Hacker News new | ask | show | jobs
by loup-vaillant 5474 days ago
The end user is not fine. He is more likely to see results that he actually agree with. See, the original confirmation bias will cause you to seek opinions you agree with more often than others. The search engine will then conclude that you are more interested in the kind of sources those opinions come from. That would be true, by the way, but then comes a point when a quick glance at your search engine result will show you more of what you agree with, and less of what you disagree with.

Now go use that as an estimation of popularity and veracity. I bet many people do, without knowing the result is strongly biased by their own prior behaviour.

Search engine, as the sole entry point of the web, do bias it. Page Rank for instance, could trigger a feedback loop: if a site is more prominent in searches, it will get more links. That will get them more search prominence, and feedback and foom.

Now is the popular bias better than the personal bias? I think it is. One would at least get to be exposed to other's opinions, instead of just his own.

If you just care about the economy of the web, in the sense of selling, advertising, promoting, buying… then of course the personal bias is currently best. That's the most efficient way to milk the tear$ out of eyeballs. The easiest way to reward the brains behind those eyeballs. When it's all about money, there is absolutely no problem with the method. But I have other values besides money. A very important one is respecting curiosity and search for truth. The personal bias doesn't.