Toward the Pluriversal Machine
A pluriversal AI would differ structurally from the current AI architecture along four dimensions Escobar's framework identifies as constitutive of any pluriversal practice. First, the pluralization of design — distributing design authority beyond the concentrated corporate contexts where foundational decisions are currently made. Second, the pluralization of knowledge — expanding training paradigms to include forms of knowing the current paradigm cannot capture. Third, the pluralization of evaluation — developing criteria that reflect the priorities of the diverse communities the tools claim to serve. Fourth, the pluralization of governance — creating institutional mechanisms that give affected communities genuine decision-making power over AI systems.
In the AI Story
The 2024 paper 'From Singularity to PlurAIverse,' presented at the ACM's Halfway to the Future Symposium, proposed design principles for AI development drawn explicitly from Escobar's Designs for the Pluriverse. The paper was not the work of postdevelopment theorists applying their framework to an unfamiliar domain. It was the work of technology researchers who had concluded, from within the discipline of human-computer interaction, that the dominant model of AI development was producing systems whose universalist assumptions were generating systematic failures when deployed across diverse cultural contexts.
The pluralization of design means more than user feedback surveys or diversity initiatives. It means genuine participation in the foundational decisions: which knowledge systems are represented in training data, which languages are supported at what level, which workflows are optimized for which contexts, which problems the tool is designed to solve. These decisions are currently made by AI companies' internal teams, informed by market signals that privilege commercially significant user populations. Genuine democratization would redistribute decision-making power to include communities whose lives the tools affect.
The pluralization of knowledge requires data sovereignty — frameworks that recognize communal knowledge as a contribution entitled to recognition, compensation, and governance rights. Indigenous ecological knowledge, traditional cultural expressions, locally produced content in languages whose speakers the models' governance structures do not include — all of this currently enters the training pipeline as raw material rather than as intellectual contribution. Indigenous intellectual property movements have developed legal instruments for the protection of traditional knowledge in other domains. Extension to the AI domain is technically feasible; it is politically absent because the current distribution of power does not require it.
The practical starting points are not hypothetical. Community radio networks in Latin America and Africa demonstrate how communications technology can be governed by the communities it serves. Participatory mapping projects demonstrate how digital technology can be appropriated for community-defined purposes. Fab labs and maker spaces demonstrate how the relationship between design and context can be structured to serve local purposes rather than global markets. Each operates at a scale far smaller than the AI transition requires, but each demonstrates a principle that scales: technology serves human purposes most effectively when the humans whose purposes it serves participate in its design and governance.
Origin
The concept emerged from the convergence of Escobar's pluriversal framework with critical AI research, particularly the 2024 ACM paper 'From Singularity to PlurAIverse' and the 2020 'Decolonial AI' paper by Shakir Mohamed and colleagues.
It represents Escobar's constructive contribution to the AI debate — the horizon toward which his critique of the current architecture points.
Key Ideas
Not a single alternative. The pluriversal machine is a landscape of technologies, not a better version of the current architecture.
Four dimensions. Design, knowledge, evaluation, and governance must all be pluralized for the transformation to be substantive.
Existing precedents. Community radio, participatory mapping, cooperative ownership, and data sovereignty frameworks demonstrate feasibility at smaller scales.
Autonomy in relation. The goal is not local isolation but community capacity to engage with wider networks on self-defined terms.
Political, not technical. The obstacles are obstacles of power, not of technical feasibility.
Appears in the Orange Pill Cycle
Further reading
- Sareeta Amrute, 'From Singularity to PlurAIverse,' ACM Halfway to the Future Symposium (2024).
- Shakir Mohamed, Marie-Therese Png, and William Isaac, 'Decolonial AI,' Philosophy & Technology 33 (2020).
- Arturo Escobar, Designs for the Pluriverse (Duke University Press, 2018).
- Sasha Costanza-Chock, Design Justice: Community-Led Practices to Build the Worlds We Need (MIT Press, 2020).