Google built their index by externalizing the human cost of curation, aka Webrings. Now all the webrings are dead, because humans are easily tempted by lazy search into not maintaining them. I wish they weren’t. They were a better form of curation than anything since.
I wonder how realistic a hybrid of human and algorithmic curation would be.
Obviously the internet has an enormously long tail, but if a company could ‘curate’/rank the top 10,000 most popular sites, that might still be useful.
