YouTube algorithm accidentally blocks “black v white” CHAC strategy

The overworked AI on YouTube could have misinterpreted a conversation about chess as racist language.

Last summer, a YouTuber who produced popular chess videos saw his channel blocked to include what the site called “harmful and dangerous” content.

YouTube did not explain why it had blocked Croatian chess player Antonio Radic, also known as “Agadmator”, but the service was restored 24 hours later.

Computer scientists at Carnegie Mellon suspect that Radic’s discussion of “black vs. black. white ”with a great master accidentally triggered AI filters on YouTube.

Running simulations with software designed to detect hate speech found that more than 80 percent of chess videos marked for hate speech did not have – but included terms such as “black,” “white,” “attack,” and ” threat”.

Researchers suggest that social media platforms incorporate the language of chess into their algorithms to prevent further confusion.

Scroll down for the video

The popular YouTube chess Radic, Antonio Radic, blocked his channel last summer for

The popular YouTube chess Radic, Antonio Radic, blocked his channel last summer for “harmful and dangerous” content. He believes that the AI ​​of the platform mistakenly marked him for discussing “black against white” in a chess conversation.

With more than a million subscribers, Agadmator is considered the most popular vertical chess on YouTube.

But on June 28, Radic’s channel was blocked after he posted a segment with Grand Master Hikaru Nakamura, a five-time champion and the youngest American to win the title of Grand Master.

YouTube did not give him a reason to block the channel.

In addition to human moderators, YouTube uses AI algorithms to identify banned content – but if the right examples are not provided to provide context, those algorithms can signal benign videos.

Researchers at Carnegie Mellon have tested two top speech classifiers, AI software that can be trained to detect hate speech.  Over 80 percent of the comments the programs reported had no racist language, but included chess terms such as

Researchers at Carnegie Mellon have tested two top speech classifiers, AI software that can be trained to detect hate speech. Over 80% of the comments marked by the programs did not have any racist language, but included chess terms such as “black”, “white”, “attack” and “threat”.

Radic’s canal was restored after 24 hours, prompting him to speculate that the use of the phrase “black against white” in Nakamura was to blame.

At the time, he was talking about the two opposing sides in a game of chess.

Ashiqur R. KhudaBukhsh, a computer scientist at the Carnegie Melon Institute of Language Technologies, suspected that Radic was right.

“We don’t know what tools YouTube uses, but if it relies on artificial intelligence to detect racist language, this type of accident can happen,” KhudaBukhsh said.

To test their theory, KhudaBukhsh and his research colleague Rupak Sarkar conducted tests on two state-of-the-art speech classifiers, AI software that can be trained to detect hate speech.

Radic's channel was blocked 24 hours after he posted this video, featuring a conversation with Grand Master Hikaru Nakamura

Radic’s channel was blocked 24 hours after he posted this video, featuring a conversation with Grand Master Hikaru Nakamura

Using the software for more than 680,000 comments taken from five popular YouTube chess channels, they found that 82% of the comments reported in a sample set did not include any obvious racist language or hate speech.

Words like “black”, “white”, “attack” and “threat” seem to have triggered the filters, KhudaBukhsh and Sarkar said in a presentation this month at the annual Association for the Advancement of AI conference.

The accuracy of the software depends on the examples given, KhudaBukhsh said, and the training data sets for YouTube classifiers “probably include some examples of discussions about chess, leading to misclassification.”

Radić, 33, started his YouTube channel in 2017 and has more than a million subscribers.  His most popular video, a review of a 1962 game, garnered over 5.5 million views

Radić, 33, started his YouTube channel in 2017 and has more than a million subscribers. His most popular video, a review of a 1962 game, garnered over 5.5 million views

If someone as well known as Radic is mistakenly blocked, he added, “it could happen quietly to many other people who are not so well known.”

YouTube declined to say what caused Radic’s video to be flagged, but told Mail Online: “When we are notified that a video has been deleted incorrectly, we act quickly to restore it.”

“We also offer users who have the opportunity to challenge deletions and we will review the content,” said a representative. “Agadmator appealed for removal and I quickly restarted the video.”

Radić, 33, started his YouTube channel in 2017 and, in one year, his income exceeded his daily job as a wedding videographer.

“I’ve always liked chess, but I live in a small town and there weren’t too many people to talk to. [it]he told ESPN last year. “So starting a YouTube channel made sense.”

His most popular video, a review of a 1962 match between Rashid Nezhmetdinov and Oleg Chernikov, has so far received 5.5 million views.

The COVID blockades have sparked renewed interest in chess: since March 2020, the Chess.com server and social network have added about 2 million new members a month since the pandemic began, Annenberg Media reported.

The game of kings also benefited from the popularity of “The Queen’s Gambit”, an acclaimed mini-series about a troubled female chess master, released on Netflix in October.

.Source