It started with a fairly typical exchange for Facebook in 2020. Nicole Goetz, who runs a small farm in southern Michigan with her husband called Frontière Farm House posted some charming pictures of herself on the farm — one with a goat and several with a few sheep. A friend soothed by the cuteness commented she was “Struggling with the big sad and appreciate the goat pictures”. To clarify, Goetz's reply included “all except the all-white dude are sheep”. Somewhere deep in the heart of a poorly-written algorithm, Facebook’s most useless of bots instantly detected white dude are sheep — and Goetz was automatically given a three-day ban.
Facebook’s “Community Standards” are an utter failure on multiple levels. To the average user, it can be a confusing experience because many make the knee-jerk assumption that someone has reported them, when in fact, this “feature” (that is more of an intentional bug from a basic UX literacy perspective) is automatic. It’s the kind of batshit bureaucratic interaction with a large corporation so surreally absurd, it belongs in science fiction films — Brazil’s “Central Services” comes to mind. Facebook clearly came up with a list of phrases, and if you type one of those phrases regardless of context, you get the boot.
In Goetz’s case, the primitive simplicity of the filter demonstrates its failure to understand how language actually works — even on a very basic level. For instance, a naive person might assume that when Facebook came up with that particular phrase, someone would have anticipated discussions of literal farm animals — an assumption that Facebook is much smarter than it is.
There is also no process for correction as far as anyone can tell, so your account will come with a warning with increasingly harsh penalties if it happens again — even if Facebook is squarely at fault. It claims it has an internal “review team” which is clearly a sleight of hand term for “bot” since there was no way to correct it and nobody reached out to Goetz with an apology. A less lazy company could still have the same kind of program, but with built-in fail-safes. Surely, if Facebook is sophisticated enough to scan text, delete the offending comment, and auto-ban the user in one fell swoop, wouldn’t a saner and less punitive way of implementing an otherwise broken tool be to simply auto-delete the comment and have the user try again? Facebook’s idea of Clippy: “Would you like to rewrite this so that it does not violate our community standards?” BOOM. Way better. Pay me, Facebook.
The company gets away with it for the same reason it gets away with being dangerous with how false information spreads in a politically contentious era, and for the same reason it gets away with being a subpar product in general: Facebook has very little competition, despite the existence of multiple social media platforms — all of which are very different from each other. I would LOVE to quit Facebook, but 100% of my friends live in a different state than me (or country). No company has a chance of dethroning them unless it is powerful enough to attract enough users to reach a critical mass, which is not an easy task since that requires, at minimum, data storage solutions for 2.7 billion potential users and all the cat videos they’ll bring with them.
Even worse, Facebook’s specific List of Things You Can’t Say isn’t published anywhere, so when you receive their nauseatingly paternalistic message that you have violated their “community standards”, you are breaking internal, arbitrary rules that have not been made clear to its users first — all on the adorably simplistic pretense that you ought to just know how we interpret our community standards. To illustrate how fucking subjective this is, in fact, their idea of “community standards” routinely violate mine. Even when the tool works as intended, its existence fosters the kind of false equivalency and anti-intellectualism that already permeates American Internet Culture — that Facebook has made worse through systemic validation of a childish conception of ethics.
There has been a nauseating uptick in the number of white cishet men who abuse terms like racism and sexism with reductionist literalism — men who flaunt the letter of the dictionary but ignore its deeper spirit and obvious context, who feel comfortable claiming it is racist and sexist to mock them — and otherwise believe “reverse racism” or “reverse sexism” are real things. The problem with that being: They are fucking not. If women start murdering strangers because they are men or pass legislation that strips men of their rights and bodily autonomy, and black people start murdering strangers en masse because they are white, THOSE THINGS would be bigotry and oppression. Facebook, in calling it “hate speech” when you say something flippant like “men are trash” validates the beast. We live in a world in which casually racist comments about an oppressed group have cruel, real consequences in the real world — including and especially the deaths of minorities by law enforcement, and violence against women (cis and trans). Whereas the dominant class with the power is frankly 100% deserving of a little “punching up” — especially in response to specific behaviors that make things even worse for the rest of us.
My personal account is currently branded with a warning after two apparent infractions due to the irreverent spitfire I tend to be online in casual conversation (either despite my best efforts or because of them — still working that out). If there is a third instance of not anticipating how Facebook’s robots have been programmed to feel about my use of language, it will be a full week vacation for me (and honestly, forced time away from the hellsite would be something of a blessing). Early in the pandemic when face masks were still scarce, I posted a maddening article about men who refused to wear donated cloth masks, sewn by the kind of quilters who had yards of extra floral calico laying around — for being too “feminine”. These men were risking their (and others) lives and fucking up a free lunch to boot — all for the vanity of what they incorrectly felt isn’t “masculine” which is, of course, among the least masculine things I’ve ever heard. I posted it along with a flippant “Men are dumb” and got the boot. A few months later and just before the election, I replied to one of my Canadian friends in a thread explaining the election shitshow at large. “I dunno”, I said in exhausted futility, “Americans as a collective are stupid.” followed by the shrug emoji. BOOM — three-day ban. As an American with a longtime Canadian partner, it is my god-given right to admit in my own space that my own country is stupid as a collective. I wouldn’t even be friends with anyone who would be offended by that truism.
Finally, the tool is a bit like jailing the servants of the gentry while sparing their masters who would be guillotined. Facebook routinely dusts a room with a swiffer, while ignoring the elephant in the room that continues to leave large piles of shit. It ignores legitimate complaints about profiles that absolutely do (or at least should) violate their community standards — often to dangerous levels. How many of us have seen the worst kind of comments from the worst kind of people pass Facebook’s “Community Standards” with flying colors simply because Facebook can’t automate an inexpensive way to interpret what is actually being said? My partner wrote a popular piece of software that has been our livelihood for the past two years. Entire groups exist on Facebook with the advertised purpose of sharing cracked software. He has reported these groups, but outright organized IP theft doesn’t violate their community standards.
Until Facebook has better tools, better internal ethics, and a better User Experience with which to implement them, their “community standards” make the dumpster-fire of American internet culture much, MUCH worse for the most vulnerable people.