The current model has weaknesses. It could struggle with precisely simulating the physics of a posh scene, and could not recognize particular scenarios of trigger and result. For example, somebody could possibly have a bite out of a cookie, but afterward, the cookie may not Possess a bite mark.