Using a machine-learning system known as a deep neural network, researchers have created the first model that can replicate human performance on auditory tasks such as identifying a musical genre. This type of model can shed light on how the human brain may be performing the same tasks.