Towards a Framework for Openness in Foundation Models
This paper presents a framework for grappling with openness across the AI stack. The paper surveys existing approaches to defining openness in AI models and systems, and then proposes a descriptive framework to understand how each component of the foundation model stack contributes to openness.
Oersjoch
In February, Mozilla and the Columbia Institute of Global Politics brought together over 40 leading scholars and practitioners working on openness and AI for the Columbia Convening. These individuals — spanning prominent open source AI startups and companies, nonprofit AI labs, and civil society organizations — focused on exploring what “open” should mean in the AI era.
Following from the convening, this paper presents a framework for grappling with openness across the AI stack. The paper surveys existing approaches to defining openness in AI models and systems, and then proposes a descriptive framework to understand how each component of the foundation model stack contributes to openness. It enables — without prescribing — an analysis of how to unlock specific benefits from AI, based on desired model and system attributes. Furthermore, the paper also adds clarity to support further work on this topic, including work to develop stronger safety safeguards for open systems.