Looks like she took down the post in question (it was public). My argument boils down to this - people familiar with the state of the art will tell you that it is very likely that this kind of system is automated or based on mass importation of third party data. In both these cases it's very likely that a small error can totally screw things up, or that somebody was gaming the automated system.
Also, I rather suspected that the LGBT angle was a red herring - that while yes, a number of LGBT titles were blacklisted, that it was rather likely that that was just a pattern picked out of a wider problem because the real pattern is harder to see (e.g., we don't know what products get updated by amazon when, so we can't look at all those).
According to the original Mark Probst post, it seems that it was in fact a bad data push, that affected way more than just a handful of LGBT titles.
Comments 2
Reply
Also, I rather suspected that the LGBT angle was a red herring - that while yes, a number of LGBT titles were blacklisted, that it was rather likely that that was just a pattern picked out of a wider problem because the real pattern is harder to see (e.g., we don't know what products get updated by amazon when, so we can't look at all those).
According to the original Mark Probst post, it seems that it was in fact a bad data push, that affected way more than just a handful of LGBT titles.
Reply
Leave a comment