[D] Documenting the Weaknesses of Deep Learning (or are there any?)