orders of magnitude than ever before, and it's easier to save it than ever before, and 99% of it is completely useless and uninteresting outside of its original context, even for hypothetical future historians. As you say, the goal shouldn't be to save everything just in case, but to weed out the irretrievably useless and carefully index the rest.