YouTube Is Asking for Your Help in Identifying Artificial Intelligence Flaws on Its Platform.
If you’ve watched any videos online in the last year or two, you’ve likely encountered content generated by artificial intelligence ( even if you didn’t realize it ). Some examples are compelling, but much is obvious. For example, I don’t think anyone realized that those dramatic videos featuring people made of fruit were painstakingly animated by hand. This type of artificially generated content has been affectionately (or not so affectionately) dubbed “artificial crap,” which describes it quite accurately: it’s meaningless nonsense, created with minimal effort and investment by some AI generator, that makes real money by going viral on social media.
While all platforms featuring short videos are inevitably already saturated with AI-generated garbage, YouTube has been particularly hard hit. YouTube Shorts videos can be virtually unwatchable, depending on the algorithm. (Try watching in incognito mode to see what the platform offers users without bias; it’s just insane.) Worse, YouTube’s algorithm shows these videos to children . Short videos are bad enough for kids for many reasons, but AI-generated garbage takes it to a whole new level.
“Didn’t this strike you as poor AI work?”
At least there’s some good news: as Dexerto reports , starting this month, YouTube appears to be asking users to help identify AI-generated content. The company isn’t mincing words or beating around the bush. If you’re asked to do something, a pop-up will appear asking, “Does this look like AI hackwork?” According to a screenshot from this Reddit user , you’ll have the option to respond with one of the following: “Not at all,” “A little,” “Moderately,” “Very much,” or “A lot.”
Why is YouTube asking for your help in identifying AI-related bugs?
This might seem like a good sign, and I certainly hope it is. In an ideal world, YouTube would take these findings into account and remove the most objectionable videos. This is consistent with some of the company’s past actions: YouTube even removed popular AI-powered channels under the guise of “reducing the spread of low-quality, repetitive content.”
But Dexerto offers a different, less optimistic theory . Perhaps the reason YouTube offers so many options in the pop-up window is to more accurately understand how viewers interpret these AI-generated videos. YouTube will be able to distinguish clearly AI-generated videos from those in the “uncanny valley.” Using this data, they can train their own AI models to create content that won’t be perceived as low-quality by most viewers. If you watch Shorts and the videos seem real, how likely are you to question their authenticity?
I hope this isn’t true. While my faith in tech companies remains low, I’m impressed by YouTube’s recent stance on low-quality AI-generated content, and these warning pop-ups seem to be another step in the right direction. Hopefully, YouTube will continue to promote higher-quality human-generated content and eliminate low-quality AI-generated content.