John Regehrsometimes I think about how there was this netflix prize and then at the same time every streaming service I've tried will (at least sometimes) fail to recommend a new season of a show when I've watched all previous seasons of that same show
just this morning I noticed that the broken lizard crew did a tv show ("Tacoma FD") back in 2019 and zero streaming services thought I might enjoy it despite the fact that I've watched every broken lizard movie. so dumb.
Jan de Muijnck-Hughes
David R. MacIver@regehr I have a couple of heterodox recommendation system opinions, and one of them is your recommendation system obviously needs a whole bunch of human written rules, and without them your fancy machine learning system will have worse UX than a Python program I could write in a day.
@regehr To be clear I think the fancy recommendation algorithms are good too (especially for anything with an actually large catalogue), but there's just a lot of low hanging fruit they totally fail to pick.
@DRMacIver for sure
Fabian Giesen@regehr just impossible for the algorithm to tell if you're interested or not
@rygorous simply unknowable if I want to watch season 6 given that I binged 1-5. the universe is just like that sometimes.
@regehr to be fair, it actually be like that sometimes
I bailed on Game of Thrones after season 7 (because I thought that was a train wreck already) and that seems to have been absolutely the right choice