I love this idea of transparent assessment. I wonder where the boundary would be for too much artificial intelligence in a submission?