Human-like machine learning: limitations and suggestions

Re-assign community