TikTok Should Have Been Banned Years Ago, and Here's Why

Over the last couple of days, I have seen many of my marketer colleagues and AI researchers debating over the potential TikTok ban and defending the platform blindly.

As someone who not only runs ad campaigns on social media and has attended human-centered computer design courses for years but also spent some time in libertarian organizations defending the free market and individualism over state control, I have a few insights.

Let me be unequivocally clear: my advocacy for open-source AI, free from government control, is mirrored in my stance on social media platforms. However, the crux of the current debate focuses on China's proven exploitation of TikTok to surveil Western people, leaders as well as journalists and shape the ideological trajectory of our youth.

Without delving too deeply into politics, let me outline with data and research papers why I strongly support the bipartisan bill to ban TikTok, and why you should delete the app now!

Before TikTok, social media was social

Instagram, Twitter, and Snapchat; these apps focused on interactions between friends, groups, or communities until TikTok brought the dopamine hit and sucked users' attention spans.

The "Likes and Fragments: Examining Perceptions of Time Spent on TikTok" research is a proven example of how TikTok users often misjudge their usage time.

Findings highlight that users who engaged more intensively with the platform by liking more videos and opening the app more frequently had a higher discrepancy between their perceived and actual usage times.

This suggests that the engaging and immersive nature of TikTok, combined with its algorithmically curated "For You" feed, creates an environment where users are likely to misjudge the amount of time they spend on the app.

This is important because it leads us to TikTok's algorithm, which heavily relies on casino-like tactics. You have to be over 18 or 21 to gamble, but this app gets the next generations addicted from a young age.

So, what exactly are these tactics?

The Devious Tactics of TikTok's Algorithm

Significant research from the Brown University School of Public Health highlights the addictive nature of TikTok. The design of TikTok uses classical conditioning and reward-based learning processes to encourage continuous use. Features such as the "For You" stream customize content feeds based on user interactions.

Moreover, TikTok employs a variable reward pattern similar to slot machines, with the "swipe down" feature to refresh the feed mirroring the action of pulling a slot machine lever. This intermittent reinforcement keeps users engaged, always anticipating the next "rewarding" video.

I have explored the algorithm below more. This strategic manipulation exploits classical conditioning and reward-based learning processes, making the platform highly addictive and challenging to resist.

Last year, Utah Attorney General Sean Reyes filed a lawsuit against TikTok, showing how the "For You" recommendation algorithm hooks young users. You can read the lawsuit showing TikTok's casino tactics here for further exploration.

“TikTok induces that effect by relying on the same strategy casinos use to hook gamblers on slot machines – keeping them at the machine, spending a quarter at a time, hoping that the next pull will be ‘the one,’”

These random dopamine releases and other manipulative tactics hook the not-yet-fully-developed brains of our youth to this state-owned media channel.

Now, let's discuss why we should not leave the control of our next generation's brain development to communist China.

The Cultural Dichotomy

A comparison between Douyin (Chinese version of TikTok) and ours after China's National Intelligence took control of ByteDance shows a massive difference.

In 2021, China's National Intelligence took over the control of ByteDance. They appointed a government official with a golden vote (this person has the final say in the company), and after this change, Douyin's algorithms entirely changed.

While your children are bombarded with 'trash content', Chinese generations are exposed to educational content. While your one and only child spends hours doom scrolling, their children have mandatory time limits.

Marketing luminary Scott Galloway gives a strong reason to avoid TikTok:

"If I were part of the CCP and recognized a vested interest in undermining America's strategic position globally, I would subtly and cunningly tip the balance of content to cast America in an unfavorable light. I believe this is exactly what's happening at this moment—it would be foolish for them not to engage in such tactics."

The Addiction and Its Impacts

An average TikTok user spends about 60 minutes a day on the app and it is projected to exceed 100 minutes this year. The hyper-personalization algorithm leads to addiction-like behaviors.

It has been proven repeatedly that TikTok's hyper-personalization algorithm leads to problematic use patterns similar to addiction-like behaviors.

TikTok encourages isolation and loneliness among young people, and year-over-year depression statistics show that depression among youth has significantly increased following the widespread adoption of this soft power weapon.

"Individuals, particularly young people with social anxiety or depression, might be inclined to spend excessive time online, diminishing real-life interactions," explains Anne Marie Albano, PhD, director of the Columbia University Clinic for Anxiety and Related Disorders.

"Avoiding active engagement with the world — through social exchanges, confronting challenges, participating in discussions, or navigating peer conflicts — can intensify feelings of isolation, despair, and emotional distress."

The findings reveal that using TikTok elicits negative emotions in users, affecting their intention to continue using these platforms. It leads to feelings of wasted time, addiction, and even impacts mental health when usage becomes excessive or disrupts daily life.

"Understanding users’ negative emotions and continuous usage intention in short video platforms" research proves that Tiktok's negative emotion flow algorithm leads to consuming more and ultimately, addiction.

Weaponizing TikTok

TikTok serves as a weapon for the state-sponsored agenda. The research conducted by the Network Contagion Research Institute in collaboration with Rutgers University Miller Center investigates discrepancies in hashtag ratios on sensitive topics to the Chinese Government.

For non-sensitive topics, such as general politics and pop culture, the hashtag ratios between Instagram and TikTok approximately followed the user ratios (~2:1).

