IB Digital Society 4.7A Social Components of Identity Notes

4.7A Social Components of Identity

1. Organising concept is identity; digital systems shape how we build, show, and are labelled by age, gender and sexuality, race and ethnicity, religion, and disability.
2. Supporting concepts: expression (a platform can amplify or silence a voice) and power (who is centred and who is treated as an edge case).
3. Content topics that meet here: media and networks (self-presentation, community), data and artificial intelligence (inference, representation), and values and ethics throughout.

Presenting the Self Online

1. A profile is curated: chosen photos, words, and likes present an identity, under a real name or anonymously / a pseudonym.
2. Anonymity lets people explore identity safely and speak honestly. It can also shield abuse.
3. Context collapse: offline audiences (family, friends, colleagues) merge online, so one post reaches all at once.
   1. Impacts vs implications: new freedom to shape and share who you are, against pressure to perform, loss of separate spheres, and anonymity that can shield harassment.

Finding Community and Common Humanity

1. An online community can form around a rare condition, faith, heritage, or marginalised identity, letting people find others who share a part of themselves.
2. For a stigmatised or unseen young person this is a lifeline (information, role models, belonging); across borders it can build international-mindedness and common humanity.
   1. Contested: the same grouping can harden into an echo chamber, expose a vulnerable community to harm, or quietly narrow a worldview.

Identity by Algorithm

1. Algorithmic profiling: platforms infer identity from behaviour rather than asking, then sort people into groups to target content and ads.
2. From clicks, watch time, and connections a system estimates age, interests, and group memberships, sometimes guessing sensitive traits (ethnicity, sexuality) never disclosed, then boxes identity into a category.
   1. Power: the platform, not the person, defines the label. Values and ethics: inferring intimate traits without consent raises privacy and fairness questions.
   2. Impacts vs implications: more relevant feeds, against manipulation, discrimination (labels deciding who sees a job or housing ad), and a self shaped by algorithmic assumptions.

Representation, Bias, and Being Seen

1. Because they learn from past data, algorithmic bias and data bias make systems work well for some identities and poorly for others.
2. Facial recognition misidentifies women and people with darker skin more often (under-represented training data); Google Gemini image tool (2024) overcorrected stereotypes, produced historically inaccurate results, then paused.
   1. Values and ethics: being mis-seen or unseen harms dignity and access. Power: whoever trains the model decides which identities are default and which are the exception.

Weighing a Contested Balance

1. Two-sided: systems can help people explore identity, find community, and reach a world built for others. The same systems can stereotype, profile without consent, misrecognise, and exclude.
2. Outcome depends on who is in the training data, whether profiling is transparent and optional, and whether people can be seen on their own terms. Identity is negotiated between a person's own choices and the platform's inferences, so what the person shares, hides, or corrects still affects the result.