Thursday, December 12, 2019
YouTube makes its "anti-harassment" rules much more ambiguous and unpredictable in enforcement; new questions on how radicalization really happens
There is a lot of scuttlebutt in the past two days about the ambiguity of YouTube’s expanded “anti-harassment” policy, on its own creators blog.
Hoeg Law has the most expanded discussion of the problems with this post. Note the thumbnails of Stephen Crowder and Carlos Maza, whose dispute last spring YouTube handled so badly, leading to "Vox Adpocalypse".
The basic problem with all this, is the “compound interest” idea, and its retroactivity to previously posted content, which confronts the poster with the idea that a video acceptable today might be taken down later because of some larger group context.
I presume the same ideas could apply to Blogger, which this post is on.
One idea is that a pattern of many individually acceptable videos (or blog posts with embeds) about an otherwise not so famous individual (like a smaller Youtube channel) could turn out to draw unwanted attention to “them”. I often use “independent channels”, but I notice that there is a difference between how the large independent channels (Pakman is an example, of a “Youtuber” who can make a living off the channel and who hires employees – obviously, also, Pewdiepie) and some of the smaller personal ones that don’t ask for memberships or donations, feel about the practice.
This issue has come up before with the fact that I put the text of my three DADT books online, making a few people less obscure (because of search engines) than they might otherwise have remained.
Bostwiki has a summary video this morning (8 min) where he also mentions Gab’s sudden no pornography policy – because its payment processor otherwise would view it as “high risk”.
There is a big article in Wired (paywall), by Paris Martineau, Oct. 23, 2019, “Maybe it’s not YouTube’s algorithm that radicalizes people”, link. Read it now. It is based on a study at Penn State; remember there had been some controversy from another study at Cornell. I’ll come back to this later.