I remember hearing Harry Markowitz speak in the 1950’s about manufacturing simulators.  (He has since become a significant figure in finance and economics and a winner of several major prizes, including the Nobel Memorial Prize in Economics.)  The problem being attempted was trying to model the entire floor of a manufacturing plant full of machines, people, conveyors etc.  Experience back then was that no such model could be used as a central control over the factory being modeled, simply because the real factory floor never behaved as the model predicted.  Things would break down, people would do dumb things and processing times would vary from what had been assumed.  Thus the centralized model never corresponded to the real situation on the floor and, thus, could not be used to issue orders to the factory floor.  Thus, what started out to be an optimization tool became totally disconnected from reality and of no practical use.

The thought I took from Markowitz’s talk was that the best you could do was to develop the dispatching rules to be used at each station on the floor.  These rules told the operator what to do under the various circumstances that might occur.  So the real purpose of a manufacturing simulator was to model the factory floor with the purpose of exploring the consequences of various dispatching rules.  With a good set of such rules a real factory might run efficiently even with the mis-occurrences of reality.

Now, a complete change of scene.

In 1970 I attended  and presented a paper at a conference in Bordeaux, France on” L’Homme et L’Informatique” sponsored by the private Institute de la Vie.  For several days luminaries from the computer field discussed various aspects of the relationship between man and information.  Among all this prestige, one speaker stood out and delivered the best speech I ever heard, anywhere.  Among all this formality and importance, Richard Bellman took off his jacket, loosened his tie and stalked the front of the auditorium like a caged lion.  Bellman was a widely known mathematician and the inventor of Dynamic Programming, a mathematical technique for dealing with large systems of non-linear equations. His talk was never published (I inquired of both the Institute de la Vie and Bellman’s office at CalTech).  The reason for this was simply that the basic math underlying his theme turned out – I was told later – to be mistaken.  But bear with me anyhow.  A précis of his talk goes as follows:

1.       I have specialized in dealing with large systems of equations.  Mostly, these systems are used to model some part of industry, economics or society.

2.       I have observed that as these systems of equations get larger and larger they tend to be increasingly unstable.

3.       When this conclusion is applied to society I am alarmed by the prospects for instability in our society.  (He expanded on this thought quite a lot, producing a very gloomy outlook.)

4.       Nevertheless, I am an optimist about our future!

5.       I come to this conclusion by finding that if each of the small coefficients of the variables in these systems of equations is adjusted just a little bit in the right direction, the systems become stable.  If everyone in our society can be persuaded to do just a little bit for the society as a whole, there is a future for us all.

It was brilliant and uplifting – and, apparently, wrong.  What a pity.  I have always remembered this talk with fondness and admiration.

However, a thought has occurred to me lately that relates to Bellman’s talk and ties it to the work of Markowitz which I mentioned at the start of this essay.

In the sense of both of Bellman and Markowitz, perhaps some fruitful work could be done to develop the “dispatching rules” that each of us should use in our lives when interacting with others and with society.  If any progress could be made along this line it would be beneficial to everyone.

It might be particularly interesting if the resulting “rules” turned out to be those that religion has been advocating for so many years.  Yet I doubt that this will be the case, since the world which religion dominated has been notable for instability and wars.  So there must be something different from what religion is advocating which is needed to make for a stable and peaceful society.

That’s as far as I can take this line of reasoning.  If there is any value to it, others will have to take the next steps.

John W. Weil