If open source is to win, it must go public

TL;DR

The paper argues that for open source AI to truly democratize access, it must be supported by public infrastructure and institutions that address resource and governance challenges.

Contribution

It highlights the necessity of combining open source AI with public infrastructure to ensure accessibility, sustainability, and governance in AI deployment.

Findings

Open source AI faces resource and governance challenges.

Public infrastructure is essential for democratizing AI.

Complementing open source with public institutions enhances AI accessibility.

Abstract

Open source projects have made incredible progress in producing transparent and widely usable machine learning models and systems, but open source alone will face challenges in fully democratizing access to AI. Unlike software, AI models require substantial resources for activation -- compute, post-training, deployment, and oversight -- which only a few actors can currently provide. This paper argues that open source AI must be complemented by public AI: infrastructure and institutions that ensure models are accessible, sustainable, and governed in the public interest. To achieve the full promise of AI models as prosocial public goods, we need to build public infrastructure to power and deliver open source software and models.