In RetroTech: NetLingo Blog

YouTube was under fire again in early March, when a video blogger named Matt Watson detailed how pedophiles can enter a “wormhole” of YouTube videos to see “footage of children in sexually suggestive positions” according to CNET.com. They can then jump from video to video, helped by YouTube’s recommendation engine, and fill them with lewd comments. Oh yeah, there's an old NetLingo word for that: flame bait.

In response, brands such as Disney, AT&T, and Epic Games pulled their ads from YouTube, and the company responded by banning more than 400 accounts. Unfortunately, it’s not the first time that Google-owned YouTube has had this kind of child-safety flare-up. In 2017 alone, disturbing knockoffs appeared on the YouTube Kids platform that depicted Disney and Marvel characters in troubling ways; then sexually explicit comments appeared under videos of kids’ gymnastics. “In response to those scandals, CEO Susan Wojcicki overhauled YouTube’s safety guidelines.” Yet two years later, the same problems keep cropping up.

It’s not just the comments that are problematic for YouTube, said The New York Times. The platform has been reckoning with the vast troves of disinformation and extreme content it harbors, such as conspiracy videos and hoaxes that are popular with millions of viewers. Here, again, the recommendation engine is part of the problem: It sends viewers of misinformation to similar videos with more misinformation.

Conspiracy theories and viral hoaxes top the list” of recommendations for viewers of many popular channels. Young people repeatedly battered by these recommendations often start to reject mainstream sources. To fix this, YouTube needs to recognize how deep these problems run and realize that any successful effort may look less like a simple algorithm tweak, and more like deprogramming a generation.