>
founded on ...
the endorsements of "an angry, misinformed mob"I started grad school at Brown in 1997 and I remember a talk there, by someone from Google or connected to it, about PageRank. PageRank was still new, Google was still in beta. Free swag from dotcoms was literally growing on trees. But I digress.
I remember that the narrative about PageRank at the time was not about popularity, but about expertise. I really remember that the presenter brought up a possible threat to validity—what about gaming the system?—and pointed out that if you wanted to persuade the system that you were an expert on a thing, you could get lots of people who talk about the thing a lot to point a link at your site. BUT, he says: that shows the system works! If you can get that many people who are at least mini-experts on the thing to point to you, the only way to do that would be if you, too, were persuasively an expert on the thing.
(These were... naïve times in many ways.)
Twenty years ago this was highly persuasive. It was an un-game-able system, because "gaming the system" meant the system changed you. I can be pretty cynical about a lot of things, including and especially Google, but I really do think that even Google itself believed this. The company was founded on an algorithm that rewarded expertise.
It turned out that the algorithm also rewarded angry, misinformed mobs, so cmckn isn't exactly wrong here. But I think it's important to be clear that to the extent that Google's algorithms have always done this, they a) weren't meant to, and b) actually didn't in the very earliest days, because the WWW link structure really did, in the 90s, work like they thought it did. (Then the measure became the target, etc etc)