I miss the algorithm in my fediverse.
I miss the algorithm.
The algorithm—a term that has become a buzzword in the realm of social media, and one that stirs a mixture of nostalgia and frustration within me. Like many, I recall the days when Twitter’s chronological timeline was the norm, a straightforward feed that allowed me to seamlessly follow my friends and acquaintances in real-time. Yet, as social platforms evolved, so did the mechanics governing our digital interactions. Enter the algorithm to Twitter, (now “X”, what an awful rebranding), that mechanism that generates a timeline who drives engagement. In this age of algorithmic timelines, I find myself oscillating between yearning and disdain.
I remember vividly the advent of the algorithm, a shift that I greeted with vehement opposition. The abandonment of the chronological sequence seemed to sever my connection to the familiar. The sense of control over my feed slipped through my fingers, replaced by an automated curation that left me disoriented. However, time, as it often does, worked its magic on me. Gradually, the algorithm etched its place in my daily scrolling routine, possibly augmented by a generous dose of dopamine, igniting a newfound affinity for this previously despised mechanism. But as my scrolling habit intensified, an alarming pattern emerged—the peril of “doom scrolling.”
When Elon Musk bought Twitter, and there were the big layoffs, it creates waves of migration within the social media landscape, and a lot of people starting to migrate to Mastodon and the Fediverse. Guess who hopped on that train? Yep, yours truly. And what did I find? A community wonderland that felt like a reunion after years apart. Now, I had dabbled in Mastodon before, but this time, it was like I’d walked into a new party with fresh faces and cool stories. However, building that network of people who I wanted follow was difficult. Shortly after following #hashtags was released and things improved, but still, getting to that point where my feed is curated has been hard, and I am sure that is by design, there is a full control on the feed.
So, I still spend most of my time in Mastodon, but also go to bed doom-scrolling in Twitter. Every day, I unfollow more people, and every day I found that twitter is not for me, it is getting worst, but I am missing that algorithm that once was so engaging.
In my quest for content that clicks and connections that count, I’ve got a couple of issues:
- The Blink-and-You-Miss-It Dilemma, The Vanishing Act. Sometimes someone that I follow posts something interesting, but it is a time or day that I did not checked Mastodon, and gets lost in the noise. Sometimes, a gem from someone I follow pops up, but I wasn’t looking at that exact moment, and poof, it’s gone in the flood. Total bummer.
- The Endless Content Hunt. Sure, #hashtags help, but sometimes it’s like searching for a needle in a digital haystack. And there are so many cool people on the Fediverse—I’m certain I’m skipping over gems. I want to find stuff I dig.
Acknowledging the Fediverse’s ethos of frugality, a safe haven of sorts for those who wish to remain undiscovered, I am cognizant of the community-centric spirit that prevails. I am aware that virality is not a thing the fediverse is pursuing, but more a sense of community, so I want to keep this as a tool for myself and only myself, also as a project to learn and upskill.
I want to tap into the content I enjoy and the people I follow, using them to create an algorithm that tailors relevant content and connections for me. My plan is to craft a bot that spots content aligned with my interests. When it detects something cool, the bot will give it a boost. I will follow the bot, and whenever I want to get that algorithm fix, I can just go to the bot timeline. Hopefully over time, the bot will learn (through my likes and feedback retro) what I like (and dislinke) and will be useful to me.
I will start an open-source project, for fun, so if someone wants to create their own algorithm-bot can do it.
UPDATE: I created a proof-of-concept and released the code here in racoonbits, this is what is behind this bot @raccoonbits.