Following the mass school shooting which killed 17 people and wounded over a dozen others in Parkland, Florida this month, YouTube launched a campaign to use some of its 10,000 new moderators to somewhat thin out the ranks of the conspiracy peddlers and far-right nuts which have become rampant across the site. Though other tech companies have been pressured into such action over the past year, the matter was particularly pressing for YouTube – which yet again promoted virulent conspiracy theories speculating that the shootings were a hoax or that the survivors were “crisis actors.”
As Bloomberg reported the site has now admitted it “mistakenly removed several videos and some channels from right-wing, pro-gun video producers and outlets,” raising the prospect it’s caving to the backlash:
Here we go. YouTube JUST REMOVED one of my @cbts_stream videos & DISABLED my LIVE STREAM for 90 days #QAnon #Qanon8chan CENSORSHIP or LEFTISTS attacking @realDonaldTrump WE NEED TO PROTEST YouTube @YouTube @Google RESTRICTING CONSERVATIVES ON INTERNET #InternetBillOfRights
— Jerome Corsi (@jerome_corsi) February 27, 2018
According to Bloomberg, a YouTube representative said the mistake could be attributed to “newer members” of its team being brought on as part of its push to hire 10,000 new content moderators, but it repeatedly declined to say what content in particular was taken down in error. As the Outline noted, the only content that’s reappeared thus yet are two firearms videos with no obvious rule violations from Military Arms Channel – though the usual whackos like far-right personality Mark Dice are going wild on Twitter, claiming the admission of any mistake at all constitutes victory. Hopefully they’re wrong and people like Corsi, whose prior best hits include a steadfast belief Barack Obama is secretly gay and also some kind of Muslim, aren’t coming back.
In any case, this underscores the same dilemma that’s been dogging YouTube and other tech giants for a while. They fine tuned algorithms to replace human supervision, and now whenever one of problems those design decisions caused spiral out of control, the result is institutional paralysis even when users do bother to report the content and human moderators step in (who as NYMag noted, are usually foreign contractors with workloads too high to allow “the kind of thoughtfulness and cultural nuance necessary for good moderation”).
There doesn’t appear to be any coherent philosophy underscoring the company’s approach to the rules beyond the belief platforms aren’t truly accountable for user-generated content, which means they’re constantly playing catch-up and any progress tends to be frustratingly tenuous.
YouTube could probably break this pattern, but social media megaplatforms built to satisfy demands for scale and push the loudest content to the biggest audience possible tend to have a bad track record at consistently following through. One other group to fall afoul of YouTube’s moderators lately was neo-Nazi organisation Atomwaffen, though it took tons of bad press before the company could be bothered to take action.