Comment by VHRanger

Neural networks put a huge damper on themselves by being so compute hungry.

They've always been there, it was just an inefficient way to do machine learning before around 2011 when GPUs started being good enough