The YouTube AI had blocked YouTuber Antonio Radic's channel for harmful content last summer. It has become a talking point within the chess community recently.
The YouTuber is popular for uploading chess videos on his channel but saw it get banned by the platform. The Croatian's account got restored 24 hours later, but the reason for the block remained shrouded in mystery. It continues to baffle the wedding videographer turned YouTuber.
According to Radic, a discussion of "black vs. white" with a grandmaster accidentally triggered YouTube's AI filters. This might have been enhanced to try and put an end to racial slurs online.
Several platforms, including YouTube and Twitch, have updated their policies to bring an end to harassment online.
YouTube uses AI algorithm to block sensitive content
While this is a sensible practice, the algorithm could use some refinement.
The algorithm caught words like black, white, threat, and attack. However, it failed to capture the context.
KhudaBukhsh and fellow researcher Rupak Sarkar at Carnegie Melon's Language Technologies Institute, recently used the AI software. They came to this conclusion.
"We don't know what tools YouTube uses, but if they rely on artificial intelligence to detect racist language, this kind of accident can happen."
To their surprise, the AI captured over 680,000 comments from popular YouTube chess channels. However, over 82 percent of the comments captured did not point towards hate speech.
This was an unfortunate incident for Radic and the chess community. Online chess has seen a sharp rise recently.
Streamers, including xQc even, have confessed to taking chess lessons to sharpen their skills.