> system one: long-horizon planning, resource allocation, ml research, learning to forecast complex situations from few-shot examples

some epoch people are currently trying to prioritize / create as many benchmarks in these areas as possible, please feel free to throw ideas at us if you're time-constrained