Assessing the thickness of melanoma is difficult, whether done by an experienced dermatologist or a well-trained machine-learning algorithm. A study shows that the algorithm and the dermatologists had an equal success rate in interpreting dermoscopic images.