Here is how it works and why it has attracted extra discussion than abilities frail by its opponents a lot like Instagram, Google, YouTube and Snapchat.
Algorithms
The algorithms are deemed core to ByteDance’s overall operations, and ByteDance would somewhat shut down the app than sell it, Reuters reported citing sources.
China made adjustments to its export laws in 2020 that give it approval rights over any export of algorithms and present codes, including a layer of complexity to any effort to sell the app.
Lecturers and former company employees said that it’s not just the algorithms, but also how they work with the short video format, that has made TikTok so successful globally.
TikTok is gearing up for a legal battle against a US law that could force the social media platform to cut ties with its China-based parent company or face a ban. Source: AAP / Mariam Zuhaib
Or not it’s miles the app too
Before the emergence of TikTok, many had believed that connecting a person’s social connections was the key to a successful social media app, given the popularity of Meta’s Facebook and Instagram.
Nonetheless TikTok confirmed that an algorithm, driven by an understanding of a particular person’s interests, could possibly be more significant.
In contrast to building their algorithm on “social graph” like Meta has, TikTok executives including CEO Shou Zi Chew have mentioned that their algorithm is based on “interest signals”.
While competitors have similar interest-based algorithms, TikTok is able to turbocharge the algorithm’s effectiveness with the short video format, mentioned Catalina Goanta, an associate professor at Utrecht University.
“Their recommender system is very broad. Nonetheless what in fact distinguishes TikTok as an app is the format and the content,” she mentioned.
The short video format allows TikTok’s algorithm to become much more dynamic and even capable of tracking changes in users’ preferences and interests in real time, going as granular as what a particular person might like during a specific timeframe of the day.
TikTok CEO calls on users to react to US ban
Rapidly records sequence
In addition, the short video format allows TikTok to learn about user preferences at a much faster rate, said Jason Fung, former head of TikTok’s gaming unit.
“Because it’s in short-form format, it’s short video, you can gather data about a person’s preference much sooner than YouTube, where the average video is just under 10 minutes long,” he said.
“Have faith you are collecting data about a particular person on average every 10 minutes versus every couple seconds.”
The positioning of TikTok as an app constructed for cellular devices from the beginning also gave it a bonus over rival platforms that had to adapt their interfaces from laptop displays.
TikTok’s early entry into the short video market also gave the firm a substantial first-mover advantage.
Instagram didn’t open Reels except 2020 whereas YouTube launched Shorts in 2021, each and every of which inch on the relieve of TikTok in years of data and product style abilities.
Enables exploration
TikTok also often recommends content that falls outside of a user’s interests, which the company’s management has repeatedly stated is intended to enhance the user experience.
A survey, which researchers from the US and Germany published last month, found that TikTok’s algorithm “exploits user interests in 30 percent to 50 percent of recommended videos” after analyzing data from 347 TikTok users and 5 automated bots.
“This finding implies that the TikTok algorithm opts to recommend a structured sequence of discovery videos in an attempt to either better infer user interests or maximize user engagement by recommending many videos that may fall outside of the user’s (known) interests,” the researchers wrote in the paper titled “TikTok and the Art of Personalisation”.
Mobilises users into groups
Ari Lightman, a professor at Carnegie Mellon College, said that another effective tactic TikTok has employed is to encourage its users to build groups publicly by hashtags.
By encouraging users to build public groups, TikTok can more effectively learn about its users’ behavior, interests, alignment and beliefs, he said.
If TikTok ends up getting banned in the US, Lightman said that while US tech giants undoubtedly have the ability to replicate TikTok with their own products, replicating the user culture enabled by TikTok could well be the harder task.
Chinese language advantage
TikTok’s recommendation algorithm was once also largely taken from its Chinese sister app Douyin, which was released in 2016.
Although ByteDance generally stresses that TikTok and Douyin are separate apps, one source with knowledge of the matter said the two algorithms remain similar to this day.
In turn, Douyin’s AI was supercharged by the firm’s ability to leverage low labor costs in China that allowed it to hire many content annotators to painstakingly label all the content and users on the platform.
“Around 2018 and 2019, Douyin worked on tagging every user. So they would label every video clip manually. Then they would label their users based on the video they had watched,” said Yikai Li, a manager at ad agency Nativex and a former director at ByteDance. “Then they also applied this tactic on TikTok.”
Whereas hiring annotators to label data is now a common and fundamental practice for AI companies, ByteDance was early in adopting this approach.
“Or not it’s miles a lot of labor sorting out these tags. Or not it’s extraordinarily laborious,” he mentioned. “So Chinese language corporations grasp a bonus here. It’s likely you’ll well present you with the money for a lot extra folks. The price is more cost effective than it’s miles for North American corporations.”
