![The change lets users reset their For You feed to see things as a new user.](https://static.standard.co.uk/2023/03/16/11/TikTok%20algorithm%20refresh.jpg?crop=8:5,smart&quality=75&auto=webp&width=1024)
But now the company is giving users a chance to reset this, should they decide that suggestions presented in the ‘For You’ feed “no longer feel like they’re for you” or fail to “provide enough topical variety”.
It appears to be at least a partial algorithmic reset, with the option promising to let established TikTok users see the For You feed as if they were brand new, with the app learning their tastes all over again.
While the app cautions users that the refresh can’t be undone, your following feed, profile, and inbox will remain untouched.
To access the feature, users need to open the settings and privacy section of TikTok, then scroll down to “Content preferences” under “Content & Activity”. In that sub-menu, they’ll find an option labelled “Refresh your For You feed”.
In the blog post introducing the new feature, Sandeep Grover and Mabel Wang — the heads of Trust & Safety and Content & Creator Product respectively — also provide an update on what TikTok is doing to prevent repetitive recommendations, especially when it comes to more controversial content.
“An inherent challenge of any recommendation system is ensuring the breadth of content surfaced to a viewer isn’t too narrow or too repetitive,” they write. “We’re intently focused on this challenge, and work to design a system that intersperses a variety of topics.”
This is especially pertinent for “content that doesn’t violate our policies, but may impact the viewing experience if viewed repeatedly.” That may sound a bit like the vague “legal but harmful” definition that was removed from the Online Safety Bill, but the authors say they’re especially focused on content with the themes of “sadness, extreme exercise or dieting, or that’s sexually suggestive”.
But the post defends TikTok’s right to include this kind of content, and actually highlights the benefits of such borderline posts, when shown in moderation.
“We routinely hear from experts that closing the door on this expression can increase feelings of isolation and stigmatisation, and that enabling people to see how others cope with difficult emotions can be beneficial, especially for teens,” the post reads.
In other words, as long as a video doesn’t violate the company’s policies — such as content that “promotes or glorifies self-injury” — this kind of video is eligible for recommendation, albeit in a more controlled manner.
“If multiple videos with these themes are identified, they will be substituted with videos about other topics to reduce the frequency of these recommendations and create a more diverse discovery experience.” The work “is ongoing”, the post explains, with “over 15 updates” in the last year to make it work better.
The update comes at a particularly tricky point in TikTok’s short history. The app faces criticism on at least two fronts, both from those concerned about the power of the algorithm to spread harmful content and others suspicious of the possible involvement of the Chinese government.
Indeed, even as TikTok is announcing the ability to reset the algorithm and reiterating its concern for user safety, the UK is expected to follow in the USA’s footsteps and introduce a ban on the app’s usage on government phones.