Natural Order and Science

[I_am_not_mafia]

But more seriously, it is easy to imagine (and to construct!) systems which self-organize in a complex fashion that after a few iterations cannot be "understood" by human intellect any more than the entirety of how dna organizes embryological development or how the brain creates consciousness. Take a blank deep neural network and let it learn something. It will become a highly complex thing which accomplishes a task by seemingly magical and to our minds hopelessly intractable steps. My point is that our level of understanding of how a complex self-organizing system accomplishes things is by no means a measure of how likely it could have arisen on its own dynamically (and I am talking not about the construction of the blank network which is obviously designed by humans in my example, but of the learning process which creates the actual complexity which is entirely driven by experiencing the environment.

Without having proper code of conduct for stabilizing and disciplining the action, ever-changing circumstances would transform into chaos in any structure of events in the universe. Not a single event in the universe is the conduct which is free of some coding system which specify explicit rules of action and specify the expected behaviour in accordance with those determinant set of rules. That is another story whether we have an understanding of that coding system or not but there is always a connection between action and certain laws which guide those actions and we cannot simply say that whatever can cause a representational state is represented by that state.

Unfortunately, lack of code of conduct obscures the clarity of Natural Selection’s distinction which by no means can disentangle itself from the ambiguity and able to fit into the world of terminology of the modern scientific traditions. Natural Selection is too weak to support a comprehensive mathematics and on other side system of virtual analogy are not exchangeable: the order of results makes a difference. That regrettably means that wheels of fortune do not constitute a technological (let alone a biological) kind for which you can use information about previously investigated instances. But if you had knowledge of a random sample of all existing wheels of fortune, then you would work on the average for a new, randomly drawn, one.

See, Alex said something very reasonable and factually true. I and other scientists who have worked with artificial evolution and self organisation can attest to the fact that it is relatively easy to evolve something using very simple components that can take many months to figure out how it works, if at all. Whether it's a circuit design using FPGAs, neural networks or whatever. I myself spent three months trying to figure out how my neural networks actually functioned during my PhD. In fact I actually spent two weeks trying to stop them working by removing components that I assumed were required only to find that they kept on working albeit at a lesser performance. I envisage spending just as long if not longer with my artificially evolved dynamical systems. All I know is, they work.

You on the other hand Harris, have responded with word salad that is not relevant to anything that Alex has said. It is also factually incorrect and relies on equivocation. You still haven't explained what a proper code of conduct actually is. Or for that matter a coding system. I can say though that "specify the expected behaviour in accordance with those determinant set of rules" is factually wrong because it ignores the concept of emergent phenomena which has been studied in-depth in the scientific literature (e.g flocking behaviour, ant colonies, bee hives etc). You can for example emulate flocking behaviour with just three rules without specifying an expected behaviour.

You Sir, are talking shite.