Tuesday, June 03, 2014

The Melencolia I Manifesto. Part 2

(See here for Part 1)
Human beings are complex adaptive systems and as such they have imaginations that can generate endless novelty to the extent that it matches the novelty they see around them.

(Note: Addendum below  added on 12/06/14)

In this post on Uncommon Descent IDist Granville Sewell links to one of his YouTube talks and says this about it:

I want to focus here on the second part, beginning at the 19:40 mark, which discusses the remarkable similarities between the evolution of life and the evolution of human technology……… Some people do not like the comparison, because (1) it may seem to bring God’s design down to the level of human design, and (2) they may say that Genesis 1 does paint a picture of creation by magic wand. With regard to (1), I would say that it does not bring God down to our level, because the things God has designed are so much more advanced than the things we design, but a designer must always get involved in the details of his design, no matter how intelligent he may be. And with regard to (2), although of course Genesis 1 is not an accurate scientific account of creation, even here we see a God who created one type of animal, “saw that it was good,” and proceeded to improve on it; that sounds a lot like the way we create things, though testing and improvements. And if all God had to do to create species was to wave a wand, why does the Bible say that on the seventh day, God “rested from all the work of creating that he had done”?

For me this development on UD is extremely interesting on at least two counts: Firstly, I also don’t believe in creation by magic wand. Secondly, thoughts like this are not entirely new to Uncommon Descent. In fact in March 2008 I broached a similar subject in one of UD’s discussion threads, a thread which can be seen here:
My comments to this thread and those who responded to me (Paul Giem, Gpuccio and Jerry) can be downloaded from here (See the “file” drop down). My first comment wasn’t understood (probably because I hadn’t made myself clear enough) and this was compounded by a tendency for UD’s ID interlocutors to assume that any challenges to their position comes from straight evolutionists, “materialists”, “naturalists” or atheists; in fact YEC Paul Giem seemed to think I was arguing from an atheist position – he refers to “your naturalistic friends”.  However, I must add that on the whole I have found the UD people, whether Old Earth or Young Earth, to be nice intelligent people who aren't in the “heretic burning” mould; it is probably that which sets them apart from the stricter fundies whose personalities favour stick rather than carrot. However, UD fans are going to listen to Sewell as if he is saying something completely new; after all, he’s one of their gurus and, moreover, they are likely to have a "not invented here" attitude.

In May 2009 I made a shorter but similar comment on this UD post:
This comment I reproduce below:

Timothy V Reeves 05/16/2009 9:23 am
I enjoyed the song but “One piece at a time” may not be so far from the mark.
The wheel, the wheeled chassis, the leaf spring, glass, various electromagnetic devices, pistoned heat engines, bulbs, not to mention agriculture, cities and writing which set the social base for industrialization etc, were all invented/discovered without the car being conceived or envisaged as a goal. This example sets the precedent for a form of reducible complexity at least to an extent which allows limited human intelligence to make advances with a piece meal divide and conquer strategy, and achieve results beyond that available to a single act of inventive foresight and goal formulation.
The artifacts generated by human culture must form islands of innovation in “configuration space” sufficiently close together to unable limited human intelligence and prescience (let’s represent that by “i”) to jump the gaps between these islands of functionality.
Now I would not be so brazen as to suggest that “i” could be reduced to zero and hey presto you have mindless, goalless evolution (I realize there are lots of robust challenges to that, thanks to you excellent folk), but the human technological model does indicate that limited foresight and goal perception can generate functionality beyond itself if some measure of reducible complexity holds in “technological morphospace”. If this were not so then human technological progress, with its ability to create unforeseen, unimaginable artifacts well beyond single quantum flashes of inventive genius will come to a standstill, limited by its ability to see ahead and formulate goals. Such is the Creator’s grace bestowed upon finitely endowed humanity.


So what are my conclusions here?  Whenever I have looked at the North American ID community’s concept of intelligence it has always come over as a “black box” notion of intelligence – that is, as an inscrutable ancillary agent of causation that stands a dichotomy apart from so-called “natural processes”,  an agent invoked when we can’t think how “law and disorder” processes have generated a pattern. At least Sewell is, as far as I can tell, starting to look into that black box; at last someone on UD is hinting that intelligence is a process and in fact a process that looks a little like evolution itself! It is this immanent aspect of intelligence that is the subject of my current researches. Below I list the salient areas of this research. The following is my broad brush manifesto: of this research.

1. Intelligence is not a highly located phenomenon and least of all is it something precisely focused in a small black box. Rather, intelligence is a distributed phenomenon that cannot be identified in low level parts but only at the high system level. The low level parts are, however, essential component contributors by way of their mutual configuration. Therefore intelligence is both immanent as well as eminent to any local sub-system.

2. Intelligent problem solving activity has a common structure that constitutes a very general heuristic of intelligent action; namely, search, find, reject and select.

