Hi, I'm Garrett.

garrett_xu [at] gmail [dot] com

I'm currently at Brown, studying applied math. Somewhere along the way, I started spending most of my time thinking about AI safety.

What are my actual values? Freedom — to fall in love, raise a family, grow old in a world that still has room for that, post-AGI. So I'm trying to develop a theory of victory, and build my work around whatever will matter most.

Garrett Xu leaning around a stone pillar in a courtyard
Here's a photo of me, so you know you've got the right Garrett Xu.

Right now, I'm an external researcher with Cal OES, helping draft internal deployment risk reporting guidelines for frontier AI developers. I also lead the policy team at Brown AI Safety Team (BAIST), contribute to METR's frontier AI regulations reference, and pick up whatever else needs doing — most recently canvassing for Alex Bores and running logistics for the AI protest that marched from OpenAI to Anthropic to Google DeepMind's SF offices.

Previously, I spent four years in the California legislature helping pass bills on AI transparency, social media safety, and education policy.

I'm also interested in dermatology research, hospice volunteering, music (Daniel Caesar, John Mayer, Bob Dylan), and pickup basketball.

I want to watch something boring on a Tuesday night and not have it mean anything. I want to be surprised by a good meal. I want Sundays. I haven't always loved the world. But I do now, and I want to stay in it.

I'm always happy to talk — feel free to reach out!