• Subject: Re: Design shift of view
  • From: "David Morris" <dmorris@xxxxxxxxxxxxx>
  • Date: Mon, 27 Jul 1998 17:43:34 -0600

Rob,

Can you explain the difference between a Neural database and Neural network?  
Also how does connectionism fit in?  I have played with genetic algorithms and 
programs.  I have worked on many applications where genetic algorithms provide 
much benefit.  Most of our problems seem to work with discrete sets of data.  
What sorts of problems do a Neural database and connectionism lend themselves 
to?  What about the application development process itself?

In my experience the biggest gains in application development come when a 
common solution can be applied to many problems.  The challenging part is 
recognizing a recurring pattern and building a common solution.  This has been 
an iterative process for me with an occasional gain in productivity.  I would 
like to see how connectionism could be used in our environment.

Thanks for your input,

David Morris

>>> "Rob Dixon" <rob.dixon@erros.co.uk> 07/27 2:54 PM >>>
I have been off-line for several days, driving through France, and am now
slowly catching up with my e-mail.  It is interesting how many posts the
"Design shift of view" thread, started by James Kilgore, has acquired in
the meantime.  

There are many good points that I could pick up on, but if I am only to
choose one, it would be that from John Hall dated 23 July.  I quote -

> To my view all you really need is one "data type."  This data type would
> then describe all other data types.  All rules essentially become data. 

This certainly ties in with my experience and implementation of the Neural
Database.  More importantly, it also, in my view, ties in with the human
brain.  All of us, particularly in the IT field, have some ability to
understand information of a type not understood by our parents, so I assume
that we did not inherit genetically this specific ability.  Instead I
assume that we inherited a more general ability which allows our brains to
cope with a great variety of "data types" by treating all data types as one
and the same - i.e. a "universal data type".  The tools that we have used
historically in the computer industry, have perhaps, because we do not
understand the mechanisms of the brain, made us differentiate between
"things" that are really the same.  Unless you believe that someone is
making separate decisions about every human brain for each new type of
information and deciding how and where in our brains each new type is to be
handled (i.e. reprogramming each individual brain separately), our brains
must treat all as the same.

I quote a very recent article in the British "Sunday Telegraph" by Stephen
Hawking - in the UK at least, the best known theoretical physicist, even
though most people who have heard of him, including me, are not sure what
theoretical physics is -

"Humanity is making great advances in science and technology, and the pace
of change is quickening.  I recently gave a lecture at the White House to
an audience that included the Clintons.  I argued that ..... progess will
continue at an ever-increasing rate.  In this situation, it is vital that
we all take part in the debate about where we are going.  We don't want the
knowledge and the decisions to be left to a few experts."

In our own small real-world ( I of course think it is very large and
important!), we have started our own, vital, debate.  When we have all
learnt all that we can, at this stage, learn, we will be as well equipped
as we ever will be to make decisions about the way our industry is to move
forward and so how we can best serve our users who are not necessarily
equipped to make such technical decisions.  If we are not to hold back
society (word-wide, not just our own local versions), we must make those
decisions fairly soon.  Making decisions means making judgements and taking
risk - putting our necks on the line.  But the worst decision of all, and
the most risky, is not to make one at all.   I do not believe that we can
stay as we are.


Rob Dixon
----------


+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---


This thread ...


Follow On AppleNews
Return to Archive home page | Return to MIDRANGE.COM home page

This mailing list archive is Copyright 1997-2019 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available here. If you have questions about this, please contact [javascript protected email address].