A procedural composition of rules, data, and infrastructure that performatively shapes what becomes sayable, writable, and sensible within socio-technical assemblages.
In posthumanist and new materialist terms, algorithms are not neutral instructions but lively components of assemblages that co-produce phenomena. Drawing across Amoore on algorithmic governance, Seaver on “algorithms-as-culture,” Gillespie on platform curation, Crawford on infrastructures, and Rouvroy on algorithmic governmentality, algorithms appear as material-discursive practices that enact classifications, affordances, and exclusions. They operate through training data, prompts, weights, thresholds, and feedback channels; their performativity emerges with datasets, interfaces, sensors, clouds, and users in situated apparatuses.
Resonant with Deleuze and Guattari’s machinic assemblages and Bennett’s vibrant materialism, algorithms distribute agency across model architectures, datasets, interfaces, and human practices, while Ahmed’s and Massumi’s insights foreground how they also channel affective atmospheres, regulating attention, anticipation, and intensities in use. In this register, algorithmic outputs function as events that modulate the conditions of thought and composition, entangling matter and meaning rather than representing a prior reality.
Treating algorithms as part of the apparatus reorients inquiry from interpreting outputs to tracing the apparatus’s agential cuts and consequences: how prompts, model checkpoints, content filters, and UI affordances configure what becomes available as “writing,” “voice,” or “evidence.” Postqualitative inquiry composes diffractive encounters among logs, prompts, drafts, policy statements, and interface changes to map how algorithmic practices enact boundaries, authorize certain moves, and foreclose others. Methods shift toward attunement to model affordances, prompt ecologies, and infrastructure histories, foregrounding situated ethics and accountability distributed across human and nonhuman participants.
In AI-entangled writing instruction, algorithms participate in the emergence of authorship by patterning stylistic variation, genre coherence, citation habits, and error profiles. Attention to prompt design, temperature and constraint settings, moderation layers, and fine-tuning histories clarifies how creativity and originality are enacted within specific classrooms and platforms. Instructors’ lived and affective experiences—trust, friction, surprise, apprehension—are shaped by algorithmic legibility and opacity, from explainability dashboards to sudden model updates that reconfigure expected outputs. Pedagogical artifacts such as assignment prompts, rubrics, and feedback scripts become apparatus elements that, with the interface and model affordances, enact particular definitions of learning and accountability. Tracing revisions across student drafts, system logs, and instructor feedback makes visible the distributed agency of composing processes where the algorithm is a co-composer whose contributions are conditioned by infrastructure, policy, and situational use.