Your Close Friend… The Algorithm
At face value, social media is an entertaining way to keep in touch with not only your family and friends, but also indulging in larger social spheres outside your personal life — music, celebrities, fashion, etc. Catered content is constantly being introduced every time you open or refresh your application or web browser. This is learned and calculated by a mathematical system called an algorithm. We have all come across this word in both the arithmetic and computing sense, but how has it evolved and implemented in our most favored social media? Here, I will be evaluating the algorithms implemented on Twitter, YouTube, and Instagram
Twitter’s timeline is organized by three different sections: the initial curated algorithm, “In case you missed it…”, and finally, the reverse chronological order. The first one is dedicated for the likes and retweets of those you follow, as you will see by the top of the tweet “Liked by…”. A significant sign is the time stamps, all varying from likes and retweets from two days or two hours ago. This differs from the “In case you missed it…” in the sense they are direct tweets from the accounts you interact with most, making the section relatively small. Finally, it culminates to the never ending scroll of tweets that are made directly from your followers.
Twitter’s algorithm is not only primarily on the accounts you follow. What the algorithm chooses to directly feed into your timeline is based on tweets you like, what you click on to read replies, what trends you explore, and probably many more. Your behavior then translates to your timeline, based on specifics of the tweet, such as the interactions on the tweet, the additional media, and the account’s overall activity. In general, posting a tweet with a picture or short clip is most likely to get into your timeline than a tweet without. If this is an active person who tweets several times a day, it will also probably show up on your timeline, not just because of the sheer amount of tweets.
In personal usage, Twitter has a decent algorithm set up. Usually, with one scroll through of Twitter, I find most of the accounts of follow as well as the news posted that day. And usually, within thirty minutes, the feed has a whole new set of things to show.
YouTube in a way is very “sinister” about the way they keep you indulging in their site. A previous algorithm, similar to Twitter, tunneled content onto your timeline based on your previously watched videos and newness of the video, as one would think. This included an A.I that served to find “adjacent relationships”, or something that humans would never put together,, and created a neural network and simulated that to what you know as your “Up Next”. While this seemed promising, their main goal was to keep the user on the platform longer, rather than just recommend another “good” video.
I think it is safe to say we have all experienced the YouTube “rabbit hole”, where you were just watching your favorite music video, then watching men build an underground pool out of almost nothing. This is the new algorithm, using A.I called “reinforced learning”, designed to create addictive options. Rather than catering to their tastes, it aims to create avenues of interests, hence why they are so out of the blue but oddly compelling.
Similar to twitter, your personal feed is curated by your comments, likes, and reshares. One thing unique about Instagram, is its priority to keeping a close relationship with specific followers. Not only is it you that they take into account, but also how much another follower interacts. Meaning, they take into account the photos you are tagged in by them, as well as comments and Instagram story mentions.
I explored this myself, and while Instagram does not have the “Family” option, it assumes my importance by their interactions with me. I do not use Instagram often, but it has already curated to show my sister first in terms of posts and story uploads. This is probably because on days like my birthday, pictures of our holidays photos, or just recent hangouts, the algorithm took that into account, and considered that as a “close friend/family”. This goes for my friends who have tagged me in their stories as part of challenges or just old memories.
However, like many technologies, there are mishaps and miscalculations within the system. Just like how all aforementioned platforms cater to your behavior, it can almost cater too well, and not know how to reel it in. Titled Dear Tech Companies, I Don’t Want to See Pregnancy Ads After My Child was a Stillborn, Gillian Brockell wrote to The Washington Post in December 2018 of her unfortunate experience with the algorithms in social media. She describes her journey through her pregnancy shared across her social media, ultimately saying how “silly” she was to even do that. The soon-to-be mother posted on Facebook and Instagram of her baby bump, even having clicked a few pregnancy ads. Later one, what the algorithm on all these sites did not pick on, was the unfortunate fact that she had a stillbirth.
“But didn’t you also see me googling “braxton hicks vs. preterm labor” and “baby not moving”? Did you not see my three days of social media silence, uncommon for a high-frequency user like me? And then the announcement post with keywords like “heartbroken” and “problem” and “stillborn” and the 200 teardrop emoticons from my friends? Is that not something you could track?”
Gillian Brockell practically picks apart and villainized the algorithm and tech companies. She understood its versatility in analyzing behavior, and even called out its inability to pick up key words and patterns of activity. This brings into question over the last two years since this article was posted, how the algorithm has changed and evolved, and if situations like this have alleviated. Will the algorithm reach a point of perfection, but then how comfortable would people be, knowing that even their social media behavior is being watched?
