> For example, people prefer dating advice from humans over advice from algorithms The humans are less-likely to become systematically enshittified giving you

> For example, people prefer dating advice from humans over advice from algorithms The humans are less-likely to become systematically enshittified giving you bad advice for profit. > even when the [dating advice] algorithms perform better Whoah, hold the heck up! When did that ever get demonstrated? I think I would have heard of it before. Following their link, it goes to study that AFAICT merely showed a preference for different banner-ads, where ads promising human advice were more-likely to be clicked.