Adam Gleave (CEO, FAR AI): “I visited LISA for several days and was impressed by the community they've built in a short time, with a large number of talented junior researchers. The UK AI safety community seems to be growing rapidly and I think LISA is helping catalyze this.”

Joseph Miller (Research Engineer, FAR AI)

“I've made several very useful connections with other researchers that have improved my work. I was able to ask questions directly to Arthur Conmy, who wrote a paper that I was studying in order to do follow up work. I also spoke with Hoagy Cunningham and was able to have very useful conversations about auto-encoders with him.”

Neel Nanda (Research Engineer, Google DeepMind): “Two of my MATS scholars (Arthur Conmy and Callum McDougall) used LISA as their office during MATS, and did fantastic work, their paper (Copy Suppression) is a solid and genuinely valuable contribution to the mechanistic interpretability literature. I'm glad LISA existed to host them!”

Lewis Hammond (Research Director, Cooperative AI Foundation): “Shared working and events spaces are an important tool for strengthening research communities, and AI safety is no exception. (Indeed, given that it is still a relatively young field, greater collaboration and cross-pollination of ideas is even more important.) Because of its relatively large population of AI safety researchers, strong universities (not only UCL/KCL/ICL, but also the nearby Oxford and Cambridge), and industry offices (Anthropic, OpenAI, and Google DeepMind), it is important that there is such a space in London, which ideally would be easily accessible regardless of one's affiliation. As far as I am aware, LISA is the only real candidate for this, and so every effort should be made to make it success.”