The effects of algorithms on the public discourse

We traded blogs for black boxes, now we're paying for it 09/09/2025 Come listen to the "old man yelling at clouds" in me for a bit. tl;dr: The internet is changing for the worse (or getting 'enshittified'). In this post, I write about the effects of algorithms on the public discourse to illustrate a greater point on the enshittification of the internet. Then, I offer my personal notes and curated resources to guide you on your personal internet deshittification journey. I miss the old internet, and I don't mean the slow dial-up internet. No, I miss the internet specifically before it collapsed into a single, endless feed that we access through proprietary apps (now with AI slop!). I miss how it was a network that's both human, and humane. Back when it wasn't (primarily, at least) an algorithmic black hole made to just suck us in. Maybe it's the increasing corporate ownership of the internet that took people's agency from their hands. Who knows? I want to remember, for example (and hopefully help rebuild) the internet where actual humans discovered each other's work by following trails others cared enough to leave behind; not because an LLM or some billionaire brat's algorithm 'fed' it to them. We used to use our curiosity to find stuff we liked, which felt more satisfying than any algorithm's statistically-determined aggregation. I've had this blog hosted on BearBlog for more than a month now, and thanks to the community here, I've rediscovered the indie web. What's left of it, I mean. Still, it has been an absolute joy to read human-written words on all kinds of things I'm interested in. Many of my thoughts on the internet and the digital world as a whole were shaped by those. How cool is that! If this kind of an internet is still possible, why is it such a small part of the general network? Well, money. Here are examples to what makes money on the internet: Context Collapse I recently watched a great video from Technology Connections titled 'Algorithms Are Breaking How We Think' (YouTube Link/Invidious Link). Technology Connections talks about something known as 'context collapse'. The idea behind the term is that by combining multiple audiences into one single context, algorithmic media platforms bring out (usually) negative reactions from unrelated audiences. This means that even though every one of the billions of internet users have entirely different lives and backgrounds, social media platforms (even the seemingly-personalised ones like TikTok) have one single context on their main tabs. That context is that the user is a number among many that must be kept positive at all costs. Consequently, the ideal user of modern media platforms has no individual identity, and is perfectly happy with what's given to them by the algorithm. As an example, think of how many of the posts you see on Instagram are actually from people or accounts that you deliberately chose to see. On the other hand, think of how many posts are from accounts you couldn't care less about, but the algorithm thinks you should (don't forget the 1/1 ad-to-post ratio!). The same thing goes for many other platforms. Even though some, like YouTube, at least allow you to use the platform with almost full agency over what you see , not many platforms do this. You have no choice but to use the official app with their own black-box algorithm . This way, the user's individual identity is replaced with that of the "average internet user", which they share with billions of other 'averages'. The effect that this has on online discourse is immense. As every internet user must know, the internet (as it is today) is full of negativity. You might look at this and say, "Humans are inherently evil," and lose all hope in humanity given how horribly they treat each other on the internet. The truth is, the way we have constructed this space deindividualizes, alienates and aggregates us to such an extent. In other words, when billions of us are thrown into one large melting pot of context, the output of that pot becomes as impure as it could possibly be. It does not reflect how we form groups, act, and react in other areas of life. Interpretive Communities Literary theorist Stanley Fish argued that we as individuals interpret any given text (in this case, social media content) "because each of us is part of an interpretive community that gives us a particular way of reading a text." That interpretive community usually isn't there when we are fed what the algorithm thinks we'll consume. We may share something thinking that people like us will see it share their opinions. However, because of the way that algorithms work, engagement is the main driver of a post's visibility; so here comes 10 million people who have no clue of the context on that thought you shared about your niche interest. Suddenly your post is full of over-the-top jokes and non-content-related quips from members of a completely random mix of audiences. As the algorithm prioritizes engagement, your post's new mixing pot of clueless audiences outnumbers the genuinely-interested audience of your own niche corner of the internet (if that even exists anymore), and they comment about everything BUT the content. Think That's a Stretch? Okay, that was a hypothetical scenario, but here's a real community that massively benefited from all this: the incel community. Adam Aleksic’s book Algospeak: How Social Media Is Transforming the Future of Language tackles with this exact topic. He argues that algorithms contributed to the wide-spread use of pseudo-scientific incel terms by way of increased exposure. Here are two of my highlights: "Once algorithms got involved, the incel terms were amplified by the online Matthew effect: a phenomenon where content that is slightly better at grabbing your attention performs exponentially better on social media." Then, when these pseudo-scientific terms (think: 'alpha/beta male', 'chad', 'canthal tilt', 'looksmaxxing' etc.) are taken outside their respective interpretive communities and made hyper-visible, the mixing pot works its magic. From here, these terms make their way to the general public discourse: "Poe’s law explains how dangerous ideas spread as memes. If something is meant genuinely, but it is also crazy enough to be interpreted as a joke, people may reward it with “likes” and other engagement because they find it funny. Meanwhile, if something ironic is interpreted as genuine, people will be offended by it, which then also drives engagement as a form of ragebait. Either way, “edgy” humor is able to worm its way into the mainstream via the algorithm." So, if you want to get more people to know about your extreme views, package them as 'dumb jokes' for the algorithm to advertise to billions of people. At least that's what the incel community did, and succeeded. Can you even count how many times you heard or read those terms? Okay, What Can I Do? Now, this is the part where I stop being the old man yelling at the cloud and start proposing a new way of using the internet. In this section, I have gathered some links to things that I either use or have used in the past, some personal notes on media I consumed, and quotes by brilliant writers. I'm doing this because I want to share a little bit of what I learned during my research on the open internet, even if they are not directly related to algorithmic feeds. Use and Support Free (Libre) Software "The tech stack we use is an endorsement of the companies behind it. Technological platforms are not neutral. If we truly want to resist the digital coup that is currently under way, we need to normalize the use of free, open source solutions." --- and --- "We constrain our imaginations when we subordinate our creations to names owned by fascist tycoons. Imagine the author of a book telling people to "read my Amazon". A great director trying to promote their film by saying "click on my Max". That's how much they've pickled your brain when you refer to your own work and your own voice within the context of their walled garden. There is no such thing as "my Substack", there is only your writing, and a forever fight against the world of pure enshittification." --- highlights from "This is what resistance to the digital coup looks like" by Elena Rossini Curate Over Engage Start a blog if you want. Share your experiences or thoughts about anything. You don't have to write something that has never been written before. Just have fun, and some people will probably benefit from, and enjoy your writing at some point. Scavenging for information used to be optimal, but now we live in a world of information overload. Notice the algorithm's firm grip over your attention. Prioritize discernment; focus on what information is needed and how it connects to your existing knowledge. Set boundaries on information consumption by using attentional filters. Block out redundant, overwhelming, or poor-quality information to avoid constant distraction. On helping revive the human network of the internet: "Regardless of the kind of site you have, whether its a modern site for a business or arts practice, or if its a personal homepage; setup a links page on your site and add links to your friends, people you’ve worked with, tools you’ve used, blogs you enjoy, other businesses, anything and everything that deserves to be seen; because it does deserve to be seen, and so do you!" ---highlight from "Why linking matters" by Melon's Thoughts I'll add more to this list as time goes on, and use it as my personal rolling list of apps, services, or notes for a more open and free internet. You can bookmark it if you like the list. But for now, thank you for visiting my corner of the internet. This post was last edited 1 hour, 42 minutes ago.