Principles #

AIINFRA is an exploratory project that will identify and implement best practices for the design, development, assessment, and maintenance of research tools enabled by Large Language Models (LLMs).

We will be designing and building a prototype tool using digital humanities and research software engineering (RSE) methods. We are as interested in identify governance models and frameworks capable of guiding those activities as the underlying technologies, however.

This page will be updated as the project evolves. We expect the following principles and guidelines to anchor our work:

• DAIR Institute Slow AI
• CARE Principles for Indigenous Data Governance
• FAIR Principles for Research Data Management
• The various initiatives listed at the Global Index on Responsible AI