3 ms·
What she is asking will probably be even more creepy. Currently these platforms are basically associating words to your user, which advertisers then select fro
by woolvalley 8y ago
What she is asking will probably be even more creepy.
Currently these platforms are basically associating words to your user, which advertisers then select from a pull down.
Past some specialized categories created out of advertiser demand - such as political leaning, location, age and gender - there isn't much reasoning behind it.
First her user id got the pregnancy keywords associated, now she also has sad / still born words associated. The system isn't smart enough to reason that the two tags should be mutually exclusive in a culturally sensitive way, just like it doesn't with millions of other pairs around the world within 100s of other cultures & languages around the world.
But if it was, then how much worse could it get in other ways?
- danso 8y agoThe author describes manually removing herself from being targeted, e.g. clicking "I don't want to see this ad" and "It's not relevant to me". Are we supposed to be content with platforms that are unable to incorporate such manual feedback? I mean, we (the average user on HN) get pretty irate when software ignores or reverts config options we've previously set. I don't see how this is much different.