Sarah Sheard's Thoughts and Theories

Wednesday, September 27, 2006

What if this *IS* the second time through?

How often have you thought, "Gee, if I had this life to live over again, I would definitely..."

An article in Science News discusses a famous snake-bite doctor, Dr. Sean Bush of the Venom ER TV Show, receiving a page that his 2-year-old son had been bitten by a rattlesnake. He was able to save his own son's life with all the skills he had studied in medical school and practiced by saving other peoples' lives.

Now, if you knew that when you grew up, your own babies' life would be risked by a snakebite, wouldn't that make you want to go into snakebite medicine?

Now suppose for a minute that Dr. Bush lived a different life, where he became a dentist, possibly, or a proctologist, and one day his baby son died because there wasn't a practiced snake-bite doctor in the hospital to which he was flown? Dr. Bush certainly would have said, "Gee, if I had it to do over, I definitely would have become a snake-bite doctor!"

What if this IS the life that we are living the second time through? Doesn't it look like once in a while, things turn out just exactly the way you would want them to, and you had no idea they were going to until that exact minute?

Thursday, September 21, 2006

Sarah's Theory on Writing Things Off

I believe that we all prosper by simplifying the world around us. A mental model is just a simplified picture of reality. We base our own actions on the thoughts that make sense, and predict good results, after running them through our "prediction" routines in our brains (in the frontal lobes, actually).

What this means is, essentially: we all learn to write things off as unimportant, and focus on other things. Most new things are written off as aberrations, not actually thought through as possibly being worth changing everything you think so that you can account for the new idea. (This is why sometimes it takes the old folks dying off to allow science as a whole to make progress on some front. -J. DeFoe)

For example, If you wear plaid slacks with a floral shirt (You know who you are, Pat): they don't exactly see your fashion statement as setting a new trend; they see the situation as you being weird, and they write you off.

Therefore when you are trying to announce or teach anything new, you cannot count on people welcoming your ideas with open arms. In fact many people, engineers and business managers included, consider anyone promoting anything they don't already know about as a charlatan until proven otherwise.

I taught a tutorial on Tuesday. The tutorial highlights the research currently being done in areas called Complex Systems, an interdisciplinary set of sciences. The class shows how some of this research could have direct impact on some aspects of systems engineering if we could only spend some time doing the necessary "development" to make the research results practical. The tutorial went extremely well, with many students essentially speaking a different language about complex systems by the end of the day, being able to identify chaotic, fractal, and complex aspects of the systems they know and deriving a couple of new principles.

After the tutorial, however, some students told me that the principles discussed in the course are "nothing new". I countered that they are; that we never had the computing power before to study situations where multiple objects must be tracked simultaneously through multiple changes and adaptations. Still, these students insisted that nothing of interest is going to come of this research. They have already made up their minds that the work they would have to put into investigating this will not pay off, so they are not going to do it.

Fortunately, complex systems does have an answer to this.

The most "fit" systems engineers and the most "fit" systems engineering organizations will survive the competition. Engineers who do not learn about multi-agent systems with emergent behavior and different kinds of math and analyses will simply be weeded out in competition with those who do.

A course description brochure is located in the second paragraph here.

After the tutorial I considered changing the focus of future tutorials to outline a set of specific methods and techniques that the systems engineer can apply at clearly designated points in the development of a complex system. However, my experience with specific techniques tells me that doing so will make it even easier for people to apply the wisdom of these fields automatically and unthinkingly. That will be like jumping into Phase 8, Proceduralization, of the life cycle in the "Silver Bullets" paper, the phase where the improvement effort you expend has no value at all, even negative value.

Perhaps I shall proceed to evolve the current course instead.


Friday, September 08, 2006

Sarah's Theory of Dress Codes

When employees (engineers, for example) are hard to find, the dress code gets more casual. When jobs are hard to find, the dress code gets more formal. The dress code lags the job situation between six months and two years, though. I believe now jobs are harder to find than five years ago. People are also wearing more jackets and ties.

Thursday, September 07, 2006

Sarah's Theory on Process and Fractals

In the last ten to fifteen years, corporate America has concerned itself much more with process than it had since mass production and Henry Ford. This time it's different, of course...we in "the industry" aren't talking about the process of making widgets, we're talking about the process of producing knowledge products, like software, and engineering designs. This is tough stuff, and you're never going to put numbers into a machine, crank the handle, and out will come an original thought.

This brings us to some new theories. Here's one, why process focus is like fractals:

Suppose you were six years old and had to cut a shape out of construction paper. You would probably cut a square, or a circle, or maybe a triangle. (or as close to those as your six-year-old fingers would get.) You wouldn't try to cut a fern shape, or a broccoli shape, or a riverbed shape with all its twists and turns. Why? Because circles and squares involve a lot less to cut than things as convoluted as ferns and broccoli.

So the question has to be, how is it possible that ferns and broccoli can arise from nature?

It's because the way things grow naturally does what ever is simplest process-wise, not whatever is simplest in form. It turns out that ferns and broccoli and rivers are fractal or self-similar: you can look at big pieces and small pieces and they have the same shape. Computer artists have discovered the ease of making use the same code for the small pieces and simply scale up for bigger pieces. It's very easy to make a fractal look complicated with ridiculously little code, because that code is reused at all scales.

Compare this to making a square or a circle. You have a single shape existing on only one scale. Things are done once, and a "thing" comes out.

That is similar to the way people built, say, software, before the process focus. You examined a problem, built some code to solve it, and you were done. Next problem: build different code.

Now of course, people are much more aware that it's better (for reliability, maintainability, and even initial development cost of new software) if you can create reusable code, work out the bugs, and then keep using what works.

This is what nature does with fractals!

First photo is from the satellite Seawifs, of a river basin in Norway. See