Tunes in my phraseology

Mark Baily mbaily at iii.rmit.edu.au
Sun Sep 5 22:36:27 PDT 2004


Hello, I am writing to get an idea of what the other members think of these
ideas which area in some ways a repeat of ideas already discussed about tunes
but put into my phraseology. If someone has time to comment, please do so.

I think I am beginning to understand the idea more because I am beginning
to put it in terms of my own phraseology.

In my own terms it is a series of declarative statements of the form:
If "this" then "that"
If "this" then "that"
If "this" then "that" etc.
in each case with "this" being whatever the hell "this" might be and "that"
being whatever the hell "that" might be and with "this" and "that" having
no inherent meaning in themselves. This no inherent meaning allows one to
introduce fresh contexts in sittu to the model more reliably than before.

As distinct from saying
if "this is true" then "that is true" which would have been a different
statement entirely, because it has the automatic implication of the existence
of a corresponding if "this is false" statement.

Since a programmer knows that computers are not inherently intelligent or
"magic", when they analyse a domain, they are actually dynamically formulating
in their heads the problem in terms of these declarative statements beforehand
and as they go, saying "Ahh I see!" to themselves whenever they see that
declaratively, if "this", then "that" for a given "this" and "that" is the
solution to their problem. In current usage of computer systems, a so called
"good" or "hot" programmer will, for the purposes of quick reification,
keep in their heads their own private cache of such declarative knowledge
about the domain but also various intermediate layers of abstraction
(intermediate in terms of current HCI's including languages as well as hardware
knowledge etc.) which are not essential or relevant to the domain at hand, from
which they can quickly and easily reify an essentially static subset of the
model. This reification can take the form of coding in a (eg.) currently
existing computer language or even explaining it to someone else. However
the reflective knowledge contained in the original thought is lost on the
computer.

The purpose of Tunes would be to allow the original flexibility of the
model to be made available to the system, as well as to other people through
the system in a way that promotes freedom of use of the model in any desired
context, so that intuition can be augmented with computing power to allow
"experiments" in a much more dynamic and domain model oriented fashion. This
does not imply that a form of artificial intelligence would be being
introduced. Neither does it imply that models introduced from to the system
from the real world would be correct in any absolute sense, and some
perception introduced by one group/person might be considered incorrect,
irrelevant or even harmful by another. The purpose of Tunes is not to solve
these problems, except by allowing people to pick and choose at their own
leisure the abstraction level and aspects of the model they see/use, with
whatever precision they desire, where computer hardware (electronic devices)
are involved as an aid to the storage and processing of models.



More information about the TUNES mailing list