TikTok is changing the recommendation algorithm for its 'For You' feed to prevent users from seeing too many similar videos. YouTube, Facebook and other outlets with crowdsourced media have faced a ton of criticism over the years for their recommendation algorithms that critics argue lead people down a never-ending rabbit hole. According to observers, these recommendations inherently encourage tribalism by cutting out opposing view points, thus reducing the chance for a healthy debate.
TikTok has emerged as one of the preeminent social media apps over the past few years, and was reportedly the most downloaded and highest-grossing app of 2021. However, even as it continues to top the charts in most regions around the world, it has had a few hiccups over the years. Primarily among them is a total ban in India, ostensibly over privacy concerns over its Chinese ownership. It was also threatened with a similar ban by the Trump administration, although that threat never quite materialized.
In a press release on Thursday, TikTok said that it is changing its 'For You' recommendation algorithm, which could "inadvertently reinforce a negative personal experience for some viewers." According to the company, it is implementing the change because the current algorithm does not provide users with a "diverse discovery experience" that TikTok aims to create. With the new algorithm in place, the recommendations will still offer plenty of content based on the user's expressed preference, but will intersperse it with different categories of videos to enable users to explore a more diversified catalog of content.
TikTok Aiming For More Diversified Recommendations
Explaining the benefits of the new algorithm, TikTok said that the diversified recommendations will help people move away from potentially dangerous and/or harmful chains of thought. It would do that by providing viewers with a breadth of content across several categories of videos and improving user experience on the app. According to the post, the 'dangerous' categories include videos on extreme dieting, breakups, loneliness, etc. The company said it is working with doctors, psychologists, AI ethics experts, the TikTok community, and members of its Content Advisory Council while fine-tuning its new algorithm to best serve its audience.
TikTok’s personalized recommendations algorithm has faced harsh criticism from activists for a number of reasons. In one especially concerning case, a recent investigation by The Wall Street Journal showed how the app was pushing content related to sex and drugs to children. Moving forward, TikTok says it is building a slew of safety features within the app to limit the type of content people do not want to see in their feed. As part of the plan, the company says users will be able to select words or hashtags associated with content that they'd prefer not to show up in their 'For You' recommendations page. The upcoming feature would help vegetarians avoid meat recipes or people working on self-esteem issues see less beauty tutorials, said TikTok.
Source: TikTok, The Wall Street Journal