This is a fascinating tool, I estimated that the corpus of the factual web was between 1 and 10 TB when I last played around with BigQuery using domain names which had low amounts of click bait. Seeing these search results I suspect my estimate was off by a couple orders of magnitude.
Although a search for "Fractional Reserve Banking" shows that some further ranking improvements can be made to exclude unrelated results, and potentially penalize old conspiracy sites.
Although a search for "Fractional Reserve Banking" shows that some further ranking improvements can be made to exclude unrelated results, and potentially penalize old conspiracy sites.
https://search.marginalia.nu/search?query=fractional+reserve...