RECOMMENDED SELF

Extract—Validate—Rank— Repeat—Feedback —Self-Correct

This project explores how social media recommendation algorithms shape contemporary beauty standards and contribute to aesthetic homogenisation. The project examines how individual facial features are extracted, classified, ranked and repeatedly circulated as desirable visual patterns. Through recommendation, repetition and feedback, certain appearances gain greater visibility, gradually forming recognisable aesthetic templates. Rather than suggesting that algorithms directly determine how people look, the project focuses on how users actively adapt their self-presentation in response to these systems, adjusting their appearance to align with what is more visible, recognisable and rewarded. Through a series of six posters — Extract, Validate, Rank, Repeat, Feedback and Self-Correct — the project visualises this algorithmic cycle and questions how technological systems can amplify existing beauty standards while reducing visible diversity and encouraging forms of self-regulation.