Tuesday, December 25, 2007

Tranche or not

Earlier posts considered whether we are putting too much trust in mathematics and the related computation (issues of quasi-empiricism) and in processes based upon those two. There were allusions to problems that come out of abstraction as it pulls one away from substance toward what are 'bubbles' (a bane of modern existence). That is, having more substance can be thought of as equivalent to being more 'real' (opening up a door here).

One problem to understand and to control is application of the computational to objects-in-themselves (those who want to knock me as Kantian may consider their understanding of tribalism on intellect). On the computer, it is easy to reboot and to start over. Simulated errors and faults (on the computer) have no consequences in the world.

Yet, we cannot do similarly, that is reverse time or consequences, in the real world except for limited types of things.

Simulation has brought about troublesome thinking especially in so far as success there makes it appear to be 'okay' that one takes simulated data and acts as if it is real. It used to be that a prime counsel, in defining and using numeric processes related to complicated natural situations, was to not confuse the observed (measured in the world) data with something that was generated by a model.

Of course, where one can occasionally calibrate some model against the world, such intra-computer work is of use (operationally sound).

In the sense of a new product, especially one like a plane, the vehicles that will fly and otherwise support testing would have more substance than sub-assemblies or those artifacts with which abstractionists like to be bothered, even though the sub-assemblies in themselves can be tested or that the artifacts are amenable to analysis. And, in the case of a simulator, flight data would support calibration, so to speak.

Looking at the 'tranche' logic of finance makes one wonder if project management (in so far as it partitions and assigns and then has the worries of earned value) can have a similar set of lurking failings. Ah, the tranche, a mechanism to extend ratings above what is actually supported by the underlying instruments.

There seems to be some analog between the tranche mis-thinking and planning versus status measuring that we'll explore further. It's easy to poke fun at the money crowd as there really is no measurable basis that has been agreed upon and that has lasted. Yet, that 'froth' and 'bubbles' come about indicates that there is a problem.

Project management is a little better if is deals with a product that substantially deals with matter (such as a plane). But, as we all know, computer system project management is still more an art than a science despite the best efforts of DOD (US) and Software Engineering Institute, though this work may help prototype the types of control systems that will be needed.


08/01/2013 -- Ben cannot unwind or taper downhe has too many Doves. We'll have to get back to the king thing (yes, the divine rights of the CEO, new royalty, in other words) and dampening of these types by a new outlook (Magna-Carta'√≠sh).

05/25/2011 -- Stinking realities, such as, lemons problem, dark pools, ... Oh, so much to look at! Does the idiocy have any limit?

04/03/2011 -- Need to look at some background. Too, tranche and trash.

01/01/2011 -- We have four last posts of December under our belt. Also, we lost our chance at setting things right.

11/02/2010 -- Two years later, the message is the same, except some changes have occurred. Of real note is that the jobless rate is high; out-housing really set up for that. Also, we need to re-look at that learned from the 'vons' guys, Ludwig and Friedrich. See Near Zero.

07/23/2009 -- After the bust and the rebound, toxic assets are still a problem due to tranche realities.

06/07/2009 -- Say what?

02/18/2009 -- We can look at why securities become toxic, almost by necessity.

01/27/2009 -- Lessons to be learned (as opposed to learnt), including, by necessity, Ponzi.

11/12/2008 -- Well, things fell apart fairly quickly, starting in September of 2008. By N0vember, there was general spooking. Starting in September, movements toward nationalization sped so fast that it was easy to forget that a Republican administration was still in the White House. Talk about rewarding hubris and moral hazardness!!!!

Modified: 08/01/2013

Saturday, December 15, 2007

IQ and PIQ

Flynn's work on IQ shows that the IQ level (as determined by tests) increases by generation thereby requiring that the average value for the tests (100) needs to be re-calibrated. Other work shows that there are definite ways to look at how intelligence may be rated.

For now, let's just consider two that were considered in the reviewed book. The main thrust has been to measure ability to handle abstractions and their myriad layers as if that is how one gets to real knowledge (theory of everything - which, of course, pertains to nothing). A more operational focus would look at functionality (oh, by the way, which is more important for a plane - looks or being able to perform).

