blueeowyn: (Default)
posted by [personal profile] blueeowyn at 09:04pm on 2022-10-23
"Actually I'm not sure even THIS approach works since even tech applied for pure aesthetics can do harm (imagine an app that makes cute cartoon selfies, but because of a limited training set breaks on people of color). But if you're gonna use ML at all u have to outline its limits"

You mean like has already happened? Along with a bunch of other bad results for people of color (in prticular but others as well) in the results of algorithms based on biased and/or incomplete data?

Links

January

SunMonTueWedThuFriSat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24
 
25
 
26
 
27
 
28
 
29
 
30
 
31