swan_tower: (Default)
[personal profile] swan_tower
I don't know if this theory is correct, but I frankly consider it a more likely explanation than "Amazon's executives got lobotomized and decided to institute a PR disaster over Easter weekend" --

http://tehdely.livejournal.com/88823.html

Short form is, some organization (no idea who) may have mobilized to flag certain content on Amazon until an automated feature kicks in to remove it from the rankings. It would explain the bizarre selection of items hit, and also why the initial response of Amazon flacks was to say they're de-ranking "adult" content -- if this theory's right, that's exactly what they're doing, but they hadn't yet realized somebody was gaming the system.

And Easter weekend is a sadly unsurprising time to see positive LGBT* content hit in such fashion.

So. We'll see what develops. But this makes a lot more sense to me than this being some kind of actual corporate decision on Amazon's part.


*Or whatever form of abbrevation you prefer. Man, I hadn't realized just how many permutations of initials that cluster of ideas had until I started seeing posts about this incident all over my flist. What the heck is the (QQI) add-on?

Date: 2009-04-13 01:08 pm (UTC)
From: [identity profile] xmurphyjacobsx.livejournal.com
I'm wondering if that Amazon rep was spinning it out the ass, as sometimes happens in customer service -- if the trolling thing is correct (and I'm not saying it is -- plausible, not proven) then anyone consistently flagging a book as objectionable would set off a trigger that would remove the book from ranking as "objectionable" -- which tends to translate as "adult". That kind of customer flagging/triggering COULD be described as Amazon policy. I can see how the two fit together.

The problem comes in when you set up robots to do a human's job -- flagging should lead to a HUMAN (preferably a few, with brains) looking at the situation and deciding if the flag is legit or just a troll.

I'm not suggesting this, but it occurs to me that if this theory is correct, a similar 'attack' could be committed in reverse -- that is, having many people flag some set of anti-alternative sexuality books and seeing if they, too, fall out of ranking. I do NOT suggest this sort of sandbox revenge tactic as an action, just as an imaginary experiment to test out the 'trolling' theory.

Date: 2009-04-13 06:16 pm (UTC)
From: [identity profile] swan-tower.livejournal.com
I think people are running that experiment, actually. Because if you searched for "homosexuality" after this started, you did get some hits . . . all of them books about curing gays or what-have-you. So there was talk of flagging those "gay erotica" or whatever might be causing this, and seeing if they went away.

But yes, automating stuff hath many dangers. As we've seen elsewhere, even if it turns out not to be the root cause here.

Profile

swan_tower: (Default)
swan_tower

May 2025

S M T W T F S
     123
45 678910
11121314151617
1819 2021222324
25262728293031

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Jun. 11th, 2025 07:20 am
Powered by Dreamwidth Studios