This pattern of anomalies was consistent across national/regional and international issues sensitive to China, suggesting a systematic approach by TikTok to influence content visibility based on its alignment with Chinese Government interests.

NCRI's research extends to geopolitically sensitive topics, demonstrating underrepresentation of hashtags favoring narratives opposed to Chinese Government interests, and overrepresentation of those in alignment, particularly regarding international conflicts where China has strategic interests.

The findings raise concerns about TikTok's impartiality and the platform's potential use in furthering the national/regional or international objectives of the Chinese Government. The discrepancies in hashtag representation underscore the need for a more comprehensive analysis to understand TikTok's influence on public narratives and its alignment with the Chinese Government's geostrategic objectives.

The NCRI's work highlights the importance of transparency and scrutiny in digital platforms' operations, especially when they have a significant impact on public discourse and possess the potential for governmental influence.

Indeed, it is a tool utilized for advancing state-sponsored agendas, with the recent "Letter to America" trend attributed to bin Laden being just one of the more notorious examples.

There are two primary uses of this instrument: extensive political manipulation and concerns regarding data security.

It has been established that TikTok amplifies content related to self-harm, mental health issues, and extremism to its younger audience. A Wall Street Journal investigation reveals that TikTok's algorithm has displayed footage of explosions, rockets, and distressed families from war zones to a 13-year-old user who has just signed up, highlighting the platform's exposure of young users to such graphic content.

Furthermore, research by Newsguard indicates that spending as little as 40 minutes on TikTok leads to exposure to false content regarding political events, such as Russia's invasion of Ukraine.

The study "An Empirical Investigation of Personalization Factors on TikTok" examines how user behaviors and settings affect the recommendations made by TikTok's algorithm.

The researchers used controlled "sock-puppet" accounts to focus on the impact of personalization factors, including language preference, geographical location, user interactions such as likes and follows, and video viewing durations, on the content that TikTok suggests to its users.

The "black box" nature of TikTok's recommendation algorithm with limited control over the content they are shown leads to echo chambers and filter bubbles, where users are increasingly exposed to content that reinforces their existing beliefs and interests, limiting exposure to diverse viewpoints.

These researches are showing that this app doesn't only make you addicted while diminishing your ability to pay attention, but it also deliberately changes the future of our very own civilization.

The other concern pertains to data security. With over 112 million TikTok installations on iPhones and Android devices in the USA, there are nearly 112 million distinct data sources, encompassing not only content but also real-time voice streams.

Data Security

The "Monolith: Real-Time Recommendation System With Collisionless Embedding Table" by Zhuoran Liu et al. at ByteDance Inc. showcases ByteDance's position in leveraging complex algorithms.

This is a system engineered for real-time recommendation.

Monolith addresses critical challenges inherent in recommendation systems, such as handling sparse, categorical data and adapting to non-stationary user behavior, through solutions like a collisionless embedding table and dynamic feature eviction mechanisms.

Let's clarify this;

Traditional recommendation systems often rely on batch training, where data is collected over a period, processed, and then used to update the model's recommendations. This can lead to delays in the model reflecting the latest user interactions.

Monolith's system, by contrast, uses online training. This means it learns from user interactions in real-time, allowing for a much more responsive and personalized user experience. The user's interaction with content is immediately fed back into the system, enabling it to adjust its recommendations on the fly.

In the context of machine learning, an embedding table transforms sparse, categorical data into dense vectors of fixed size, which the model can more easily work with. A "collision" in this context would mean different input data points being mapped to the same vector representation, which could lead to inaccurate model predictions. A collisionless embedding table suggests a method where these collisions are minimized or eliminated, improving the model's accuracy and efficiency in handling vast amounts of categorical data.

The system's architecture utilizes online training, enabling immediate adaptation based on user interactions, thus significantly refining content personalization.

However, this state-of-the-art AI personalization raises important discussions about data privacy, as the continuous, real-time processing of user data for personalized content delivery involves real-time extensive data collection and analysis, often without transparent consent mechanisms or clear data usage policies.

.

TikTok is under scrutiny for the extent of data it gathers to customize content and refine its algorithm. There is concern that this data could potentially be accessed by the authoritarian government and used against us.

One might counter-argue that ByteDance recently proposed transferring their data centers to the US by using Oracle servers. However, accepting this at face value is not only naive but also highlights that ensuring data security is an almost impossible task.

The Urgent Need for Action

China has been exposing its plans to rule the world with artificial general intelligence since 2017.

Deleting TikTok from our phones, creating a public narrative against its use, and sharing these findings and research to enlighten our friends are crucial steps.

I believe I am doing my part by writing this article to showcase why we need immediate action against this state-controlled app.

"Control over the use of one's mind is integral to human autonomy, and denial of such control undermines freedom of thought and freedom of will."

To gain deeper insights into the digital competition and its critical implications for the future, I highly recommend tuning in to the All-In Summit: Bill Gurley 2,851 Miles session, if you haven't watched it yet.

Here, Gurley delves into discussions about regulations, the future of AI, and the upcoming intense historical crash between China and the USA.

I urge you to share this content with your friends who use TikTok and encourage them to take immediate action by deleting this spyware app.

Your participation in this conversation is crucial.

Feel free to share your thoughts and comments below, as we continue to explore and address this important topic together.