About Me

Hi, I’m Joe! I recently left my job at Anthropic and will soon be joining METR, an independent organization that evaluates how safe AI companies’ systems are. I’m extremely excited to increase transparency around the risks from AI systems. I believe that on the current path AI development could be catastrophic for humanity, and informing the public about the state of these risks is essential to help the world navigate this transition responsibly.

Previously, at Anthropic I managed the Scalable Oversight team at Anthropic and was the research lead for the Anthropic Fellows Program. For more about my research, see my research page. Before that, I was a PhD student in the Department of Statistics at the University of Oxford, where I was supervised by Arnaud Doucet and George Deligiannidis and worked on the theory of diffusion models. I also spent time at the UK Frontier AI Taskforce (now the UK AI Security Institute), helping to set it up in its early days.

Outside my research, I enjoy running through the hills in Berkeley, learning new things (currently, I’m trying to understand quantum field theory and improve my skateboarding), and reading.

I get a large number of emails from people who are interested in working with me. I unfortunately don’t have time to reply to all of them, but my best advice if you’re interested in my research or getting into AI alignment more broadly is to read Ethan Perez’s advice here and Neel Nanda’s advice on his blog for how to be a strong researcher.