Tuesday, June 12, 2018

The Thinknet Project Part 4

The forth part of my Thinknet Project can be downloaded from here. Below I reproduce section 11, a section which is about Intelligent Design a subject which is very relevant to this blog as I have posted so much on this contentious question.

 The other parts to this series can be picked up here:



11.  A note on Intelligent Design, specified complexity and information creation.

As we saw in the last section Thinknet is way of seeking and selecting the improbable, given certain input specifications, specifications realized as consciously stimulated patterns.  As with an internet search a Thinknet search has the potential for returning rare cases and this equates to improbable cases; that is, cases of high information

The de facto Intelligent Design community often talk about specified complexity and the impossibility of “natural processes” (sic) creating information, thereby implying that such are only available to an intelligent process. It is not always clear just what the de facto IDists mean by an intelligent process and by specified complexity. Also, it is not clear why “natural processes” (which for the Christian are processes created, sustained and managed by a “supernatural” God, so they are hardly “natural”) can’t create information; after all human brains presumably classify as “natural processes” and yet they seem to be able to create information. 

As we have seen in my Melancholia I project so-called “natural processes” can create information, especially if they have an exponentially expanding parallelism. The reason why macroscopic “natural processes” appear not to be capable of creating information is because:

a) They don’t often have this expanding parallelism and therefore generate information only slowly with the logarithm of time.
b) They don’t often have a teleological selection structure which clears away generated non-targeted outcomes. Therefore “natural processes” appear not leave the high information targets conspicuously selected.

Thinknet on the other hand, has both of these features. Viz:

a) An exponentially expanding parallelism is required to search the complex network of associations.
b) Thinknet has a built-in teleology which leads to the clearing away of outcomes which do not meet the target criteria.

If Thinknet is an indication of the fundamentals of human cognition then it follows that the human mind is a natural process which conspicuously creates information and targets it. This is not say that what we classify as non-intelligent processes don’t create information; as I have said above, they do, but not conspicuously because in many non-intelligent processes information is only generated in slow logarithmic time and also without selection. The latter in particular may explain why atheist world views tend to have a preference for information symmetry; the many worlds interpretation of quantum mechanics does entail exponentially expanding parallelism but it lacks asymmetrical selection of information, a trait which smacks of a teleologically interested intelligent process; the latter would of course be unacceptable to many atheists.

The Thinknet simulation does, however, throw some light on de-facto ID’s so called “specified complexity”.  In my simple Thinknet simulations two stimulated input patterns A and B are used to specify a sought for outcome in a similar way to an internet search. As we have seen, symbolically this can be expressed as:

[AB] => C

Input patterns such as A and B have zero information as they are from the outset known objects; that is, they have no Shannon “surprisal” value expressed by an improbability. But from the outset C is an unknown and in fact may be a member of very small class of objects which fulfill the conditional specifications A & B. Thus, C may have a high improbability implying that it is a high information object. The computational complexity of the outcome C is implicit in the symbols “[ ] => ”. These symbols represent the search needed to arrive at C. Since a Thinknet network presents a search problem whose computational complexity is an exponential function of the network penetration depth, then the operation symbolized by [ ]=> may embody a high computational complexity. If this high computational complexity is actually the case then we can say that C has a high specified complexity with respect to the specifications A & B.

No comments: