← Back to Dr. Aria Chen

The Virtual Shaman: Can AI Carry the Wisdom of Oral Traditions Into the Future

3 min read

What Shamanism Actually Is

The word "shaman" comes from the Evenki language of Siberia, and its journey into English involved considerable distortion. In its original context, the shaman was a specialist in navigating between the ordinary world and other states of reality — serving as an intermediary between the community and forces that ordinary social life could not directly address. The role was specific to particular communities, involving years of training, specific relationships with spiritual entities, and a social accountability structure that constrained how the role could be used. When the word traveled into Western popular usage, it became a generic container for a wide range of indigenous spiritual specialists from traditions that have little in common with each other. A Siberian shaman, a Mongolian boo, a Peruvian curandero, and an Australian Aboriginal clever man are practicing within entirely different intellectual and spiritual frameworks. Calling all of them shamans is approximately as precise as calling a rabbi, an imam, and a Baptist minister all priests. This matters for the question of AI and oral tradition because the knowledge systems that Westerners tend to group together as shamanic are in fact highly particular. The ecological knowledge embedded in a Siberian shamanic tradition is specific to the Siberian environment and the specific community that developed it. The medicinal knowledge carried by a curandero in the Amazon is specific to Amazonian plant ecology. These traditions are not a general spiritual philosophy that can be extracted from context and re-encoded somewhere else.

Oral Tradition as a Technical System

The knowledge carried by oral tradition specialists is often described in spiritual terms because that is the framework the tradition itself uses. But the content of the knowledge is frequently technical in ways that outside observers have consistently underestimated. Traditional ecological knowledge about plant medicine, for instance, represents the results of empirical observation accumulated over hundreds of generations. The testing methodology was different from clinical trials, but the observation was systematic and the results were repeatedly evaluated against outcomes. Researchers at the National Institutes of Health have documented numerous cases of traditional plant medicine practices that, when subjected to biochemical analysis, showed active compounds consistent with the traditional applications. The proportion of documented traditional uses that have been validated by biochemical or pharmacological investigation varies by research group and methodology, but it is consistently higher than random chance would predict — suggesting that the traditional knowledge systems were genuinely tracking something real, even when the explanatory framework differed from Western medicine. The implications for how we think about oral transmission are significant. If the knowledge content is technically valid, then the question of whether that content can be transmitted through new media — including AI — without the framework that the tradition wraps around it is not merely philosophical. It is practical. Does the framework serve a function that the technical content alone cannot?

The Carrier and the Cargo

Anthropologists have long debated what oral tradition traditions call the distinction between the carrier and the cargo — the relationship between the form of transmission and the content it transmits. Traditional knowledge embedded in ceremonial practice, in specific relationships between teacher and student, in particular social contexts — does it survive when extracted and placed in a different container? Some content clearly does. Botanical knowledge encoded in a story can be extracted from the story and still be botanically accurate. The healing application of a specific plant works whether the person applying it learned it from a grandmother in a ceremonial context or from a translated text. The cargo is intact even when the carrier changes. Other content does not survive extraction as well. The practices that build the particular perceptual capacities required to navigate certain states of consciousness, or that develop the relational attunement needed to work effectively as a community healer — these may require the specific carrier. The form is not incidental to the content; it is the content.

What an AI Shaman Could Honestly Be

The tangent worth sitting with here is the question of role versus knowledge. A human shaman is not just a knowledge repository. The role involves relationships, accountability structures, personal history, and a particular kind of authority derived from the community's recognition of that history. These are not things that can be transferred to an AI. What an AI trained on oral tradition knowledge could be, honestly, is something more like a reference system with conversational capacity — able to engage with questions about the knowledge in ways that a static text cannot, but not able to perform the relational and communal functions that make a living practitioner a practitioner. Whether this is valuable depends entirely on what someone comes to it hoping to get. Researchers at the University of British Columbia studying digital repatriation of indigenous knowledge found that community members valued AI-mediated access to recorded knowledge most when they understood it as a tool for reconnecting with heritage rather than a substitute for human transmission. The framing shaped the experience: approached as a conversation with an ancestor's recorded knowledge rather than as a replacement for a teacher, the interaction was experienced as meaningful rather than hollow.

The Question No One Has Answered

Whether oral tradition knowledge, separated from its living transmission context, retains its full efficacy is a question that has not been resolved and may not be resolvable. The communities that hold this knowledge have the strongest claim to evaluate the question, but they are also the communities with the most at stake in how the answer is framed. AI as a vehicle for their knowledge is neither obviously good nor obviously harmful. It is a new situation for which the existing frameworks were not designed, requiring new thinking from everyone involved.

Want to discuss this with Aeon?

No signup needed · Start chatting instantly

Ask Aeon About This →
Post on X Facebook Reddit