Algorithmic Gestation: Inference, Stages and Prescription in the Organisation of Biographical Time

Algorithmic personalisation holds a central place in platform studies, where it has been described as the outcome of classification, profiling and adaptation to inferred preference. That vocabulary proves insufficient once what the system attributes is no longer a stable property of the subject but a condition with a beginning, an internal articulation into stages and an end, since the platform must then hold a model of how the attribution will change and distribute content in accordance with it. My focus falls on this form of temporal organisation, and I develop the concept of algorithmic gestation through five jointly necessary features: inference of an emerging biographical state, stage-sensitive temporal sequencing, progressive stabilisation of an algorithmic identity, anticipatory presentation of practices and needs, and normative prescription indexed to the projected trajectory. The argument rests on a nine-month digital autoethnography carried out during my own pregnancy, in which I documented the reconfiguration of an Instagram feed, and it sits within a feminist STS genealogy that treats obstetric visualisation, prenatal diagnosis, self-tracking and intensive mothering as layers which accumulate rather than succeed one another. The material proves consistent with four of the five features and required revision of the fifth, since stabilisation of identity appears to operate as subordination of earlier configurations rather than as their replacement. I argue that algorithmic inference constitutes the most recent layer of reproductive governance, set apart by the absence of a threshold, by the absence of disclosure and by its address to a subject identified in advance of any account she gives of herself, and that platform power extends beyond the management of visibility to the organisation of biographical time.