A 14-year-old watches a video about astronomy because a friend mentioned black holes at lunch. Within a day, their feed shifts. More space content. Then science broadly. Then philosophy videos. Then content about feeling different, about being the “deep thinker” who doesn’t fit in.
Within weeks, they have a new identity. Not one they chose through genuine exploration, but one that was assembled around them by a system optimizing for engagement.
If you teach middle or high school, this is happening to your students right now. Understanding how can change the way you approach identity formation, peer dynamics, and the surprising rigidity you’re seeing in kids who should be in their most exploratory phase.
Adolescence has always been a period of trying on different selves. Students experiment with values, test different social groups, explore interests, and gradually figure out who they are. That process was always shaped by family, neighborhood, and available cultural narratives, but it had friction. Trying on a new identity required effort: you had to seek out new music, find different friend groups, have awkward conversations. That friction was essential. It meant identity formation involved agency, genuine encounter with the unfamiliar, and real choice.
Recommendation algorithms have eliminated most of that friction. A student engages with content, the algorithm serves more of the same, and the student engages again, not because they have deep affinity for it, but because it was placed in front of them and held their attention. Over time, the algorithm builds a model of who the student is based on behavioral data and serves content that reinforces that model. The student starts seeing themselves reflected in their feed. It feels like the algorithm “gets” them.
But the algorithm doesn’t know them. It knows a statistical model of their behavior. And it reflects that model back with such consistency that it starts to feel like truth.
The clearest sign is surprising rigidity. Students seem locked into an identity at 13 that feels too solid, too finished. They’ve decided they’re “the artist” or “the gamer” or “the activist” with a certainty that doesn’t match their developmental stage. When you suggest they try something outside that lane, the resistance is intense. That’s not normal adolescent stubbornness. That’s an identity that’s been algorithmically reinforced hundreds of times a day. Stepping outside it feels like betraying their authentic self, when really they’re just stepping outside their engagement profile.
There’s also an echo chamber effect. A student gets interested in social justice. Within days, their entire feed is social justice content, stripped of nuance and counterarguments. By the time they’re discussing it in your class, they’re not exploring an idea. They’re defending a position that feels like core identity because it’s been reinforced so consistently.
You’d expect kids with strong online identities to be confident. Instead, you’re seeing brittleness. Students who present as certain and self-assured online fall apart when that identity gets questioned in person. The algorithm doesn’t challenge you; it gives you more of what you already engage with. When these students encounter genuine friction in your classroom, they don’t have the psychological muscle to handle it. Their sense of self hasn’t been tested. It’s been curated.
The algorithm markets itself as helpful. It learns your preferences and shows you relevant content. But adolescents need exposure to difference. Identity forms not just through affirmation but through friction, through encounters with perspectives that force you to clarify what you actually believe. Real identities are tested, questioned, and complicated.
Algorithmic identity tends toward coherence without complexity. A clean narrative assembled from engagement data instead of the messy, contradictory process of real self-discovery. Your students aren’t getting that mess anymore. They’re getting streamlined, optimized, algorithmically confirmed versions of themselves. And then you’re asking them to write nuanced essays, consider multiple perspectives, and tolerate ambiguity. No wonder they’re struggling.
How to Push Back in Your Classroom
The most direct tool you have is algorithmic literacy. Most students don’t know how recommendation systems work. When you explain that their feed isn’t showing them reality but a prediction of what will keep them engaged, something shifts. Try this framing: “The algorithm doesn’t know who you really are. It knows what you’ve clicked on. Those aren’t the same thing.”
Assign deliberate exposure to difference. Design assignments that require students to engage with perspectives outside their comfort zone, framed as genuine exploration. Use class projects as low-stakes opportunities to try on different roles: the kid who’s “the athlete” writes from the perspective of an artist, the “social justice kid” argues an opposing view, not to change their mind but to understand the logic. Give students this assignment: screenshot your feed for three days. What pattern do you see? What’s missing? Then ask: “Is this who you want to be, or is this just what the algorithm thinks will keep you watching?”
Normalize contradiction. When a student holds two conflicting ideas, don’t make them pick one. “You can believe both of those things. People are complicated. Your algorithm wants you to be consistent. Real humans aren’t.” This is countercultural to their digital environment, which rewards clarity and punishes nuance.
If you understand that your students’ identities are being significantly shaped by algorithmic curation, your classroom looks different. You’re not just teaching content. You’re providing one of the few spaces in their lives where they encounter genuine friction, real difference, and the possibility of becoming something other than what their engagement profile predicted.
At the end of a unit, instead of asking “What did you learn?” try asking: “What idea made you uncomfortable? What perspective challenged something you were sure about?” If students can’t answer that, the unit stayed inside their algorithm. You didn’t push them into new territory.
The 14-year-old who stumbled into a new identity because a black hole video showed up at lunch deserves the chance to discover who they actually are. You can’t undo the algorithm. But you can create space where students learn that who they are is bigger, messier, and more interesting than who the algorithm says they should be. That’s not just good teaching. That’s exactly what teaching is for right now.
EdTechInstitute provides practical resources for K-12 educators navigating technology, AI, and the changing realities of the classroom.
Related Reading
- How Algorithms Are Shaping Your Students’ Identities (And What Teachers Need to Know)
- Teaching Students to Understand How Algorithms Shape Identity (Before It’s Too Late)
- The Profile Isn’t the Person: Helping Students Form Authentic Digital Identities
- When Algorithms Become Your Students’ Mirror (And Why It Matters for Learning)
- Teaching Students to Understand How Algorithms Shape Identity (Before It’s Too Late)
Try This Free Tool
RazaEd offers free AI-powered literacy tools for K-12 teachers, including differentiated reading passages, comprehension questions, and vocabulary activities for any grade level.