AI competition without catastrophe

I’m Sophie, a researcher at MATS studying US-China tech competition. Before that, I researched AI-enabled biosecurity risks (bioweapon.ai) and open-weight model risk (uncensorable.ai). I work on AI safety because I believe decisions made about AI this decade will shape everything downstream – will we cure cancer, or will someone build a bioweapon? What’s going to happen to the economy? Will powerful AI destabilize nuclear deterrence? In this program, I plan to write about the conditions under which the US and China could meaningfully cooperate on making AI development and diffusion safe.

Aspiring Irish castle resident

My favorite hobby is collecting strangers’ life stories. A couple of years ago, a friend and I did street interviews across the UK and ended up chatting with the daughter of a prison escapee – still the interview I think about most. Otherwise, I can be found rereading Little Women (five times and counting), journaling, and sustaining myself on tea and acai bowls. My bucket list ranges from “see Halley’s comet” to “eradicate a disease” to “live in an Irish castle.”