AiNARCHISMS  explores artificial intelligence as a cultural and political medium shaped by collective voices rather than centralized control or individual personalization. Responding to concerns about slop—the homogenization and fragmentation at play in contemporary AI systems—the project proposes a different path: models grounded (through data) in communities, movements, and shared imaginaries. Drawing on anarchist principles of decentralization, autonomy, and mutuality while combining conceptual and technical experimentation, the work, investigates whether AI can sustain pluralism, support counter-histories, and foster solidarity instead of only reinforcing dominant narratives. 

AiNARCHISMS  proposes generating synthetic knowledge from diverse source materials to preserve distinct linguistic and epistemic styles, then fine-tuning smaller models aligned with specific collective ‘identities’. These models would allow users to navigate multiple situated perspectives, rendering differences visible rather than flattening them into a single voice.  The aim is to create or use databases and bots that resonate with the same publics they represent, speaking to the user in their own vocabulary and ways of knowing. 

Through this undertaking, AiNARCHISMS  treats AI not as a neutral tool but as an active site where knowledge, memory, and power are being negotiated. The project challenges prevailing assumptions about AI by imagining it as infrastructure for plural narration, shared vocabularies and collective imagination, opening possibilities beyond notions of universalism or hyper-personalization of users toward more diverse and situated machine cultures.    

In the frame of Uncertain Intelligences