Showing posts with label Edison. Show all posts
Showing posts with label Edison. Show all posts

Friday, July 11, 2014

Feed-forward loops: spiralling out or shutting down


Basic for our understanding of processes that run out of control is the idea of ‘feed-forward’ instead of feedback. If we go shopping with money in our pocket, the spending of that money, sadly but naturally, causes its own end - though credit considerably loosens the natural discipline of this feedback. When you are a good student at school, you progress and finally you graduate and will leave school altogether. Also, when you do not learn easily you will leave school anyway. Attending school leads to leaving school anyway.
Other processes, however, have a peculiar tendency to reinforce themselves. A well-known example is the chain reaction. Nuclear fission emits radiation, which triggers more fission, which triggers more radiation, until the entire system explodes. Many biological, psychological and intellectual feed-forward processes exist, as do social and economic processes.
When something grows and everyone expects it to grow further, the expectation itself may trigger more growth. Things may work up to a frenzy, go crazy, hit the roof, until a point of collapse. This happens in stock markets.
Expectations are important in pricing. If people expect prices to go up, chances are that prices go up. When people start to hoard, shops may run out of stock, so people then want their own stock, and so they hoard. Or think of the bankers’ nightmare, a run on the bank.
Expectations are just one example of self-reinforcing processes. When we are nervous we act nervously, which may make us more nervous still. Agitation leads to more agitation. Once we feel gloomy, we see much and think much that feeds our gloom and, perceiving how bad it is to feel gloomy, we feel gloomier still. The main cause of depression is depression. Envy shows us things to be envious about, thus feeding our envy. When we are scared, our fear makes us more frightened, and our fantasies about scary things that may happen frighten us even more.
Loneliness, too, may snowball. When we feel lonely and therefore seek company, we are using feedback. But we can also withdraw into ourselves, and become miserable and reclusive. Likewise, timidity may lead to clumsiness that makes us even more timid. Stammering is so embarrassing that we keep stammering. When we think people want to kill us, everywhere in the streets we see people following us.
Aggression shows a different vicious circle. Merely expressing aggression does not make us more aggressive, but when we express our aggression toward other people they will resist, defend or avenge themselves, in a response that may provoke our aggression again. Aggression easily escalates and becomes a self-reinforcing process between parties.
An innocent example of social feed-forward is that of spectators in a stadium. At thrilling moments, spectators will rise so as to miss nothing of the spectacle, but by standing they hinder one another’s view. Likewise, in a noisy room people have to raise their voices to be understood, and thus raise the noise further. As another example, well-to-do people who live in cities want to move into the outskirts. This movement of people diminishes the green space around cities, so people have to move further into the outskirts, thus reducing green space further.
Or take positive examples of feed-forward. The more social connections we have, the more chance we have to make new connections. The same holds for fame. Once we are well known, it becomes easier for us to make the press. Anyone who is already well known as a writer has an easier time finding a publisher.
An economic example of feed-forward is that of compound interest. Debts that incur interest tend to increase, as interest is charged on interest. This applies both to debtors and to creditors. The more money we have, the more we can make, and the more debts we have, the more money we need. The Bible mentions the making of more money with money, in the parable of the talents. Also, the more we have of, for example, intelligence, the more chance we have to become even more intelligent.
The greater the advantage we have, the more chance we have to exploit that advantage. The greater we are in arrears, the less chance we have to make up arrears. An important issue here is that of ‘critical mass’ or ‘critical limit’. With personal capital a critical boundary exists, when net received interest surpasses living expenses. Once we have so much money that we can afford to pay financial, fiscal and legal advisers, we find it easier to earn even more money. Once we cross that line, moneymaking reinforces itself.
Feed-forward loops, both those that spiral out (accelerate) and those that spiral in (decelerate), often exhibit a critical limit as well. Above or below that limit, the process of spiraling suddenly shifts into another gear. The concept of critical mass is well known in chain reactions. Feed-forward processes, when they surpass a critical level, widen differences. When differences in social equality are thus widened, such a process is undesirable, but when the threat of universal entropy is thus reduced, the process is refreshing.

