Hi, I’M Michael Chen
Hello!
I work on AI policy at METR, an AI evaluations nonprofit. I help support foundation model developers with crafting AI safety frameworks, and I provide input to government bodies.
Before METR, I was a software engineer at Stripe, working on backend infrastructure and the occasional LLM application. I coauthored a paper on AI deception that was covered in The Guardian and MIT Technology Review.
Writings
RE-Bench: Evaluating frontier AI R&D capabilities of language model agents against human experts. METR. 2024.
Common Elements of Frontier AI Safety Policies. METR. 2024.
Response to U.S. AI Safety Institute Draft “Managing Misuse Risk for Dual-Use Foundation Models”. METR. 2024.
Response to NIST Draft “AI Risk Management Framework: Generative AI Profile”. METR. 2024.
The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning. International Conference on Machine Learning. Featured in Time. 2024.
AI deception: A survey of examples, risks, and potential solutions. Patterns. Featured in The Guardian and MIT Technology Review. 2024.
AI risks that could lead to catastrophe. Center for AI Safety. 2023.
AI alignment is not enough to make the future go well. Stanford Existential Risks Conference. 2023.
Inverse Design of Nanophotonic Structures Using a Hybrid Dimensionality Reduction Technique. Frontiers in Optics. 2020.