AI Governance (Q3 2023 Cohort)
| 872 words- Core Content (Q3 2023)
- Week 1: Intro to ML (100 minutes)
- Week 2: Introduction to potential catastrophic risks from AI (80 minutes)
- Week 3: Challenges in achieving AI safety (80 minutes)
- Week 4: AI standards and regulations (100 minutes)
- Week 5: Closing regulatory gaps through non-proliferation (100 minutes)
- Week 6: Closing regulatory gaps through international agreements (100 minutes)
- Week 7: Additional proposals (60 minutes)
- Week 8: Career Advice (75 minutes)
- Project Track - Field Building
Core Content (Q3 2023)
Week 1: Intro to ML (100 minutes)
- The AI Triad and What It Means for National Security Strategy by Ben Buchanan (2020)
- Visualizing the deep learning revolution (2023) by Richard Ngo
- But what is a neural network by 3Blue1Brown (2017)
- Gradient descent, how neural networks learn by 3Blue1Brown (2017)
- A short introduction to Machine Learning by Richard Ngo (2021)
- How ChatGPT Works Technically by ByteByteGo (2023)
Week 2: Introduction to potential catastrophic risks from AI (80 minutes)
- Overview of how AI might exacerbate long-running catastrophic risks by BlueDot Impact (2023)
- As AI agents like Auto-GPT speed up generative AI race, we all need to buckle up by Sharon Goldman (2023)
- The need for work on Technical AI Alignment by Daniel Eth (2022)
- Specification gaming: the flip side of AI ingenuity (2020) by Victoria Krakovna et alia
- Researcher perceptions of Current and Future AI by Vael Gates (2022)
Week 3: Challenges in achieving AI safety (80 minutes)
- Emergent Deception and Emergent Optimisation by Jacob Steinhardt (2023)
- AI Safety Seems Hard to Measure by Holden Karnofsky (2022)
- Compilation: Why Might Misaligned, Advanced AI Cause Catastrophe? by BlueDot Impact (2023)
- Nobody’s on the ball on AGI alignment by Leopold Aschenbrenner (2023)
- Avoiding Extreme Global Vulnerability as a Core AI Governance Problem by BlueDot Impact (2022)
Week 4: AI standards and regulations (100 minutes)
- Primer on Safety Standards and Regulations for Industrial-Scale AI Development by BlueDot Impact (2023)
- Frontier AI Regulation: Managing Emerging Risks to Public Safety by Markus Anderljung, Joslyn Barnhart, Anton Korinek et al. (2023)
- Model Evaluation for Extreme Risks by Toby Shevlane (2023)
- Safety Evaluations and Standards for AI by Beth Barnes (2023)
Week 5: Closing regulatory gaps through non-proliferation (100 minutes)
- Primer on AI Chips and AI Governance by BlueDot Impact (2023)
- Introduction to Compute Governance by Lennart Heim (2023)
- The State of AI in various countries - an Overview by Lizka Vaintrob (2023)
- Racing through a minefield: the AI deployment problem by Holden Karnofsky (2022)
- Winning the Tech Talent Competition by Remco Zwetsloot (2021)
Week 6: Closing regulatory gaps through international agreements (100 minutes)
- What Does It Take To Catch a Chinchilla? Verifying Rules on Large-Scale Neural Network Training via Compute Monitoring by Yonadav Shavit (2023)
- A Tour of Emerging Cryptographic Technologies by Ben Garfinkel (2021)
- Historical case studies of technology governance and international agreements by BlueDot Impact (2023)
- Overviews of Some Basic Models of Governments and International Cooperation by BlueDot Impact (2022)
Week 7: Additional proposals (60 minutes)
- OpenAI Charter by OpenAI (2018) and LP Announcement by OpenAI (2019)
- What AI companies can do today to help with the most important century by Holden Karnofsky (2023)
- 12 Tentative Ideas for US AI Policy by Luke Muehlhauser (2023)
- Let’s think about slowing down AI by Katja Grace (2022)
- AI Governance: A research agenda by Allan Dafoe (2018)
- [Optional] International Institutions for Advanced AI by Lewis Ho, Joslyn Barnhart, Robert Trager et al. (2023)
- [My Addition] Provably safe systems: the only path to controllable AGI by Max Tegmark and Steve Omohundro
Week 8: Career Advice (75 minutes)
- Career Resources on AI Strategy Research by BlueDot Impact (2022)
- Career Resources on U.S. AI Policy by BlueDot Impact (2022)
- China-related AI safety and governance paths by 80000 Hours (2022)
- Career Resources on European AI Policy by BlueDot Impact (2022)
- AI Governance Needs Technical Work by AI Safety Fundamentals (2022)
- Some talent needs in AI governance by Sam Clarke (2023)
Project Track - Field Building
Week 9: Defining field-building and the AI Governance field (45 minutes)
- How Do You Build a Field? Lessons from Public Health by Kevin O’Neil (2015)
- The longtermist AI governance landscape: a basic overview by Sam Clarke (2022)
- A summary of current work in AI governance by Anonymous (2023)
- The Strong Field Framework: A Guide and Toolkit for Funders and Nonprofits Committed to Large-Scale Impact by James Irvine Foundation (2009)
Week 10: Contextualising the field: past, present and future (60 minutes)
- The brief history of artificial intelligence: The world has changed fast – what might be next? by Max Roser (2022)
- Learning from History: Preventing AGI Existential Risks through Policy by Connor Tabarrok (2023)
- AI Policy Portal by United Nations Institute for Disarmament Research (2023)
- Measuring trends in Artificial Intelligence by Stanford University Human-Centered AI (2023)
- Explosive Growth from AI: A Review of the Arguments by Ege Erdil and Tamay Besiroglu (2023)
- 5 Neglected Areas in AI Risk