Today’s improved nutrition and hygiene, which have decreased mortality (especially infant mortality), emerged from science and technology. Science and technology have become the carriers of major economic and social breakthroughs in western civilization. Through the ages, the great shackles that have limited the scientific spirit of inquiry and experimentation have been poverty, superstition and priesthood. When enough critical mass was reached to escape from these three brakes, we entered a takeoff phase and feed-forward loops started working.
The Industrial Revolution began with invention of the steam engine. The First Industrial Revolution produced prime movers, and consequently transport. That technology carried important self-accelerating elements. Once we can make machines of a certain precision, we can make new machines of even greater precision. When we can produce better steel, with that steel we improve other machines, so that in time we can improve the steel-making process itself.
Science functions with this feed-forward phenomenon as well. The more we know the better tools we can develop and the more new knowledge becomes available. Power and transport are the classic examples; computers, software and telecommunication devices are the present-day examples. We design and test chips with equipment that uses chips. We use computers to design new computers. We use software to generate new software. Soon robots will help us to make new robots. The most fundamental invention, as Norbert Wiener has already noted, has been Thomas Alva Edison’s research and development lab that spawns inventions.
The self-reinforcing developments of science and technology, and the economic use of these developments, comprise the driving force behind the takeoff of modern society. Today, empirical and pragmatic physical science has become the greatest breeder reactor in the development of the human race.
Economists talk about the takeoff phase, that phase that a country enters once it has acquired an adequate base of knowledge, finance, management and infrastructure to develop new industries. Poor countries that lag behind tend to run further behind, because their start-up base falls below financial critical mass. Even if developmental financing is provided, a country’s human capital base may lack skill, knowledge, understanding or attitude. The more industry is already present, the more we can add to it; the better the education that is already available, the more we can develop education. In any situation with feed-forward and critical mass characteristics, advantaged parties advance faster and disadvantaged parties lag farther and farther behind.
As a result, the stakes are increasing. In a high-tech environment, the critical mass of research and development investments may have become so high that most countries drop from the race altogether. They become technology consumers, not technology producers. In many technological fields only Japan, the United States, and Western Europe (if united enough) can still compete.
The more capital we already have, the easier it will be to borrow more. The greater our need for money, however, the more difficult it will be to get a loan. This is another example of feed-forward, and is one that becomes important in international banking. Some countries are presently on a borrowing spree. On the very day that I wrote this, Brazil had taken a loan of $12.5 billion to pay off interest on its former loans. The poorest countries pay the highest interest rates, and the richest countries the lowest. Scores of countries have fallen into this kind of financial ‘black hole’. The saddest and also the funniest aspect of their unhappy situation is that such countries are not allowed to collapse, as they might take their debtors down with them. There is a sadomasochistic elegance to all this.

Many social interactions also function with feed-forward. Groups and institutions develop their peculiar, distinct flavor from small differences at first. Out of the trivia of daily life emerge patterns of ‘how we do things around here’. People who do not fit are recruited less often, are promoted less, and leave earlier. This is a major reason that bureaucracy is the most common disease of organizations. Rules and control diminish initiative and motivation, increasing the call for even more rules and more control. Paul Watzlawick has found this process in many walks of life, and called it ‘more of the same’.
A government bureaucracy organizes deregulation by establishing deregulation committees and deregulation procedures. Organizations solve problems of coordination by designating coordinators, thus complicating the already complex line of communications. ‘More of the same’ also exists when drugs are used to fight the undesirable effects of other drugs, or when surgery is used to fight the undesirable effects of other surgery.
Feedback loops, on the contrary, maintain balance. When this balance is undesirable, the loop becomes a prison, and getting out involves us in a prisoner’s dilemma, as escape becomes self-defeating.
Self-defeating efforts may be seen in people who chase exclusive goods or exclusive distinctions, in exclusive clubs that want to expand, or in high school students who want to distinguish themselves with the same clothes or shoes or hairdo. These, what the sociologists call ‘positional goods’, provide satisfaction only to the extent that others don’t have them.
An escape from this paradox does exist. It exists in the world of advertising that offers exclusive articles wholesale, in contests that everyone can win, and in books that offer ‘Secrets for the Millions’. This is no longer the ‘revolution of rising expectations’; this is Alice in Wonderland.