It is functionality that gets the short view of late. Why? The whole scheme of things seems to be going toward abstractions (covered in early posts here and in Truth Engineering) where those who excel on this level in certain types of instruments gather whole bunches of moolah (in many cases, with explicit take-aways (of bread, essentially) from those who are actually performing in ways these abstractionists can only dream about - many of whom have servants [essentially] who keep them afloat - often these types do not even know that their backs are unclothed [much more of a problem than the naked emperor] or that they are being carried about).

But, let's get back to IQ and the accumulation of this in both the functional and ethereal (of course, management and that dismal of sciences, economics) domains over time in an OEM of complicated things like a new airplane, let's call it PIQ. Well, PIQ would show increases within a program; so we see variants that exploit new knowledge plus earnings from improved processes. This is all well and good.

But, too, we see that within some type of operational methods (even including types of farming out) involved with continuing improvements within a technology line (of which we see the culmination that has been extremely successful come about - this is about 10 or so years ago), the progress makes things better and better. The PIQ increases; even generational dynamics would cause an upward movement; things work as expected, for the most part.

Now, and how is this not seen (except for the fact that those who really know seem to have lost out to the dreamers who deal with abstrations -why? - well, it's partly thanks to mathematics and the computer), when new technology comes in across the board with across-the-board changes in process and business thinking, well, it's a whole new ballgame (thinking about things, such as, undecidability ought to get some play - risk-based thinking is too tied into the blinders put on by the gloating about the effectiveness of mathematics -- without really understanding why we see this).

So, the PIQ of a farmed-to-ee cannot just on the dime get to the state required by the farmer-outer. It could probably even be shown that the internal PIQ could have converged long before now (oh, but it was money and risk at issue). Yet, those with their heads in the abstracted clouds cannot see that their feet may be on other than functionally-capable grounds (grounding is necessary, as that which flies is really an earth-bound beast until it overcomes constraints - and, even then, it's only for some finite time).

But, another benefit that derives from PIQ is knowing how to do things when necessary, such as 1) how to avoid stumbling, 2) how to recover from a stumble, and 3) how to learn from the fact that a stumble occurred.

One thing attributable to PIQ is believability which does not accrue to heritage alone. It is not that there is a general negative view in regard to due diligence (though, that lowering direction may just be happening monotonistically), rather business has taught us to doubt since a lot of focus goes into efforts that save and prune (almost to the point of anorexia, it seems, many times) just so that some pocket (or set of pockets) gets filled (okay, unfair perhaps, but not entirely untrue).

Some programs, such as we see with medicine, need to have a focus that is above certain goals with regard to which the heated mindsets are eternally engrossed (and, for the most part, dealing with things of no substance).


09/02/2009 -- Lets face it, folks, undecidability needs to be discussed and adopted in any complex situational setting, especially if computers are involved. Only hubris pushes us to make loud exclamations about what we're going to do in the future.

05/18/2009 -- Testing in flight is within sight.

11/01/2008 -- Much has happened with regard to the schedule, the suppliers, and more. Boeing announced some insights about its 787 planning. Before that, the idiocy of a truncated (abbreviated) test cycle was changed.

Reminder: at this point last year (we can pinpoint the specific dates), there was still some talk about delivering in May 2008.

Modified: 09/02/2009

Tuesday, December 11, 2007


Evidently, there are some types of engineering that map well through the phases from conception through production where the products get out to the market quickly. Many, who strive for the new and human-less (green field, lights out, ...), want this. Okay, it has been shown to reduce touch labor, if things work.

Think quickly, though; how many of these types of things are hugely complicated in either type or aggregation across the set of sub-entities? How many have serious mission requirements that must be met?

Somehow, one sees that success of this type requires human talent, not easily obtained in many cases, meaning, of course, that it takes years of experience to 'get the arms' around some of these problems.

