[sbierwagen]

The AlphaZero algorithm swapped between board games pretty easily. OpenAI could also have been gesturing at this when they named the GPT paper "Language Models are Few-Shot Learners".