How to Train Your Algorithm
Social Media and Agency
Welcome to Polymathic Being, a place to explore counterintuitive insights across multiple domains. These essays take common topics and investigate them from different perspectives and disciplines to come up with unique insights and solutions.
Today's topic is a practical guide to exerting your control over social media and training your algorithm to avoid the downward spiral of content that permeates everything. We’ll explore the problem and provide a simple set of solutions for how to train your algorithm.
Intro
Algorithms are a lot like dragons. They appear to be mythical, fearsome creatures which we have no control over but, in reality, they are a lot like the dragons in the movie from which I borrowed this title. While algorithms and dragons are potentially dangerous, they can be trained with intentional attention, and they certainly should not be ignored.
In 2022 the TV show, 60 Minutes, did an investigation into Instagram where they opened an account under the auspices of being a thirteen-year-old girl and within a very short time were being fed content on anorexia and self-harm.
We hear similar stories of TikTok and Instagram slowly nudging people toward more graphic, sexually exploitative, and potentially harmful content. Even if it isn’t specifically harmful, let’s just say that the algorithm isn’t biased toward happy and healthy content. This is clearly evident in the work Jonathan Haidt is documenting regarding cellphone and social media use and teenage anxiety.
On a personal note, I’m taking a hard stance against social media for my kids. Right now they won’t get a cell phone until they are 16 and I don’t plan to let them have social media until then, or later. The evidence shows it’s just not worth it until they’ve built up enough life experience to know how to train their algorithms.
It’s clear there’s a problem but I want to shift gears and show, that instead of fearing our algorithm, we do have tools on hand that we can help fix it ourselves. As we explored in Agency vs. Addiction, I believe we have agency and can begin to eke back control of our social media and online habits toward healthier goals.
Agency
We all know that hitting that like button whether as a ‘thumbs up’ or a ‘heart’ trains your algorithm in one direction. (And don’t forget to hit that ❤️ button, comment, and most importantly, share these essays if you do like them!) This is the action that drives the algorithms to highlight a creator. It’s the magic of trending.
What’s much less well-known is that those likes will also inform the algorithm of a pattern that feeds in adjacent content. Basically, it says, “Mike liked this post, other people who liked this post also liked this other post,” and then it is likely to show that post as well.
This is where the slippery slope starts to happen and instead of finding more and more positive things, the opposite happens. This is partly coded because we have a negativity bias and partly because the old news adage “If it Bleeds it Leads,” is a very true human condition.
But just like we can hit that like button and get fed more, we also have tools that can negate the negative. For example, in Instagram, you can either just hit the “X” button on the upper right which is the anti-like button or you can hit the three little dots … and open a new series of options from “Report” which is useful to flag problematic content, which you should do, to an option to select just ‘Not interested,” which starts to train your algorithm.
Facebook has a similar interaction with the “X” and the “…” but offers more options including snoozing, hiding, or blocking profiles. (nothing against Dan Marries, it’s just an example.)
You can also do this for Ads across all the platforms as well. Lest you think this isn’t important, ads are one of the primary ways these sites profile you to feed you other content as well. What’s interesting about LinkedIn is that you can also click “Why am I seeing this ad?” and it, sometimes, gives you some insight. But you can also select “Hide this ad” and take one more step in training by saying you don’t like it.
Facebook and Instagram Reels have similar options to either:
Click those dots and select “See less” or
Hit that ‘like’ button to see more things like that.
I’m not going to go through all the social media sites and options as they all have some form or flavor of these options when you know where to look. Thankfully they all use a similar user interface to make it easy to find.
Taking Action
Now that we’ve learned the ways to give positive or corrective feedback to your algorithm, let’s talk about a few other intentional steps you can take.
DO NOT doomscroll. Doomscrolling is where you bite on those clickbait headlines of doom and gloom and it’ll never stop.
Stop and think about what you want to see in your social media time. Arts and Crafts? Comedians? Cat Videos? (a favorite of mine!) Scientific, Technological, Educational? Define what you want to see and look for it.
Hit the ❤️ button when you do like the content.
Provide corrective feedback to content you do not want to see like we saw above
Stop and think about what you don’t want to see as well. Train yourself to not accept ‘bad behavior’ from your dragon, *ahem,* algorithm.
‘Follow’ accounts that provide the content you want to see.
Provide corrective feedback to content you do not want to see.
‘Unfollow’ accounts that no longer appeal or you aren’t thrilled by. These accounts will start to edge you further away from your target.
Yes, we do just zone and scroll but once in a while scroll and consider and clean. That way when you zone and scroll it’s not dragging you down.
Don’t ignore the ads, if it’s a bad match, train your algorithm why. This will also have an effect on the content you are provided.
Sub-tip here, if there’s an ad that does interest you consider searching for the product in Google instead of clicking the ad.
Provide corrective feedback to content you do not want to see. (yes, it is intentionally in the list three times because if you don’t do this and just passively consume, you’ll keep sliding down that slope.
These are some basic tips I use in my social media that help curate my feeds toward healthier interactions when I just want to sit and zone on silly cat videos. If I see my feed start to hit that slippery slope, I’ll pause and provide that corrective feedback to keep my algorithm in check.
Train your algorithm like you’d train a dragon. With intentionality, non-complacency, and with both positive and corrective feedback.
What tips and tricks have you learned in your own experience?
Another tip is to train your own curation algorithm:
Every day, Refind analyzes thousands of articles and sends you only the best, tailored to your interests. Loved by 503,336 curious minds. Subscribe Here
Enjoyed this post? Hit the ❤️ button above or below because it helps more people discover Substacks like this one and that’s a great thing. (and it helps train your algorithm) Also please share here or in your network to help us grow.
Polymathic Being is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Further Reading from Authors I Appreciate
I highly recommend the following Substacks for their great content and complementary explorations of topics that Polymathic Being shares.
- All-around great daily essays
- Insightful Life Tips and Tricks
- Highly useful insights into using AI for writing
- Integrating AI into education
- Computer Science for Everyone
I do not have a social media account. However, I try to confuse streaming algorithms by picking random movies and series, which I will never watch while multitasking (when TV is not my focus).
Feed it nothing but shitposts so that it completes the classic copypasta. Might be showing my age a little, but here's my Turing meme test:
Katanas are underpowered in d20...
I sexually identify as an Apache attack helicopter....
What the fuck did you say about me, you little bitch...