[FRIAM] Optimizing for maximal serendipity or how Alan Turing misdirected ALife

[Steve Smith]

Jon/Glen (et alia) -
> I very much agree that questions of a /formality-informality/ spectrum
> will weave
> itself throughout /the work/. It seems to me that the informality
> ought to provide a
> /place/ for birds to make a nest, a /bellybutton/ for lint to collect,
> and a /place/ for
> rust to never sleep. To my mind, it is not necessarily the formality
> that chokes
> development. Rather, I think of formality purely as description and
> one among
> many valid and possibly incongruous descriptions.

I am brought to think of mechanical tolerancing in juxtaposition to
precise machining.   The latter yields better performance (e.g. in a
high performance internal combustion engine) initially but the  former
provides for better interchangeability of parts, especially for example,
replacing worn parts along the way, and as I understand the principle,
better performance (including reduced further wear) than putting in a
"perfect as when new" part which would at least *accelerate* wear and
*reduce* performance  (think of replacing one ring or one main bearing
in an engine after the rest had "worn in significantly")

As I understand it, traditional gunsmiths and early ICE mechanics had to
*add* wear to a replacement part to match (not always in an obvious way)
the wear amongst the existing parts.

In this *analogy*, a formal description of a system is like the
precision machined parts which all function together well on assembly
and all wear in together in-balance;  a *less* formal description might
be like a machined part with deliberate tolerance built in which does
not require as much "wear in" to reach a dynamic balance and when a
replacement (unworn) part is introduced the whole system is more able to
accept it in with it's pre-created "slop".     I am suggesting that a
more informal (is formal->informal a spectrum or a step-function?)
description of a system leaves room for the "moving parts" to operate
even if they do so sloppily where if you "tightened up" those elements
with (overly in this context) formalisms they would bind against one
another?   Can an informal system be a precursor to a formal system?  
Or must a formal system be constructed by composing *smaller* formal
components?    Some of the early attempts at powered, manned flight
seemed to fail because of their lack of precision while the Wright Bros
succeeded *because* they composed a series of more precise/formal elements?

> Here is a place that I would
> again emphasize Rota's take on eidetic variation. For Rota, the eidetic
> variation includes all of the counterfactuals, contradictions, and
> messiness
> that we develop/uncover as we vary in our minds an object of interest.
> It is
> not necessary that we cut away/babies from bath waters/, but rather
> recognize
> that the concepts are complex. I believe that the development of a
> concept can
> especially choke when we fail to recognize that a concepts formal
> description
> has a /combinatorial explosion/.
Following my line of thinking above, the contemporary penchant for
developing software/systems as "soft assemblages" of existing allows for
what seems like this "combinatorial explosion" what with "soft
assemblage" leaving the "slop" to allow the components to work together
in spite of not being designed to (humans in the loop often make up for
the slop?).    It also limits the combinatorics somewhat by
allowing/selecting for larger components.   Bolting a grappling hook
onto an offroad vehicle with acrylic window-coverings and sheet
steel-armor on the sides to make a storm-chaser vehicle (with only
hundreds of variants) rather than building one of 10,000 variations from
wheels, engines, nuts, bolts, etc.   Spot welders and epoxy glue and
self-tapping screws and duct tape and bailing wire and bungee cords make
for "soft assemblage" .
> A good example is a way the concept of random
> number can be used, ironically enough, /informally/ to mean a number I
> can name.
>
> When in a conversation the concept of the random number is invoked, it
> evokes for
> me a complex. I can sense within a /single complex: /frequentist
> randomness and Chaitin
> randomness and even an ephemeral feeling/non-symbolic
> experience. Comparison of
> these complexes with others provides the opportunity for new
> pivots and jumping-off
> points, for the serendipity of missed connections and false
> juxtapositions. There was
> something of this in my experience listening to the podcasters. At
> times I thought
> that one had completely missed the other's /point/, but really /I had
> missed the point/,
> namely that the discussion was /not about a point/. They were in play,
> constructing
> common complexes and variations which they could share.
I like this apprehension, which also evokes in me a sense of abstract
classes in OO...  the creation and manipulation of a meta-thing that
captures the essence(s) of the myriad things that can be made more
concrete and/or instantiated from the abstraction.
> When I compare or attempt to describe my sense of this in terms of
> varieties
> and free module constructions, I am not saying that concepts /are/
> these things.
> I am appealing to varieties (say) in terms of its /conceptual/
> content. If we found
> that the language was flexible enough to do calculations, well that
> would be a
> pleasant though unintentional corollary.
Interesting.  I'll have to listen more (and more carefully) as this unfolds.


- Steve

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20200528/3c38e12e/attachment.html>