Introduction: Algorithms aren’t just recommending videos, they’re reshaping the way we think
We spend hours every day on social media, hooked on the thrill of discovering something new with every swipe. Short videos may seem like mere pastimes, a bit of light entertainment to kill time, but the recommendation algorithms behind them are far from harmless. They aren’t simply guessing your preferences, they are vying for your attention, shaping your emotions, and even gradually rewriting the way you understand the world.
The most frightening thing is that this transformation often happens so quietly that you don’t even notice it. Imagine this scenario: you and a friend are sitting side by side on the same sofa, scrolling through the same platform, yet it feels as though you are living in two completely different universes. You see beauty tips, recipes and topics on women’s self-improvement, whilst they see aggressive misogynistic comments, videos that objectify women, and malicious mockery of gender issues. Although you clearly share the same physical space, you have been assigned by algorithms to realities that you cannot truly see. As scholar Mark points out, this highly personalised ‘automated culture’ is eroding the very foundations of our public dialogue, making it increasingly difficult to build empathy and mutual understanding with those who hold differing views (Andrejevic, 2019). This is the most dangerous aspect of social media today. It creates not information richness, but information isolation, not a diversity of views, but emotional conditioning. Safiya (2018) once put it succinctly: algorithms have never been a neutral, objective mirror, but rather commercial machines that reflect and amplify society’s inherent biases (particularly gender and racial discrimination). Their aim is not to make you more discerning, nor to help you understand the world more comprehensively, they simply want you to stay longer, click more, and experience more intense emotional fluctuations.
This is because anger generates more engagement than calm, extremism spreads more easily than reason, and conflict drives more traffic than understanding. Consequently, platforms repeatedly push the most provocative, polarising and emotionally charged content into your feed. Over time, we come to believe we are seeing ‘reality’, when in fact we are merely viewing slices of reality carefully curated by algorithms. We believe our views are formed independently, yet in reality, through successive rounds of recommendations, engagement and reinforcement, we are quietly steered towards a more extreme direction. So, the issue has never simply been about what you like to watch, but rather, who is deciding what you can see? And who, through every moment you spend on the platform, has quietly shaped the person you are today?
Algorithms are not neutral, they are only accountable for dwell time
First, let’s be clear about one thing: algorithms are not objective, they are a set of logic designed to serve a specific purpose.
Many people think of algorithms as impartial mathematical formulas that simply recommend content you might like in a neutral manner. In reality, however, they are more like a personal assistant with clear KPIs. This assistant couldn’t care less whether you become more discerning or knowledgeable as a result, it cares about only one thing: how to keep you from leaving. The way it works is not complicated: it observes you, analyses you, and then feeds you content. Every time you linger, like, comment, or even just hesitate briefly whilst scrolling, it records it. Then, based on this data, it labels you, categorises you into a certain user type, and by comparing your behaviour with that of a large number of similar users, predicts what you are most likely to want to see next. Once it discovers that you react to a certain type of extreme content, it will keep ramping it up. Consequently, the more you look, the more it pushes, the more it pushes, the more likely you are to mistake this curated content for the full picture of the world.
This is also why the internet today is increasingly rife with hostility, borderline content and extreme rhetoric. Because what the platforms are truly competing for is not your approval, but your attention. Anger, sensationalism, fear, polarisation and soft pornography these elements that most readily trigger instinctive reactions are also the most effective at generating dwell time, engagement and traffic. By contrast, rational, restrained and objective content struggles to gain the upper hand within the algorithmic system.
Algorithms have no morality, they are simply optimisers that prioritise efficiency above all else. As Professor Safiya(2018) has said, on the internet, the most ‘popular’ content is by no means synonymous with the most ‘truthful’ content. The algorithm does not care whether the content is true, whether it misleads you, or whether it drives public discourse towards greater extremism. It cares only about one thing: what will keep you scrolling.
Therefore, what the algorithm pushes is never the truth, but rather the content most likely to hook your emotions. You think you are observing the world, but in reality, it is the algorithm observing you, then precisely delivering the digital feed most likely to stimulate your nerves right before your eyes.
How Algorithms Influence Adolescents’ Cognition
Whilst adults may still manage to retain a modicum of clarity amidst the algorithmic information echo chambers, minors, whose minds are not yet fully developed, are facing an unprecedented cognitive crisis.
Compared to adults, young people are far more susceptible to the influence of algorithms. This is because their worldview, values and judgement are still in the process of formation. Many of their understandings of what the world is like, what is normal, and what is worth believing have not yet truly solidified. At this stage, the content algorithms push to them is not merely entertainment; it is actively shaping the way they perceive the world. As law professor Frank (2015) has pointed out, today’s digital society is like a ‘one-way mirror’: platform companies have a clear understanding of our browsing habits and psychological vulnerabilities, whilst we remain completely in the dark about the logic behind the recommendation algorithms. Consequently, young people often fail to realise that the content they see does not appear naturally, but has been carefully curated. Platforms do not present them with a complete, balanced and diverse world, instead, based on their dwell time, clicks and emotional reactions, they continuously feed them content that is more likely to hook them. Over time, the information they encounter becomes increasingly one-sided, their horizons narrow, and they may even mistake a particular, constantly repeated viewpoint for common sense and believing that everyone thinks this way.
