The Architecture of Inclusion

In the AI Story

Allen's concept exposes a gap between the democratization narrative and the democratization reality that is wider than most commentary acknowledges. The narrative says: AI tools lower the barrier to building; anyone with an idea can now produce working software. The reality says: the tools are designed for a specific population, embedded in specific infrastructure, governed by specific assumptions about who the user is.

The architecture of AI access operates across four dimensions that Allen's framework makes visible. Infrastructure: AI tools require computational power, reliable electricity, high-speed connectivity, and capable devices that are unevenly distributed globally. Billions of people lack these prerequisites. Language: large language models were trained predominantly on English-language data and work best in English, reflecting English-language categories of thought. For the 6.3 billion people who do not speak English natively, this linguistic bias is a structural barrier to full participation. Culture: the tools embed the values of their developers—primarily the technology industry of the American West Coast—in their defaults and optimization targets. The aesthetic of smoothness is not a universal human value but a particular cultural preference. Governance: access to tools without participation in their governance is the condition of the consumer, not the citizen.

Each dimension requires its own architectural response. Infrastructure demands public investment in the physical and digital backbone of AI access. Language demands AI systems genuinely multilingual at the cognitive level, not merely at the surface of translation. Culture demands configurability—tools that can be tuned to reflect different cultural values and governance traditions. Governance demands participatory mechanisms that give affected communities genuine authority over how the technology operates in their contexts.

The architecture of genuine inclusion does not build itself. It must be designed, funded, and maintained through the sustained political commitment of democratic societies that take their own principles seriously enough to invest in their realization. The contest over whether this architecture will be built is, in Allen's words, 'a historical contest over what framework of political economy is going to define the world as AI transforms it.'

Origin

Allen's architectural concept of inclusion draws on the disability rights tradition and has been developed through her work on democratic theory and applied to AI governance through GETTING-Plurality and the Roadmap for Governing AI.

Key Ideas

Inclusion is architecture, not declaration. Genuine inclusion requires positive institutional design, not merely the absence of formal exclusion.

Four dimensions. Infrastructure, language, culture, and governance each require architectural response.

Infrastructure investment. Public provision of the physical and digital backbone is a democratic necessity.

Genuine multilingualism. AI must be multilingual at the cognitive level, not merely the translation surface.

Cultural configurability. Tools must be tunable to reflect diverse cultural values and governance traditions.