Mastering Deep Learning: A Complete Introduction for Beginners and Newbies by James Gabriel
My rating: 0 of 5 stars
Deep Learning seeks to mimic how humans learn (i.e., the brain processes of how humans learn in their cerebral cortex) and apply this mimicry to how computer programs are written. Thus, we have terms like a “neural network” which does not refer to a brain (made up of neurons) but to a web of computer cells which “learn” how to produce certain output from input data.
One interesting application of such is detailed in the book. A Generator attempts to produce a fake image, and a Discriminator attempts to figure out which items are fakes. When paired together with machine learning, they can produce a fairly good fake image.
What’s somewhat scary about this technology is seen in the newspapers and social networks of today. “Fake news” is not merely news which is contrary to a certain viewpoint; fake news can be a video clip of some famous person saying some phrase that she never uttered! If that doesn’t sound like 1984 (or 2016?), I don’t know what does.