How can large language models encompass more situated, embodied, and tacit ways of knowing?
This collaborative project with Bernie Boscoe and Chandler Campbell from SOU explores how science labs can make use of a rag llm to help make their work more durable, actionable, shareable, and discoverable.
What kinds of knowledge and sensory modalities are currently incorporated into large language models, and what are excluded? Aquillm is a place-based AI "archive" that aims to find out. Of particular interest is how sonic data from the field or from the lab helps us make sense of our research world.