Algorithmic Bottlenecks Reshape Language, Identity, and Reality

Original Title: How algorithms changed the way we communicate (w/ Adam Aleksic

The subtle, algorithmic currents shaping our words reveal a profound truth: language is not merely a tool for description, but a dynamic force actively reshaping our reality, our identities, and our collective understanding. This conversation with Adam Aleksic, author of "Algospeak," uncovers how the very platforms designed to connect us can inadvertently constrain our expression, leading to a fascinating dance between human creativity and technological influence. Those who grasp these hidden dynamics gain a critical advantage in navigating the modern world, understanding not just what is being said, but why it's being said, and how it’s being shaped. This is essential reading for anyone seeking to understand the invisible forces influencing our daily thoughts and interactions.

The Algorithmic Bottleneck: How Platforms Reshape Expression

The evolution of language has always been intertwined with the mediums through which it travels. From oral traditions to the printing press, each technological shift has fundamentally altered how we communicate and structure knowledge. Adam Aleksic, drawing parallels to Walter Ong's work on orality and literacy, argues that understanding today's digital language shifts requires looking back at these historical precedents. The internet, and social media in particular, are not entirely novel forces; they are extensions of a long history of media shaping human thought. However, the unique characteristic of modern platforms lies in their proprietary algorithms, which act as powerful, opaque bottlenecks for creative expression.

Aleksic recounts his own experience, where the desire to create a specific piece of content was tempered by the knowledge that the platform's algorithm might not favor it, thus limiting its reach. This isn't about the inherent "badness" of a medium, but about recognizing that "good art comes out of a resolution of that bottleneck." The challenge, then, is to develop media literacy--understanding not just written language, but the underlying mechanisms driving content delivery. This includes questioning why a video appears on a "For You" page or why a particular phrase goes viral.

"Awareness starts with media literacy understanding that every medium is a bottleneck for creative expression good art comes out of a resolution of that bottleneck and i think it's still possible i've seen amazing art on tiktok i don't think that means it's like a bad medium but it does constrain our way of speaking to a certain style of spoken cadence to a certain type of language that works with the metadata and once you learn how to express yourself within that there are many creative possibilities that open up but it starts with media literacy and we use literacy thinking about how we understand written language but you need to understand also where a video is coming from why an algorithm is recommending a video the way it is and uh why is a video showing up on your for you page why is this viral in the first place all these are questions that have answers to them and should be asking them when you look at any video"

-- Adam Aleksic

This algorithmic influence isn't confined to social media users. Phrases and trends originating online inevitably "bleed off" into the broader culture, affecting even those who consider themselves offline. Language, Aleksic explains, functions like a virus, spreading through networks. Social media algorithms, by creating vast, interconnected nodes, accelerate this natural process, making words and phrases propagate faster than ever before. This viral spread means that even seemingly individual linguistic choices are part of a larger, algorithmically influenced ecosystem.

Euphemism, Censorship, and the Linguistic Arms Race

One of the most striking consequences of algorithmic influence is its role in shaping euphemisms and responses to censorship. Aleksic highlights the word "unalive" as a prime example. Originally adopted by young people to discuss death and suicide due to discomfort with the topic, its widespread use was amplified because platforms like TikTok would suppress content containing the word "kill." This created a linguistic workaround, a euphemism born not just from social awkwardness but from technological constraint. While middle schoolers might not be aware of the algorithmic origin, the word's function as a less frightening term for death is undeniable.

This dynamic creates a continuous linguistic arms race. Aleksic describes censorship as "linguistic whack a mole," where attempts to police language lead to the rapid evolution of new terms and phrases. He points to the example of censorship in China, where the word for censorship itself was censored, leading to a cascade of euphemisms like "harmony," "river crab," and eventually "aquatic product." Each new term, when identified and targeted, prompts the emergence of another.

"The more the algorithm tries to police our language the more language is going to evolve faster -- and there's always speculation about the algorithm too because it's this opaque ineffable process you want to hypercorrect if anything so people will self censor themselves before they even know whether or not the algorithm is actually censoring their videos and that's a whole another process going on but there are all these -- incentives for language to be produced quicker than ever and the algorithm there creates an incentive for us to use language faster"

-- Adam Aleksic

This constant evolution, while a testament to human creativity, also raises concerns about the opaque nature of the controlling entities. Unlike historical shifts in language, which were often organic or driven by cultural institutions with some level of public scrutiny, today's algorithms are proprietary and profit-driven. This leads to a "brave new world" scenario, not of overt thought control, but of being "amused ourselves to death" or "engaged ourselves to death" by platforms that prioritize engagement over genuine understanding. The consequence is a subtle restructuring of our reality, driven by what makes us click, rather than what makes us think.

