On the Limitations of Compute Thresholds as a Governance Strategy

TL;DR

This paper critically examines compute thresholds as a governance tool in AI safety, arguing they are shortsighted and unlikely to effectively mitigate risks due to the uncertain relationship between compute and model capabilities.

Contribution

It provides a historical and conceptual analysis of compute thresholds, highlighting their limitations and proposing a need for more robust risk mitigation strategies.

Findings

Compute thresholds are shortsighted and likely to fail in risk mitigation.

The relationship between compute and AI capabilities is highly uncertain and dynamic.

Current policies relying on compute thresholds may overestimate predictability of AI risks.

Abstract

At face value, this essay is about understanding a fairly esoteric governance tool called compute thresholds. However, in order to grapple with whether these thresholds will achieve anything, we must first understand how they came to be. To do so, we need to engage with a decades-old debate at the heart of computer science progress, namely, is bigger always better? Does a certain inflection point of compute result in changes to the risk profile of a model? Hence, this essay may be of interest not only to policymakers and the wider public but also to computer scientists interested in understanding the role of compute in unlocking breakthroughs. This discussion is timely given the wide adoption of compute thresholds in both the White House Executive Orders on AI Safety (EO) and the EU AI Act to identify more risky systems. A key conclusion of this essay is that compute thresholds, as currently implemented, are shortsighted and likely to fail to mitigate risk. The relationship between compute and risk is highly uncertain and rapidly changing. Relying upon compute thresholds overestimates our ability to predict what abilities emerge at different scales. This essay ends with recommendations for a better way forward.