kyjohnso

Epistemic Status: Exploratory/Very Early Days/More of a Hunch Preface(s) This is a cross post from Hylaean Sea — these ideas have been developing over the past two years and I'm posting here because I deeply value this community's feedback before I embark on the research fully in mid-late 2026. tl;dr:...

This is my submission for Blue Dot Impact's AI Alignment Fundamentals Course. In this post I explore computational policy, collaborative development of policy with AI, and how linking Language Models with World Models can lead to more precise training data, novel insights into the policy domain, and potentially new knowledge.