Tuesday, November 30, 2010

Motivations

Though t-issues are important to the discussion, these are for the most part of little-t type. From time to time, there can be big 'T' issues brought in, such as this one or that one. It is of interest that both deal with things economic.

Are there big-T issues in engineering? Or, are they just easier to avoid?

Well, we can re-look at why truth engineering in this context. System thinking is a late attribute of enlightenment quest that we can trace over 100s of years. How all this came about can be avoided, skirting several big-T topics.

Yet, the basic theme is that systems relate to problem solving (oops minimization). Computational efforts then followed, and, for a long while, the computer was the human brain with some artifacts involved. Applied mathematics, and resolving systems of equations, was the focus of a lot of energy by very smart people.

We have to acknowledge their stamina, to boot. Some types of problems, that nowadays take the blink of an eye on a modern computational device, took years for people to work. One just has to marvel at the initiative, work ethic, and management skills in terms of tracking all of the details.

As our prowess increased, that which did the work became more artificially based. We're almost to the point that if the computer is not used in some professional situations things may be bordering on mal-practice.

Yet, just as we have underdetermination at the core of our knowledge, the real core of computing is undecidable. Wait! Yes. If we acknowledge that the basic circuitry is for the most part decidable, we can feel better. However, folks, progress has place real-time switching there that can be very problematic.

In fact, one worry ought to be that malfeasance-motivated mischief is not embedded. The US DOD takes this seriously.

But, we'll get back to that. Even if the switching was very stable, what we build upon the framework, layer by layer, introduces back the problematic issues real quick. Then, when we look at the growing presence (ubiquity), it can become very much troublesome.

Now, recognition of that is the motivation of the proposal that truth engineering is essential. Yes, it is necessary that computation be involved and that the majority of the work is being done by some artifact, but is that not the case and is not that expected to continue to grow?

Remarks:

01/22/2013 -- T-issues will migrate to issues of science and religion.

12/17/2010 -- An interesting topic may have presented itself for further study in this regard.

Modified: 01/22/2013




Sunday, November 7, 2010

Linear logic

Having taken a re-look at linear logic lately, and seeing the extensions as addressed in an ACM Communications' overview, I got the idea that this work could be used to illustrate some of the needed ideas, such as why things are hard.

From a computational sense, this logic has fitted well into the more problematic areas that we try to handle in the modern contexts, especially with growing use of computer systems. It does allow a better handling of proofs without falling into the traps of classical logic.

Yet, its entailment is not decidable. Somehow, we need to get that notion better understood. I've used the fact that we don't have 20-20 vision going further (farther, as in, forward movement from a position or state). Of course, we may not have 20-20 in hindsight, either.

Now, why is linear logic so important? It parallels the progress in advanced linear algebra that has been so important to numeric systems and visualization. That is, even non-linear problems are approximated using linear techniques.

And, we, essentially, have overlaid the world with this type of constraining view, despite that we know the limits of monotonicity and have determined the need for non-monotonicity in complex systems. That is, those with numeracy have superimposed upon us what is very well suspect.

That anyone would argue for some caution raises the ad hominem argument of innumeracy.

---

Now, having before raised the notion of getting technical, we'll do so (actually, we'll plumb to whatever depth is necessary -- the request of the reader? soar as far as we must, to boot). At first, we'll go top down, trying to establish a sound, coherent viewpoint (notice that this is juxtaposed to that approach which has been developed for the past 200 years and which has been deleterious -- disrespect for that which makes us human -- as well as beneficial - attempts at firming up the plumbing which can still leak - outside of our control).

Remarks:

04/03/2011 -- Need to look at some background.

11/21/2010 -- Three years ago, it was said: Computational foci raise miraculous need. Still applies.

11/16/2010 -- It is the fact that there are serious lapses in what we can and do know. Now, our maturity provides the basis for decisioning under the resulting uncertainty. That the computer has become a major player raises a whole bunch of issues. The main one is that the underlying framework is undecidable. Plenty have danced around this subject, but linear logic allows us to look at the problem more realistically.

11/10/2010 -- Jim M. could be a hero if he learned the lesson of undecidability.

Modified: 10/12/2011