The current model has weaknesses. It could struggle with correctly simulating the physics of a complex scene, and should not understand precise situations of bring about and effect. For example, somebody might have a Chunk from a cookie, but afterward, the cookie may well not have a Chunk mark