Would it be possible to use machine learning to do this job better, meaning, could the machines look at what other machines are doing, and better translate to us whats going on?
That's already known: there are inputs to the network that do not make sense and yet will trigger strong responses. Think of them as inputs that have the same effect on NNs that optical illusions have on the human brain. We infer something that isn't there.
I suspect that as network architectures get better and parameter counts drop these will get harder and harder to construct.
You could use a neural network to optimize the hyperparameters of another neural network, but afaik it's not a really efficient way, your best bet would be to use bayesian hyperparameter optimization.
