Lily Hein / The Cougar
The most effective forms of control don’t feel like you’re in control at all. It feels like a recommendation. Social media platforms display buttons and engagement signals such as “not interested” that give users control over what content they want to see. Algorithms simplify decision-making by learning what we like.
But these systems do more than just cater to user preferences. They discipline it, limiting its room for reflection while quietly shaping what users have come to enjoy, value, and believe.
In this way, selection becomes increasingly passive. In this context, how can we claim to self-curate our feeds when so much of what we see is pre-determined by algorithms?
feedback loop
Degenerate feedback loops in recommender systems is a 2019 case study that analyzes the impact of feedback echo chambers and filter bubbles on users. The phenomenon of feedback loops is the widespread use of machine learning in recommendation systems.
Decisions made by these systems can influence users’ beliefs and preferences, which in turn influences the feedback the learning system receives, forming a feedback loop.
Feedback loops not only reinforce users’ preferences, but also limit users’ exposure to opposing views and unprecedented media. More broadly, curating our feeds changes what naturally interests us.
For example, discoveries used to come about through chance, curiosity, and engagement with outside ideas. Algorithmic curation now filters the world we interact with by presenting only ideas that resonate with us. In doing so, beliefs are reinforced before they are tested and choices are not fully exercised but are pre-formed by the machine.
The argument here is not that users are seeing things they don’t want to see, but if their choices are predicted in advance and the possibility of encountering the unexpected is designed in, the user becomes a partial product of the algorithm. This raises deeper questions about what freedoms are actually being exercised in algorithmically curated environments.
Free-ish
political philosopher Isaiah Berlin presents the concept of freedom as infinite choice and contrasts it with the definition of freedom as self-direction. First, in terms of freedom, media platforms seem to promote unlimited options, as you can participate in unlimited content and unsubscribe whenever you like.
Content is pre-filtered and recommended, increasing familiarity. Some may argue that the media we consume should be tailored to our tastes, and I agree that personalization definitely makes content more engaging and enjoyable. But the appeal of personalization masks a completely voluntary engagement.
In other words, tailoring a user’s content limits self-direction. Freedom does not involve choosing among presented options, but requires the ability to encounter unexpected options.
Algorithms hand-picked by recommendations or expected preferences cannot maintain these conditions. This is why feedback loops are inherently dangerous to the human psyche. Because the feedback loop is trained to suppress the desires and reflections on which freedom depends.
The antidote to negative choices
Philosophically speaking, freedom requires not just the availability of choices, but the contemplation of the desires that drive those choices. In our media-driven daily lives, there is little space for this kind of introspection. Users are comfortable becoming the media they consume, justifying it on the basis of familiarity or as a product of their own involvement.
But when engagement is scoped by design and curated into a “For You” page, what might have been a conscious decision to engage now becomes a passive choice.
I encourage my users, as well as myself, to be passive with media for long periods of time, consuming it without interacting with it. The possibility of discovering ideas you hadn’t thought of before, hearing from communities you’ve never met before, and finding inspiration from perspectives you’ve never seen before can be the ideal antidote to algorithmic curation.
opinion@thedailycougar.com
