There are two concurrent trends in technical systems evolution. The first one is the well known trend of ever increasing controllability of parts of a technical system (which, by the way, means not only that the controllability of already controllable parts increases, but also that the number of controllable parts increases). But some parts are made controllable so that they help control the overall behaviour of a system. It was conjectured below that the number of such parts that participate in the control of the entire system also increases. The more democratic a system is (i.e. the more its parts participate in control of the system's behaviour) the better.
Date: Tue, 9 Dec 2014 22:59:08 +0000
Dear Mr Karasik,
I was reading the chapter 7 of Creativity as an exact science and I was thinking about the first Alshuller’s law about system completeness. I researched in your journal finding your article: http://www3.sympatico.ca/karasik/re_law_of_completeness.html
To say that the system SHOULD contain X parts is not very helpful in practice… I was thinking about the characteristics of a minimum viable system and maybe the “law” could be restated. It is just a quick idea that came to me reading about Wright brothers:
So maybe it would be better not to talk about minimum parts of a system but about minimum controllability of the system. The problem is: “What means in practice minimum controllability?” I do not have an answer yet. In Altshuller’s law corollary, he says: “For a technical system to be controllable it is necessary at least one of its parts to be controllable. (Changing its properties in a way required by the controller)” However, the corollary, again, seems too vague in practice.
Date: Wed, 10 Dec 2014 00:47:10 +0000
Dear Mr. Acosta,
I agree with you. I don't recall examples of how this law was used in solving problems. I only can imagine how the law could be used in making vague forecasts. For example, if you have a system that does not have some parts you can forecast that they will be added to the system at some point in time. For example, since house has none of the parts but according to the law any system should have them, one can forecast that at some point in time houses will have engine, transmission, working means, and controller. Some houses already have them. They are mobile homes. One can forecast that at some point in time all houses will become mobile.
Apart from such forecasts (which might be wrong or not), I don't see any utility of this law.
Regarding controllability, there is already another law of the progressive increasing of the controllability of a system.
Regarding Wright brothers, they invented the means of aircraft control other than the moving the body of a pilot. These means were wing warping (later replaced by ailerons), etc. Not only pilot's body started to control the flight but also other parts of aircraft.
May be there is indeed a trend of increasing participation of system's parts in its control. It is like the progressive increasing of democracy of system's parts. Ever more parts of a system participate in its control.
What do you think ?
Date: Wed, 10 Dec 2014 16:51:15 +0000
Dear Mr Karasik,
Well, maybe I would say: “More parts of the system are under control”, this idea reminds me more a dictatorship than a democracy of system’s parts. In my opinion, ideally every atom of the system must be under control, but the control means must remain simple, if you can take full control of the system's parts using just one part (i.e. one button or maybe self-control??), much better than two buttons. Maybe the “home button” of iPhone can be a good example of control: https://www.yahoo.com/tech/the-5-secret-features-of-the-iphone-home-button-84770375499.html
Anyway, my ideas were focused in the laws, which determine the start of life for technical systems (Static laws in Altshuller words). Maybe future homes will integrate the four parts; maybe the future homes will increase its controllability, but what determines the born of a new technical system? It is obvious that the four parts are not mandatory for a new system, so I am thinking that one priority for a novel system is to reach at least a “minimum level of controllability”. Of course, the minimum level of control could be zero in some cases such as homes, bigger for the early cars, bigger minimum control was required for the early planes and I suppose that even bigger minimum control requirements were required for the early space transportation systems.
For example in early cars to control the start/stop and 2D motion was enough. In planes, the movement is in 3D so the invention of 3-axis control was the key of success of the early planes. I suppose that the other inventors thought that 2D control (no control or just using body weight) was ok and focused on other parts (i.e. engines) without reaching a minimum level of controllability.
It is just a quick idea, what do you think?
Date: Wed, 10 Dec 2014 20:12:12 +0000
Dear Mr. Acosta,
To say that "more parts of the system are under control" is half truth. The whole truth is that more parts of the system control the behaviour/function of the entire system.
Wright brothers started controlling wings so that they would control the plane's flight. So, wings became controllable and controllers at the same time. The first is dictatorship and the second is democracy.
There are many criteria of viability of a new system. The minimally sufficient level of controllability is just one criterion. There are other ones, e.g. the minimally sufficient level of reliability. Another one is the principle of not exceeding published in the journal, etc.
Date: Thu, 11 Dec 2014 00:05:01 +0000
Dear Mr Karasik,
I would like to add some ideas about the topic. Let us assume that we have the subsystems X Y and Z. The integration of the subsystems allows the emergence of the system S.
If the function/behaviour is controlled only by the higher system S we have a “dictatorship”.
If the function/behaviour is controlled only by subsystems, we have maybe some kind of “anarchy” (??).
Therefore, for a given system S to be stable in time, S must generate control on its subsystems but the subsystems must generate control on the higher level too.
I am thinking that if the higher-level control, say the government (and laws) of one country does not recognize the rights of people and becomes a dictatorship state then conflict will appear and the system as a whole will be unstable and will disappear in time. On the other hand, if the control is executed by individuals, with time, their ambitions and power hungry will generate conflicts between subsystems and eventually become a dictatorship state.
So as you said, it seems that "democracy" could be the most stable option.