Literature

Our recent paper “Computing Power and the Governance of AI” is an introduction to the field of Compute Governance. We argue that compute is a particularly promising node for AI governance. We also highlight the risks of compute governance and offer suggestions for how to mitigate them. You can find a Twitter summary here and blog post summary here.

Beyond this paper, I’d recommend my blog post that motivates Technical AI Governance and our paper “Governing Through the Cloud: The Intermediary Role of Compute Providers in AI Regulation” that introduces the governance capacities that compute providers could enable.

Reading List

Transformative AI and Compute - Reading List, where I list readings and resources in the following categories:

  • Compute in the AI Production Function: Why compute matters for AI
  • Compute Supply Landscape: How chips are produced and used
  • Compute Governance: The governance of compute to achieve beneficial AI outcomes
  • Compute-Based Transformative AI Forecasting: How much compute might lead to transformative and potentially dangerous capabilities

Talks

You can find a video and transcript of an Introduction to Compute Governance on my blog where I introduce the idea of using compute as a node for AI governance.

In August 2023, I gave an updated version of this talk at Google. You can find the the slides, a transcript, and a recording here.

Podcast Interviews

In May 2023, I was interviewed by Rob Wiblin from 80,000 Hours. We talk for three hours covering a wide range of topics. You can also find a transcript at the link below.

Sean Lawrence from High Impact Engineers interviewed me. While we also cover AI and Compute, we also chat about my career and how I came about working on AI Governance.

I did two podcast episodes with the Future of Life Institute:

I also joined Jordan Schneider and Chris Miller on ChinaTalk:

Career Guide

I recommend this career guide on becoming an expert in AI hardware.

The following article is also a great motivation for working on the intersection of technical and governance problems “The case for taking your technical expertise to the field of AI policy.”


“All models are wrong, but some are useful."

— George E. P. Bo


updated Feb 2024