Oh boy, well in that spirit let me tell you about Parzen Windows! Now we all want to know where things are, and how much of things. We especially want to know how much of things are where things are! If we don't know the shape of something how do we know its density? There are many methods like binning or histograms that everyone knows, but let me tell you about Parzen windows.

We don't actually know how DeepMind programs arrive at their conclusions. Yes, we know the process of making them, and we generally know what they're good for, but we can't follow their thought processes, as it were. They've become too difficult to decipher. There's also a case that I cannot find right now where some scientists were experimenting with genetic learning algorithms on 10 10 pga arrays (I think that's what they're called? They decided to see if their algorithm could make a chip which would output a current when a certain tone was played near it.