The next speaker in this session at the IAMCR 2025 conference in Singapore is Jiayi Ge, presenting the algorithmic kaleidoscope framework to explore the algorithmic self of digital media users. Various other metaphors being used for this – the ‘algorithmic mirror’, and others – are problematic, as the algorithmic self does not simply reflect the user’s ‘real’ self; rather, there are various selves (actual, ideal, intended) which are variously represented by the algorithmic self.
How do users perceive this discrepancy, and how do they respond to it? The present study explored this through some 23 interviews with users in Singapore. They saw the algorithmic self as floating over the shorter and longer term – reflecting seasonal variations, as well as long-term personal development. The algorithmic self also misses some parts, as well as overemphasising others; there is an overlap, but this is not complete at all.
The algorithmic self might also be overly general, hardly different from others like them; this might be due to limitations to users’ open and full public expression; and as a result might also give the impression of interests that are not core to users’ identities. Self-presentation via digital media might also present more of an ideal self rather than the actual self. This might also be different across different digital and social media platforms; each presents only one slice of the actual self.
This can lead to poor recommendations from digital platforms, and result in feelings of inadequacy, exclusion, and social anxiety. This might then also result in context collapse, and uncomfortable perceptions of one’s own interests by others in one’s social network. But users may also result by making ill-advised purchases that seek to enhance their lifestyles, and generally be affected by negative feedback loops.
All of this means that reflected selves on digital media represent an algorithmic kaleidoscope, with each platform reflecting only a facet of a user’s identity, and frequent changes to representation.