Principia - Theoretical Model Organism for Alignment (Jin Hwa Lee)

​Principia is a new organization focused on deep learning theory and analytical training dynamics for AI safety. We attempt to build theory-grounded hypotheses that can be tested on AI system at scale. I will have a short pitch about Principia's research vision and we will have casual QnAs on how my experience at Pivotal helped me with the next steps in career to build a new safety organization.