People
The team running Mercial. We are looking for more people to join us! If you are interested, fill in the expression of interest form, and we will send you an email. We don’t have a high bar - a couple of hours a week can be really helpful.
Leadership
Lachlan Ewart
Director
Hi! I’m a third-year maths student. I am currently a SPAR research fellow, and I am interested in monitor robustness. Outside of AI safety, I sometimes write about philosophy on my Substack.
Leads
Rohan Neelala
Research Fellowship Lead
Hello! I'm a first year MPhil philosophy student. I've spent most of my time studying metaethics and mind and action. I'm always looking to explore the application of these topics to the field of AI safety, particularly relating to alignment and digital minds. Please reach out for a chat if you're interested!
Gideon Chang
Events Lead
Hi everyone! I am a third-year maths and statistics student. I am interested in capabilities evals, mechanistic interpretability and AI news in general. Always down for a chat, hit me up!
John Middleton
Blog Lead
Hey! I'm a third-year software engineering degree apprentice, working at Jaguar Land Rover. I'm interested in the social side of AI Safety - concepts such as alignment to whom. I love reading, writing, and being able to prove things :)
Erin Gruenberg
Operations & Outreach Lead
I’m a second-year philosophy student finding out how to make a positive impact with my career. I’ve written a little bit about AI safety and am exploring what else I can do to help within the field. In my free time, I love solo-travelling and learning photography.
Advisors
Matthias Endres
Advisor
I work on Behavioural Economics and Large Language Model evaluations. In my free time I think about gradual disempowerment risks and support students to find impactful careers.
James Sykes
Advisor
I'm a statistics PhD student working on cooperative AI. I'm also on the Recursive Self-Improvement Initiative team at the Institute for Security and Technology, where I help policymakers and frontier AI labs design risk thresholds.