Algorithmic Identity: Does TikTok Know You Better?

I opened TikTok at midnight during finals week, planning to watch one video before studying for my chemistry exam. Two hours later, my phone knew I was anxious about organic chemistry, nostalgic for early 2000s cartoons, and weirdly interested in people fixing old furniture. I had never searched for any of these things. The algorithm had watched me pause on a video about test anxiety for three seconds longer than usual, noticed I replayed a clip of someone sanding a wooden chair, and decided it understood something about who I was.

My roommate says the same thing happens to her, except her feed is full of baking tutorials and videos about adopting senior dogs. We use the same app, but it shows us completely different versions of reality. This raises an uncomfortable question: if an algorithm can predict what I want to watch before I know I want to watch it, does that mean it understands me better than I understand myself?

I downloaded TikTok during my freshman year because everyone in my dorm had it. At first, I thought the app was just random videos, but within a week, my For You Page felt personalized in a way that seemed almost intrusive. I would think about wanting to learn guitar, and suddenly my feed filled with beginner guitar lessons. I never told the app I played guitar or wanted to learn. It just knew. This happens because TikTok tracks hundreds of data points: how long I watch each video, when I rewatch something, what I share, what I skip, even how I hold my phone. The algorithm builds a profile of my interests, my attention span, and my emotional responses. Unlike older social media platforms where I chose who to follow, TikTok decides what I see based on patterns I did not even know I had.

Last semester, I noticed my feed had shifted to videos about anxiety and college stress. I was going through a difficult time after failing my first calculus test, and I had watched a few videos about academic pressure. Within days, my entire feed became mental health content. Some of it was helpful, like study techniques and time management advice. But some of it made me feel worse, like videos of students crying in libraries or talking about dropping out. The algorithm had identified that I was struggling, and it kept showing me more of what I engaged with, even when that content reinforced negative feelings. A psychology professor I talked to later explained that algorithms optimize for engagement, not for my well-being. TikTok does not care if a video makes me feel better or worse. It only cares that I keep watching.

The strangest part is how accurately the algorithm predicts what I will find interesting, even when I would not describe those interests as part of my identity. I never thought of myself as someone who cared about architecture, but after watching one video about brutalist buildings, I started getting dozens of videos about unusual structures and urban design. Now I genuinely find architecture fascinating, and I have spent hours watching videos about it. Did TikTok discover an interest I always had but never explored, or did it create that interest by repeatedly showing me content until I started to care? My philosophy class discussed something similar when we read about how our preferences are shaped by what we are exposed to. If I only like something because an algorithm decided to show it to me hundreds of times, is that really my preference, or is it something manufactured?

Some researchers argue that these algorithms reveal truths about us that we hide or ignore. A friend of mine insists that TikTok helped her realize she wanted to switch her major from business to environmental science. Her feed kept recommending videos about climate change and conservation, and she started paying attention to why those videos resonated with her. She thinks the algorithm saw something in her behavior that she had been avoiding. Maybe she lingered on nature videos longer than she realized, or maybe she skipped past business content faster than other topics. In her case, the algorithm acted like a mirror, reflecting back interests she had not fully acknowledged. If TikTok can do this, then perhaps it does know us in ways we do not know ourselves, at least in terms of what captures our attention and stirs our curiosity.

I still use TikTok, but I try to stay aware that my feed is not a neutral reflection of who I am. It shows me an amplified version of my habits and reactions, optimized to keep me scrolling. Sometimes I deliberately search for topics outside my usual feed, just to remind myself that my interests are not limited to what an algorithm suggests. The app might know what I will watch, but it does not know why I watch it, or what I think about when I put my phone down. It cannot measure the moments when I question whether I actually enjoyed something or just consumed it out of habit. TikTok might predict my behavior better than I can, but behavior is not the same as identity. What I watch and what I value are not always the same thing, and remembering that difference matters more than I expected.

Did you like this example?

246

10