Smithsonian deploys AI to digitize Revolutionary War artifacts while sidestepping colonial provenance and Indigenous erasure in museum collections
Original framing: “How the Smithsonian is using AI to connect artifacts from the American Revolution” — Phys.org
The Smithsonian holds over 19,000 Native American objects from the Revolutionary era, many acquired through battlefield looting or coercive treaties. AI tagging of these items using Western ontologies — 'weapon,' 'ornament,' 'tool' — erases Indigenous classifications like Anishinaabe 'manidoowigwisens' (spirit-helper objects) or Haudenosaunee 'gayogohó:nǫˀ' (wampum diplomacy records). Without tribal co-curation, the project digitizes dispossession.
The Smithsonian's AI project exemplifies what Ruha Benjamin calls the 'New Jim Code' — technological systems that inherit and automate colonial hierarchies under a veneer of innovation.
By framing digitization as neutral 'connection,' the institution obscures how its training data, labeler labor, and ontology design all reproduce the same power structures that made the Revolutionary War a war for white male property rights. The trickster sees clearly: the AI is not a bridge but a moat, widening the gap between the museum's 157 million objects and the communities from whom they were taken. True systemic repair requires not better algorithms but transferred authority — letting Eshu carry the messages back on terms the trickster's kin define.