Tuesday, May 17, 2016

Search, Find, Reject and Select

Searching for configurations
In my two series of articles Melencolia I and The Thinknet Project I've been trying to get a handle on the mechanics of intelligence and its relevance to evolution. One aspect of intelligence I feel fairly sure about is its general teleological structure which can be expressed by the rubric Search, find, reject and select. That is, it is a necessary (but not sufficient) condition of intelligence that it has general goals, sometimes vaguely expressed. This entails searching some kind of space of possibilities and testing and rejecting cases with the ultimate aim of selecting a solution to those goals. Clearly, this very general conception is short on detail, detail which no doubt in any working intelligence would be elaborated to the nth degree. But, I propose, the general idea of goal seeking is at the heart of all intelligence. The elaboration of intelligence is, in fact, self-explained by my general characterization of intelligence: Viz: Teleological targets can often be expressed quite simply e.g. We are looking for a way to fly from A to B. But to practically fly from A to B with efficiency requires a highly elaborated configuration of materials. Likewise, if our goal is to find a system that itself seeks and finds targets then a real target seeking system is itself likely to be highly complex. In short the complexity of intelligent systems is explained by the essential nature of intelligence itself -  as is so often the case in the subject of cognition we find self-affirming self-reference to be at the heart of it. Sometimes intelligence is characterised as  "complex adaptive systems" (cf John Holland). This characterization has some merit, but it also amounts to a form of goal seeking; in this case the goal of  adapting. 

The foregoing is an introductory preamble to some remarks made by Denis Alexander in his book "Creation or evolution? Do we have to choose", remarks I would like to showcase.  As I have worked through this book what piqued my interest most recently is Alexander's section on "natural selection" in chapter 4. Now, as I have said before unless the informational front-loading of the spongeam exists (which I actually doubt) then conventional evolution is not going to be the underlying mechanism of change in evolution: The evolutionary search space is simply too large to be explored by the parallel resources of non-quantum trial and error processes such as is envisaged in ordinary evolution. But having said that Alexander describes some other biological processes in his book which are known to exist and which work using the universal search, find, reject and select structure and these mechanisms are not controversial (I think). Alexander  refers to these biological versions of "natural selection" in  the following general terms:

....we see the same principle of abundance and selection operating time and again.....Jesus himself used the same idea in his famous parable of the sower who needs to scatter far more seed than ever will germinate and lead to a good crop (Matthew 13:13ff).

He then goes on to give some real biological examples (See pages 103 to 105):.

1. In the development of the brain neurons send out many "exploratory feelers" to other neurons and only the  fruitful connections are kept, the others die.

2. Particularly fascinating was Alexander's description that "During the development of B cells a specialised region of our genome undergoes intensive  random cutting  and rejoining of the pieces of DNA that encode different parts of the antibody protein. This results in the production of millions of different anti bodies, each one specific for a particular type of invader...."

3. Further, Alexander describes how when B cells replicate in response to an invader the antibody on the surface of the B cell changes during replication via a mutation mechanism. The B cells with the antibodies that best bind to the invader are kept and the rest are eliminated.

All very interesting, very interesting indeed. I might go as far as to call these quasi-intelligent processes; that is, they are a goal controlled searches - or what I refer to as "back-loading".  However, these processes use "real" materials as the search feelers rather then the tentative and readily expendable feelers of quantum mechanics. Of the latter kind of search feeler we have to turn to the exciton which transfers energy in photosynthesis. See here  for the following Wiki entry (My emphases) :

In 2007 a quantum model was proposed by Graham Fleming and his co-workers which includes the possibility that photosynthetic energy transfer might involve quantum oscillations, explaining its unusually high efficiency.[3]

According to Fleming[4] there is direct evidence that remarkably long-lived wavelike electronic quantum coherence plays an important part in energy transfer processes during photosynthesis, which can explain the extreme efficiency of the energy transfer because it enables the system to sample all the potential energy pathways, with low loss, and choose the most efficient one.

This approach has been further investigated by Gregory Scholes and his team at the University of Toronto, which in early 2010 published research results that indicate that some marine algae make use of quantum-coherent electronic energy transfer (EET) to enhance the efficiency of their energy harnessing.[5][6][7]

There we have it: Quantum searching. 

No comments: