Source · Tier B · PreprintSource · Computing Power and the Governance of Artificial Intelligence
Computing Power and the Governance of Artificial Intelligence
G. Sastry, L. Heim, H. Belfield, M. Anderljung, M. Brundage, J. Hazell, C. O'Keefe, G. K. Hadfield, R. Ngo, K. Pilz, G. Gor, E. Bluemke, S. Shoker, J. Egan, R. F. Trager, S. Avin, A. Weller, Y. Bengio, D. Coyle. 2024. arXiv.
| Original | https://arxiv.org/abs/2402.08797 |
|---|---|
| DOI | 10.48550/arXiv.2402.08797 |
| arXiv | 2402.08797 |
| Accessed | 2026-10-20 |
| Authored or published by | Oxford Martin AI Governance Initiative, Centre for the Governance of AI |
| Imported from | hodgkins-ai-verification-papers@c71e59ff0e8e |
| Note | Listed under "Motivations and policy proposals" in the Hodgkins bibliography (CC BY 4.0). Also listed on the Oxford Martin AI Governance Initiative's publications page (https://aigi.ox.ac.uk/publications/computing-power-and-the-governance-of-artificial-intelligence/, read 2026-09-24). Also presented as a GovAI research paper (https://www.governance.ai/research-paper/computing-power-and-the-governance-of-artificial-intelligence, read 2026-09-24). |
Cited by
- Cap frontier trainingGoal
- Chip registries and manufacturing recordsMechanism
- Communication between compute groups is boundedClaim
- Compute stock is at most a declared amountClaim
- This compute runs inference, not trainingClaim
- There is no undeclared relevant computeClaim
- A training run stayed within declared limitsClaim
- CompartmentalizationConcept
- FLOP accountingConcept
- Inference and training workloadsConcept
- Interconnect bandwidthConcept
- Undeclared computeConcept
- Centre for the Governance of AIOrganization