Ever wonder why the younger successes get their rewards in the 'virtual' world; that is, what do we have involved there but concepts mapping to a phenomenal experience that is ghostly at best (so, it can persist across sessions and the ether space - what substance is there?)? What we see here is that the new generation applies some bunch of tricks recently learned, many times academically, in a situation that is ripe for these tricks, is it not? The older folks could learn much of this, to boot, given the chance.

So many of the 'virtual' successes may appear to have some type of serious foundation, yet in actuality they are not that well founded; it's just that we've allowed opportunism to be rewarded (a whole set of analysis possible there). Oh, we like to put 'entrepreneurship' as a label to cover what is really going on (gosh, do we ever like to coat things platitudinally).

Let's see, so we can have the computer being used to make a large schedule dealing with oodles of activity steps across an enormous period of time and somehow the issues of feedback, analysis, course correction, etc. are overlooked. Well, is it people or the computer that drive a process?

Too, we see the computational put into the place of real experiments. Mind you, this is where we ought to be putting our attention; yet, the issues of quasi-empiricism have to be addressed and are essentially being ignored.

Then, we have the talking PR events.

Where or how does one find the way through the morass? Is there something about this that might point back to a world view (and its associated wet-ware wiring) that is heavily influenced by computer gaming?


12/17/2010 -- These types of issues are continually there for our resolution.

09/02/2009 -- Lets face it, folks, undecidability needs to be discussed and adopted in any complex situational setting, especially if computers are involved. Only hubris pushes us to make loud exclamations about what we're going to do in the future.

05/18/2009 -- Testing in flight is within sight.

11/01/2008 -- One has to ask what is the parallel in engineering, and program management, to the fiduciary duty of finance. In regard to the 787 project, much has happened with regard to the schedule, the suppliers, and more. Boeing announced some insights about its 787 planning. Before that, the idiocy of a truncated (abbreviated) test cycle was changed.

01/11/2008 -- One has to wonder what amount of forethought can make a process successful the first time through. Is it not our experience that a least one pass is required in order to prove the process (somewhat indicating that proof is not in the abstraction rather it is in the realization)?

Even highly trained and capable personnel, such as this medical article discusses, require a check-list to help guide a process through all the variations of a situation. ...

Oh yes, using a checklist from seemingly similar processes can be a good starting point; yet, we have to ask the question, do we not still need a run through? For a plane OEM, does this not involve several things such as flying, working the production issues, and more?

Modified: 12/17/2010

Thursday, December 6, 2007


For starters, ‘workmanship’ applies to more than hand (touch) labor; we could talk, as well, about white collar or even capital management, which is supposedly the epitome of everything business. We even expect good workmanship of our medical personnel; one might use due diligence, to boot.

One source of oops does relate to the workmanship issue and the dummying down of work.

One may ask why workmanship may be necessary. NASA has a good take, though the focus is on the effects of bad workmanship on parts. One would think that an OEM, of any sorts, would worry about this; can outsourcing workmanship be guaranteed by inspection only? We’ll have to discuss that more, as evaluation of a product requires some insight into how that product happened to come about.

Dummying down may result from work outsourcing or from process pruning.

In the former case, if the one whose function was outsourced remains to oversee the outsourced work, the role becomes watching and evaluating what comes back in from the outsource. The watching is from a distance usually, unless the outsourcee goofs and needs special attention. The evaluating is of product only, as many times the process evaluation is, or is supposed to be, done by the outsourcee.

In the latter case, some process pruning may be done on stable processes that have been around awhile, when the pruning step has a more complete set of information to work with (we’ll look further at this) than is possible on the first (or even the second) pass of an entirely new process.


01/26/2009 -- Expect more effort in firming up the earned-value (and related) discussions.

12/07/2007 -- The post was partially motivated by observing up-close, on several fronts, the effect of outsourcing in the home-building trades. Coordination at even this local of a level raised issues resulting in effects not seen prior to the outsourcing craze. Then, if one extrapolates from this simple scenario to a mega-outsource as we've seen getting focus, one can begin to why we're in such bad fixes about everywhere one looks. There is a better economic basis than that laid on us by the abstractionists (securitization, as one effect).

Modified: 01/26/2009