[fxtentacle]

Well, the acoustic model is created using deep learning loss-minimization by gradient descent, which is what people usually call AI these days.

[postalrat]

Or ML.