Feedback between individual actions that functions well and results in a desirable balance, constitutes the ‘invisible hand’ of Adam Smith. But life often does not work that way. Many social processes become uncontrollable through the addition of separate individual actions. Think of the reaction of the spectators in a stadium, or of people who try to speak against the background noise of other people speaking. Individual choices, each made separately and not necessarily taking into account the interaction between those choices, often combine to produce unexpected and unmanageable social consequences. Such self-defeating behavior of great groups of people may be funny, is sometimes destructive, and is often tragic.
Malthus assumed that population grows exponentially, while resources such as space (and so food) are finite. Thus, any population will grow until all live in squalor. We have staved off this miserable future by a staggering increase in production of food and a gradual decrease in production of offspring. Still, many regions in the world are Malthusian right now.
Garrett Hardin, in his famous article on ‘The Tragedy of the Commons’, generalized the problem of population explosion. He linked population increase to the problem of individual decisions that maximize individual benefits and simultaneously deplete common resources. Such decisions exhaust these common resources to the point of collapse. A similar example of this process is environmental pollution.
The commons used to be a village meadow that was owned by all. Farmers had their own fields but also shared one common field. If you were such a farmer you could let your sheep graze there, for such grazing was free. After all, the common was owned by everyone. However, your colleagues would do the same. There would be no halt to overgrazing, so profits would diminish, the sheep would eat up all the land, and the villagers would have less to eat themselves. You can set out another sheep to compensate for your reduced yield. You would reduce the yield per sheep further, but you would win, as the benefit would be yours, and the cost would be spread out.
The sea is another form of the old village common, and fishing and whale-hunting tell the same story. Present individual gains lead to later collective disaster. Even when you know that in about ten or twenty years no whales will be left, you may still regard it disadvantageous to stop sooner than the others. The system is ‘hell-bent’; it has to collapse. Overfishing of whales may be barbaric, but it is also stupid. It is done purely from self-interest, with nothing enlightened about it. It makes no sense to kill whales more quickly than they multiply. Plain business sense would maintain an appropriate level of harvesting.
Hardin points out that human population keeps growing for the same reasons. Particularly where people depend on children and grandchildren in their old age, producing another child provides a direct advantage. The fact that every new child actually lessens a family’s present means of existence is an indirect disadvantage. Ultimately, overpopulation may mean famine for everyone. Most self-reinforcing processes cannot be stopped by rules. They can be stopped only by a change in mentality.
Hardin concludes that there are no technical solutions, only moral solutions, for those problems. This conclusion is an overstatement, as his equations include only a weak time factor. Even if individual loss were greater than individual gain, only an immediately visible loss would make people see the light. Enlightened self-interest becomes difficult only when people have small lights themselves. Under a glaring sun, everyone can see the obvious. Often, when people finally begin to notice the adverse effects of overusing their common resources, they organize themselves and establish rules. The adverse effects that suddenly become clear to them enlighten their interests.
This awareness delay becomes tragic when feedback begins to work only after the point of no return. For example, farmers may notice soil degradation only after the process has become irreversible. A second type of tragedy occurs when some people see the effects early enough, but others stubbornly refuse to listen and see. This situation is a variety of Tuchman’s ‘folly’.
Special interests press to overstretch common resources, especially in our modern society with its dazzling array of organizations. Many people belong simultaneously to several organizations that have conflicting objectives. This array of objectives may stimulate open debate and good discussion, it but also helps special interests to lead their own lives. Such interests are not being balanced, even when the representatives of those organizations are willing as individuals to consider their conflicting interests. The rise of special-interest institutions, with their often considerable expertise, prejudices broad and reasonable deliberation.

Several global crises are presently brewing because of self-reinforcing developments. Three major related crises are the population explosion, which is slowing but continues; the increasing gap between the rich and the poor; and the inertia of bureaucratic politics, both domestic and international. These growing crises are examples respectively of feed-forward loops operating above and below critical mass, and of self-perpetuating negative feed-forward and feedback loops.
Available as e-book at http://www.onlineoriginals.com