There is a global policy imperative to leverage artificial intelligence (AI) to drive economic growth. While AI may generate substantial economic value over the next decade, this value will not be evenly distributed or easily captured. Early signs point to a lack of understanding and planning around domestic AI compute capacity that is fueling a “compute divide” that will strangle innovation across governments, academia, startups, and industry.
In order to address these problems, policymakers must include “domestic AI compute capacity” in strategic planning and budget priorities. Doing so is a challenge, however, due to a lack of standards and definitions. The OECD recently established the AI Compute Taskforce to address this policy gap. US and European government leaders and policy experts must take additional steps to address the following questions:
- How much domestic AI compute capacity do we have?
- How does this compare to other nations?
- Do we have enough capacity to support our national AI ambitions?
Dr. Divya Chander
Faculty Chair, Neuroscience
Saurabh Mishra, PhD
Researcher and Manager, AI Index
Stanford Institute for Human-Centered Artificial Intelligence
Vice President, Worldwide AI Initiatives
Deputy Director and Senior Fellow, GeoTech Center