[adamboulanger]

The real travesty of the blurred lines case is that data science was not used, from what I gather, at any point. At a minimum, the two tracks could have been compared as to quantifiable similarity using any number of methods. EchoNest, and other recommendation approaches, convolutional neural nets, etc.

The lack of computational methods to abstract and describe the material on either side is particularly bizarre considering the judge's decision to preclude original song material from being played by the Gaye family. I'm convinced that science, rather than expert insider testimony that culminated in a courtroom live-performance jam, could have prevailed.