Movie: Humans in the Loop

 Humans in the Loop: AI, Situated Knowledge, and the Politics of Representation

Introduction

Directed by Aranya Sahay, Humans in the Loop (2024) offers a critical realist exploration of artificial intelligence not as a futuristic abstraction, but as a socio-political system grounded in human labour and cultural assumptions. Set in Jharkhand, the film follows Nehma, an Adivasi woman employed in a rural data-labelling centre where she annotates images to train AI systems. Through her lived experience, the narrative interrogates the ideological foundations of algorithmic knowledge and foregrounds the epistemic tensions between indigenous ecological understanding and corporate technological rationality.

Rather than portraying AI as neutral or autonomous, the film demonstrates that technological systems are deeply embedded in cultural frameworks, power relations, and representational politics. It exposes algorithmic bias as socially produced rather than technically accidental and reveals how epistemic hierarchies determine whose knowledge is legitimised within digital infrastructures.

I. Algorithmic Bias as Culturally Situated

1. Challenging the Myth of Technological Neutrality

A dominant discourse surrounding AI presents algorithms as objective, data-driven, and free from human prejudice. This ideology of neutrality aligns with what film theorists describe as the naturalisation of ideology—a process by which historically constructed systems appear inevitable or apolitical.

Humans in the Loop dismantles this myth by foregrounding the labour that produces algorithmic intelligence. The AI systems Nehma trains are not self-learning entities; they depend on human annotation shaped by predefined taxonomies. These taxonomies are determined not by local contexts but by distant corporate clients. Thus, the film reveals that AI knowledge is mediated and constructed rather than discovered.

2. The Insect Classification Sequence: Representation and Reduction

A significant narrative moment involves Nehma labelling insects as either “pests” or "non-pests". For her, such binary categorisation conflicts with indigenous ecological knowledge, which understands insects within a network of relational interdependence. The AI system, however, demands rigid classification aligned with commercial agricultural priorities.

From a film studies perspective, this moment illustrates representation as ideological construction. As Stuart Hall argues, representation does not merely reflect reality but actively produces meaning through discursive frameworks. The AI’s dataset represents nature through capitalist productivity logic—organisms are valuable or harmful based on market utility. The system’s bias is therefore not a coding error but a reflection of cultural ideology embedded in data structures.

The film visually reinforces this critique through the contrast between lush forest imagery and sterile computer screens. The mise-en-scène establishes a dialectic between plural, lived knowledge and singular, algorithmic reduction. The screen becomes a site where complexity is compressed into quantifiable categories.

3. Algorithmic Bias as Power

Drawing on Michel Foucault’s concept of power/knowledge, the film suggests that what counts as “data” is shaped by institutional authority. The corporate AI system determines legitimate classifications, thereby disciplining alternative epistemologies. Nehma’s hesitation is framed not as intellectual failure but as resistance to epistemic violence—the erasure of contextual knowledge in favour of standardised categories.

Thus, algorithmic bias is culturally situated:

• It arises from particular economic priorities.

• It reflects dominant scientific paradigms.

• It privileges certain worldviews over others.

The narrative makes clear that technology does not transcend culture; it operationalises it.

II. Epistemic Hierarchies: Whose Knowledge Counts?

1. Indigenous Knowledge vs Corporate Technoscience

The film presents a clear epistemic hierarchy between indigenous ecological knowledge and corporate technological authority. Nehma’s understanding of nature, shaped by lived experience and community memory, is relational and contextual. However, within the AI training system, her knowledge has no formal legitimacy. The system privileges standardised corporate guidelines over local interpretation.

By insisting on strict adherence to client-defined categories, the supervisors reinforce the dominance of quantifiable data and global market logic over embodied, environmental knowledge. Through this contrast, the film illustrates epistemic injustice—the systematic devaluation of marginalised knowledge systems within technological structures.

2. Gendered and Classed Digital Labour

The narrative further complicates epistemic hierarchy by highlighting the gendered and classed nature of data labour. The workers training AI systems are predominantly women from marginalised communities. Their labour is essential yet invisible.

From a Marxist film theory perspective, this reflects commodity fetishism: the AI appears autonomous and intelligent, while the human labour sustaining it is concealed. The global tech economy extracts cognitive labour from the Global South while symbolic authority remains with Western corporations.

The film’s realist aesthetic—long takes, subdued performances, attention to mundane routines—exposes this hidden infrastructure. By focusing on labour rather than spectacle, the narrative demystifies AI and relocates intelligence within human bodies.

III. Representation, Ideology, and Cinematic Form

1. Realism as Political Strategy

Unlike mainstream representations of AI in science fiction cinema, Humans in the Loop adopts a social realist style. This aesthetic choice is ideologically significant. It reframes AI as part of everyday socio-economic reality rather than technological fantasy.

The juxtaposition of village life and digital workspace visually encodes ideological conflict. Forest sequences emphasise fluidity, relationality, and plurality, whereas computer interfaces signify rigidity and control. The camera’s lingering attention to Nehma’s contemplative pauses foregrounds the cognitive and ethical labour erased by algorithmic efficiency.

2. The Politics of Knowledge Transmission

The relationship between Nehma and her daughter introduces an intergenerational dimension. The younger generation is more comfortable with digital systems, symbolising the normalisation of technocratic epistemology. Yet the film resists a simplistic binary between tradition and modernity. Instead, it asks whether technological systems can accommodate plural knowledge forms.

This question aligns with postcolonial film theory, which critiques the universalisation of Western epistemologies. The AI system operates as a colonial archive—classifying and ordering the world according to external priorities.

---

IV. Technology as Ideological Apparatus

Drawing from Louis Althusser’s theory of ideological state apparatuses, technology in the film functions as a mechanism through which dominant ideologies reproduce themselves. The data-labelling centre becomes a site of ideological interpellation, where workers internalise the authority of algorithmic logic.

However, the film also stages moments of subtle resistance. Nehma’s reflective pauses and ethical questioning disrupt the seamless flow of data production. These moments reveal that human agency persists within technological systems, even if constrained by structural hierarchies.

The title itself is ironic: humans are indeed “in the loop", but not equally empowered within it. Some design and own the systems; others annotate and correct them.

Conclusion

Humans in the Loop offers a nuanced critique of the relationship between AI and human knowledge. By situating algorithmic processes within rural labour networks and indigenous epistemologies, the film challenges the ideology of technological neutrality. It demonstrates that algorithmic bias is culturally embedded, shaped by economic priorities and dominant representational frameworks.

Furthermore, the film foregrounds epistemic hierarchies within technological systems, revealing how corporate technoscience marginalises local and embodied knowledge. Through its realist aesthetics and focus on invisible labour, the narrative exposes AI as a political structure sustained by gendered and classed power relations.