Seeing the World with Two Eyes Does Not Mean Turning One Eye into an AI Lens
Original Chinese title: 雙眼看世界，不是把一隻眼睛做成AI鏡頭
Two-Eyed Seeing is not about putting Indigenous stories into science classes as warm-ups or feeding elder interviews to models to generate pretty PPTs. It requires students to simultaneously understand experiments, classifications, observations, ethics, relationships, and land experiences.
Aciang Iku-Silan
The Most Dangerous Slogans Are Often the Gentlest
Terms like "cross-domain," "co-learning," and "Two-Eyed Seeing" are so appealing that they sometimes make us forget to ask: Whose domain is being crossed? Who is co-learning with whom? Whose knowledge ends up in the center of a slide, and whose gets relegated to case boxes?
If Two-Eyed Seeing is reduced to lesson-plan decoration, it becomes vanilla powder for education: sprinkled lightly, it looks local, but tastes like the same standardized can.
Not Mixing, But Respecting Two Strengths at Once
The spirit of Two-Eyed Seeing is to see Indigenous Peoples' knowledge with one eye and Western science with the other, then look together. This is not translating myths into science or packaging science as myth; it acknowledges that different knowledge systems have distinct problem orientations, forms of evidence, and accountability logics.
The difficulty lies here. The education system's most habitual practice is to treat Indigenous Peoples' knowledge as case studies, stories, cultural supplements—still judged by a single standard for what counts as knowledge, evidence, or effectiveness. That is not Two-Eyed Seeing; it is one eye grading and the other being displayed.
How to Implement in AI Classrooms
AI education can practice three things.
First, have students compare plant classifications: on one side, biological taxonomy; on the other, Indigenous language names, uses, seasons, taboos, and landscape relationships. This is not about picking sides but understanding that classification is never just naming—it is also a relationship between people and their environment.
Second, have students analyze dream knowledge: not whether dreams are accurate, but how dreams form warnings, ethics, and bodily views across different societies. Dreams are not substitutes for scientific experiments, nor are they cultural trash cans to be discarded with one superstitious sentence.
Third, have students build small RAG systems: not aiming for the most answers, but knowing when to refuse, when to flag insufficient context, and when human review is needed. True AI literacy is not about using tools fluently; it is about knowing when they should not speak freely.
Assessment Should Not Only Test How Many Indigenous Names Students Can Recite
If a final exam only asks "Which tribe has which festival," that remains encyclopedia fill-in-the-blank. Better questions: Given an Indigenous traditional medicine narrative, can students distinguish cultural knowledge from medical risk, public accessibility, and scientific verification issues? When facing AI-generated images, can they spot mismatched attire, mixed ethnicities, ritual appropriation, and unclear sources?
This is the literacy of Two-Eyed Seeing. It requires students not only to follow prompts or memorize concepts but to remain alert, respectful, and judgmental across different knowledge systems.
AI Can Be a Teaching Assistant, But Not an Elder
AI excels at organizing, comparing, generating drafts, and helping students ask questions. Yet it cannot replace the memory location of elders nor decide how knowledge flows within communities.
Two-Eyed Seeing is not about making AI smarter; it is about making education systems more humble. If education treats AI only as an answer machine, students learn efficiency in laziness rather than knowledge; if education lets AI become a questioning tool, students may truly see the world behind two eyes.
AI use and content-safety disclosure
This article was reviewed by the author before publication; AI assisted with draft organization and formatting. Sections involving Indigenous Peoples' knowledge, curriculum design, and cultural context still require human review within community contexts.