3. In a search, find, reject and select computation there may be a separation between explicit outputs and implicit computations between outputs. It is this difference between hidden background processing and ostensive foreground results that can convey the wrong impression that intelligence isn’t a process and that it somehow arrives at results by magic wand.

4. Searches can be enhanced with mathematical constraints in order to improve speed of result.

5. Searches can be enhanced with expanding parallelism to improve speed of result.

6. Points 3 to 5 have isomorphisms with quantum physics leading me to speculate that the universe has a declarative computational structure involving constraints guiding a Search and Find process.

7. Randomness is not an event but a class of pattern that has certain mathematical properties in terms of its statistical form and its intractability to algorithmic generation. These kinds of patterns have an important status in the search for life; for if life’s configurations can only be expressed in compressed form as random patterns then it follows that the ordinary parallel processing of standard evolution will not find life in realistic times.

8. In our universe with its declarative computational structure - that is, it is a highly constrained search - there is no such thing as “random and unguided” processes; the guiding hand of constraint is ever present and operative.

9. My overall working assumption is the strong anthropic principle;  that is, the cosmos is contrived to find life.

The above manifesto may well be a flight of the imagination; in fact it probably is, but it's my best shot (hence comic sans). But one thing is clear: The following statement by PZ Myers is typical of a failure to appreciate the miracle of evolution: 

Darwin’s great insight was that you don’t need an overseer guiding evolution — that local responses to the environment will produce efficient responses that will yield a pattern of descent and diversity and complexity. To replace “intent was unnecessary” with “God provided intent” does deep violence to the whole theory, and completely misses the point.*

Any process that has generated life as quickly as has the cosmic process is certainly “guided” in the sense that it must be highly constrained and/or employs the methods of expanding parallelism. We might speculate that there must be some “meta-explanation” for this state of affairs, but such an “explanation” could only employ yet more mathematics of “law and disorder” and in effect we would simply be embarking on a “Turtles all the way down” regress; or as it is expressed in its usual modern form “multiverses all the way up”! Science may be less about "explaining stuff" than it is understanding and describing the status quo.

* See Myer's post "How to argue for evolution" dated 20 May

Relevant link:

Addendum 12/06/2014

In a blog post entitled Multi-component, schmulti-component and dated 10 June I was interested to note that PZ Myers has latched on to this parallel processing idea:

Evolution isn’t sequential. It’s massively parallel. Massively. Humans have about 20,000 genes, and all of them are evolving at once, with trial runs in about 7 billion individuals.

To me, that still classifies as ordinary sequential processing, as it simply involves adding more sequential processors. Unless somehow physical constraints sufficiently narrow down the search space a priori, up front (which is possible I suppose) a "small" parallel processor is not going to find those viable self-perpetuating, self maintaining organic survival solutions. Unless these highly constricting physical constraints exist, adding processors whose count is measured in only a few tens of digits (that's what I mean by "small") is utterly inadequate. After all, the search space has dimensions measured in millions if not billions of digits. PZ Myers intuitions about the adequacy of a parallel processor are far wide of mark. 

I said "expanding parallelism," a method that creates far more trials; if somehow quantum expanding parallelism can be exploited we're there. But either way Myers has an a-priori brute-fact issue; either hugely constricting constraints and/or huge numbers of trials facilitated by some kind of expanding parallelism. I'm putting my money on the latter, although the former is a possibility.

Consider also this: 

Evolution is not teleological. An organ like the eye is not being assembled to a set of specific, detailed instructions — it just has to work, or the organism is at a disadvantage to other organisms with better eyes. So a hodge-podge of solutions is accumulated, and the end result has all kinds of complexity. But you don’t get to argue after the fact that the details imply some specificity of purpose.
For example, here’s a number: 343767. It’s kind of big, you might be tempted to argue that it’s a fancier or more complex number than, say, 300000 (you’d be wrong), or you might want to argue for the significance of individual digits, or find a pattern in it. Humans tend to do that. But the reality is that I just went to a random number service and asked for a 6 digit number. Similarly, eyes wandered through a random space constrained by functional requirements and ended up at a somewhat arbitrarily complex configuration — and different lineages followed different paths.

Fair enough; the only requirement for an organism is that it has self-perpetuating/self maintaining stability given its particular environment; that's likely to lead to "hodge-podge solutions". But a cosmic generating system that is not prepared a-priori via expanding parallelism and/or up front constraints simply isn't going to find even hodge-podge solutions! Those stable organic "potential wells" in which the seeking fingers of quantum signals eventually accumulate are lost in a space unimaginably large - by comparison the visible universe is absolutely tiny!

Tough luck PZ; the cosmic "search algorithm/hardware" looks highly a-priori and anthropic to me!

Relevant Link:

No comments: