YouTube Returns Human Moderators To Minimize Mistakes Of AI Systems

Ilustrasi YouTube (Noordword Themes / Unsplash)

JAKARTA - During the COVID-19 pandemic, many people were forced to work from home, including YouTube. Along with that YouTube relies on artificial intelligence technology to replace the presence of people in the office.

But unfortunately, the use of artificial intelligence (AI) to monitor YouTube content is not going well. Because a lot of content has been deleted, due to the wrong moderation of the AI system.

Therefore, YouTube will again assign human moderators to oversee video content on its platform. Where previously, the AI system only moderated content less than 25 percent compared to human moderators.

"One of the decisions we made (at the start of the pandemic) was related to machines not being as precise as humans, we would be wrong to make sure that our users are protected, even though that might result in a slightly higher number of video stops," said YouTube's Chief Product, Neal Mohan, as quoted by Mashable, Monday, September 21.

Meanwhile, according to the Financial Report, at the end of August to the previous three months, there were 11.4 million videos that were removed from the YouTube platform by moderation of the AI system. This is the highest number of videos removed from YouTube over the past three months, since the service launched in 2005.

The video streaming service admits however, that some non-violating videos were removed by its AI moderators. However, they have canceled their decision to remove 160,000 videos.

Now that the moderation system is returning to humans, the company hopes to make more accurate decisions about what content will appear on YouTube. Although it is not yet known, whether YouTube employees will return to the office or continue working from home.