What is even more dangerous is that algorithms are particularly adept at exploiting the intense emotional fluctuations that are inherent to adolescence. Adolescence is, by its very nature, a period of heightened sensitivity to issues of identity, belonging, appearance, gender relations and social judgement, yet the very topics that platforms excel at amplifying are precisely those most likely to generate anxiety and division. Consequently, extreme gender-based rhetoric, content fuelling body image anxiety, conspicuous consumption, and even highly aggressive value judgements are more likely to appear repeatedly in teenagers’ news feeds. When a person is immersed in such a content environment over a long period, their perceptions gradually shift. They may begin to view gender conflict as the norm, come to believe that appearance and online popularity matter more than ability, or gradually lose patience with complex issues, becoming accustomed to understanding others and society through emotional, labelled, and black-and-white lenses. Algorithms influence not merely what you choose to watch, but how you interpret conflict, judge right from wrong, and conceive of your own and others’ positions.
What is even more ironic is that this influence often appears cloaked in the guise of personalised recommendations, leading people to mistakenly believe they are exercising free choice. In reality, however, young people are often not actively engaging with the world, but passively accepting the answers provided within a world that has already been curated by algorithms. Over time, independent thinking diminishes, emotional reactions become quicker, complex issues are simplified into a few easily shareable positions, and rational judgement is more readily replaced by immediate stimuli. Therefore, the impact of algorithms on young people is far more than simply causing them to become addicted to their phones. What is truly frightening is that, during the stage of life when a person is most malleable, algorithms may quietly determine how they understand the world, view others, and even define themselves.
Case Analysis: How algorithms create a dangerous world for minors
To be honest, before I started researching for this blog post, I, as a young person who spends hours online every day, had never even heard of the term ‘Beng Lao Tou’. It was only when I typed those characters into the search bar and saw the jaw-dropping content that I felt as though I had broken through an invisible wall.
This is precisely the most chilling aspect of algorithmic filter bubbles: just because you haven’t seen a world doesn’t mean it doesn’t exist. It has simply been precisely filtered out of your news feed by the algorithm. Among those young people (particularly those born in or after 2005, and even 2010) who have been algorithmically assigned to another ‘parallel universe’, ‘Beng Lao Tou’ is becoming a surging, perverse subculture online. Put simply, it refers to immature young girls who seek out middle-aged men online, using tactics such as skirting the line, sending intimate photos and engaging in suggestive interactions to extract money. In the autocomplete suggestions of short-video and social media platforms, this term is often closely associated with labels such as ‘underage’, ‘immature girls’ and ‘scamming for red packets’—terms carrying distinctly grey or even dangerous connotations.
What truly warrants our vigilance in this case is not merely the sensational buzzword itself, but the fact that it lays bare the deep-seated risks of algorithmic distribution mechanisms. Algorithms act as extremely sensitive detectors. The moment the system identifies that a young user has lingered for a few extra seconds on such provocative, boundary-pushing content, it begins to relentlessly and continuously push similar material in front of her. For minors whose worldviews are still taking shape, this mechanism is deadly. When their screens are inundated daily with ‘success’ content promising easy money, they will mistake this repetitive content for normal, commonplace, or even a rule of life worthy of emulation. It creates a false sense of a shortcut for these inexperienced young women, leading them to believe that compromising their principles can be exchanged for value, whilst completely overlooking the legal red lines and the profound personal costs involved.
What kind of digital governance do we need?
Firstly, algorithms should be visible, rather than perpetually hidden behind a black box. Today, most users have no idea why they are seeing certain content, nor do they know on what basis the platform determines ‘what you like’. If digital governance is to be truly effective, algorithmic transparency must be improved so that the public is at least aware of the basic logic behind platform recommendations, which content is amplified, and which behaviours trigger further content feeding.
Secondly, platforms must assume clearer responsibilities. They cannot, on the one hand, use algorithms to amplify emotions, stoke conflict and chase traffic, whilst, on the other, shifting all blame onto ‘users who chose this content’ when problems arise. Recommendation systems do not emerge organically, they are designed, trained and optimised by platforms, and therefore platforms must be held accountable for the social impact they cause. Furthermore, platforms should be required to offer a ‘no-algorithm-intervention’ news feed as a user’s default option, thereby returning the power of choice to the people.
Finally, I believe that content distribution targeting minors, in particular, must undergo third-party ethical and safety audits. Young people need a healthier information environment, clearer platform guidelines, and genuinely effective digital literacy education. For in the age of algorithms, the greatest danger lies not merely in misinformation itself, but in the fact that it is becoming increasingly difficult for people to distinguish whether what they are seeing is the real world, or a world crafted for them by the platform.
References
Andrejevic, M. (2019). Automated Media, Chapter 3: Automated Culture (pp. 44–72). Routledge. https://doi.org/10.4324/9780429242595
Noble, S. U. (2018). Algorithms of Oppression : How Search Engines Reinforce Racism. In Proquest.com (pp. 15–63). New York University Press. https://ebookcentral.proquest.com/lib/usyd/detail.action?pq-origsite=primo&docID=4834260
PASQUALE, F. (2015). The Black Box Society: The Secret Algorithms That Control Money and Information, Chapter 1 “Introduction – The Need to Know” (pp. 1–18). Harvard University Press; JSTOR. https://doi.org/10.2307/j.ctt13x0hch
Be the first to comment