The Unseen Influence: From "Side Eye" to Sexual Preferences

The impact of algorithmic language shaping extends far beyond explicit content moderation. Aleksic emphasizes that even seemingly innocuous online trends can filter into offline life, altering our communication and even our desires. The phrase "side eye," for instance, originated in online visual communication and has become a common idiom. Similarly, the increasing use of the word "delve," noted by Aleksic as being more prevalent in AI language models than human speech, suggests a subtle but pervasive influence of artificial intelligence on our own vocabulary.

This influence can have profound downstream effects. Aleksic discusses how marginalized groups, particularly LGBTQ+ creators, have experienced algorithmic censorship. The word "lesbian" was flagged, leading to creative workarounds like replacing letters with symbols (e.g., "l$b!an"), which then became a form of "algospeak." This linguistic subversion, while demonstrating resilience, also highlights how platform policies, often designed with good intentions, can inadvertently suppress vital conversations about identity. The subsequent adoption of "wlw" (women loving women) as a more algorithm-friendly term exemplifies this ongoing adaptation.

The most alarming consequence, however, lies in how these algorithmic clusters can reshape fundamental aspects of human experience, including sexual preferences. Aleksic points to the surge in the term "tradwife" on platforms like Pornhub, noting its rapid adoption from a niche term to a widely searched concept. This phenomenon is driven by algorithms identifying and amplifying content that aligns with perceived subconscious desires, creating a feedback loop that can influence users' understanding of relationships and gender roles. The "tradwife" aesthetic, often a performance by content creators, presents a hyper-feminized image of patriarchy, which can then bleed into real-world sexual preferences and societal norms.

"The more i look into it the more it seems like every single desire or emotion we have is just shaped by our media or at least our environments which is indirectly shaped by our media like it's it's a timeless thing one example i like to use is where vibes changing and i think if you look at the pornhub statistics for the year 2024 the usage of the word tradwife was one of their most used words and it shot up 72 from the previous year that word was coined in 2019 how is this one of their most used words now how did this suddenly take root and it's partially due to the you know rising popularity of tradwife influencers on tiktok but also the fact that there is like these sexualities marketed to people that there's a -- this is people's subconscious desires that they don't think are really being perceived and yet they are part of this greater network of of algorithmic clusters looking at data about you and sorting you into the tribe of people who should get more tradwife videos because other people have engaged with this -- and then you get more goon bait on your feed and then that sort of leads into a restructuring of our ideas about -- male to female interactions or -- the trad wife is this -- the idea is traditional wife and it plays into the aesthetic of a 1950s trad wife but again not really what a housewife in the 1950s was because any trad wife influencer you see on social media is actually a content creator who is running a business -- and so they're also performing this exaggerated persona of a trad wife which gets more comments -- so what you end up engaging with is this hyper feminized image of -- patriarchy that ends up actually bleeding over into people's sexual preferences and i i find that -- really kind of alarming"

-- Adam Aleksic

This underscores the critical importance of media literacy. It's not just about consuming content; it's about understanding the underlying systems that curate and deliver that content. By asking "why" -- why a video is on our feed, why a word is trending, why a particular aesthetic is popular -- we can begin to reclaim agency over our understanding of reality, rather than passively accepting a filtered version presented by opaque algorithms.

Key Action Items

  • Develop Algorithmic Literacy: Actively seek to understand how social media algorithms work. This involves reading articles, watching explainer videos, and reflecting on why certain content appears on your feeds. (Immediate Action)
  • Question Euphemisms and New Slang: When encountering new or unusual words, especially those related to sensitive topics, investigate their origins. This practice, akin to studying etymology, can reveal underlying social and technological influences. (Immediate Action)
  • Practice Mindful Scrolling: Consciously pause during social media use to reflect on why you are engaging with certain content. Ask yourself if your stated preferences align with what you are seeing, and if the content genuinely serves your interests or the platform's engagement metrics. (Immediate Action)
  • Diversify Information Sources: Do not rely solely on social media for news or cultural trends. Actively seek out information from a variety of sources, including long-form articles, books, and direct human interaction, to gain a more balanced perspective. (Ongoing Investment)
  • Engage in "Thinking Time" with Media: Structure your media consumption to allow for reflection. When reading, pause after paragraphs. When watching videos, consider taking notes or discussing them afterward, rather than passively consuming. (Long-term Investment)
  • Educate Younger Generations on Media Literacy: Advocate for and implement media literacy education in schools, teaching critical thinking skills related to online content and algorithmic influence, rather than simply restricting access. (1-3 Year Investment)
  • Advocate for Platform Transparency: Support initiatives that push for greater transparency in how social media algorithms operate and how they influence content distribution and user behavior. (Long-term Investment, requires collective action)

---
Handpicked links, AI-assisted summaries. Human judgment, machine efficiency.
This content is a personally curated review and synopsis derived from the original podcast episode.