Cooperation

A person by himself will never rise above the Robinson Crusoe stage, and it could be argued that DeFoe was an optimist since Robinson Crusoe greatly benefited from wreckage that he could use to make his existence more bearable.

If one were to purely exist by one self without any outside support one actually would have to live to a much more primitive level, even if one were to have all the knowledge in the world, from farming and hunting up to computer science and nuclear physics.

The best such a person by himself could do would be a primitive combination of hunting and subsistence farming. It would be possible to rise beyond this level if one has access to iron ore, which would enable one to fashion steel tools. By going by our assumption that one has all the knowledge in the world, one could even fashion guns.

However, making guns for hunting (portable guns) requires a large amount of sophisticated tools, and chances are that it would take near a lifetime to fashion these tools, especially since while making these tools, one still has to hunt, grow food and maintain a shelter. Furthermore, living by one self, one would be lucky to live a reasonable life span.

Families are a much more powerful unit, but that assumes that one is lucky enough to find a person of the opposite sex, which assumes a whole lot of other things, such as where this other person came from.

Once we make the family leap we are stuck with in-laws and we have to face that we are part of a community.

Communities are founded on the concepts of cooperation, communication and negotiation. There cannot be a community if there is no cooperation, and cooperation requires communication and negotiation.

There are communities that are coercive, but those communities still cooperate at specific hierarchical levels.

Israel is a cooperative community; the Israelis cooperate, and the Palestinians cooperate. However, the Israelis do not cooperate with the Palestinians, which is where the coercion comes in.

For the sake of this discussion I would like to define coercion as forcing people to do unreasonable things against their will. I don’t like to pay taxes, and the community can force me to pay taxes, but it is not coercion since I reasonably believe I should pay taxes. If I am the only person in the community that has to pay taxes then I am being coerced. One can see that fairness sneaks into this concept also.

An ideal community has little coercion, and exists almost entirely on pure cooperation.

It is important to realize that the need for cooperation has made a giant leap in the last century. This is caused by two independent trends that started developing very quickly between 1900 and 2000.

The first trend is the explosion in science and technology.

The second trend goes back to our earlier discussion of Jared Diamond’s book; humanity has circled around the globe, and we no longer can simply pull up stakes and live as a rugged individual on our own at some spot on the world where there is no one else, or where we can just slaughter the weaker population. If we cannot slaughter anymore we have to cooperate.

With regard to the first trend, there was a day that one exceptionally smart individual could be familiar with all technological and scientific concepts known to man. It is hard to define when that was, but while Newton might not have been familiar with the latest metallurgy, naval technology or the latest textile technology, it would not be impossible to argue that at that time one person could still grasp the world at a scientific and technological level.

There is no doubt that this had become significantly more difficult with more detailed chemical, nuclear, electrical, aerodynamic and mechanical discoveries and inventions at the start of the twentieth century.

Before that time, technical specialization was an economic advantage, but after that time specialization became a necessity. At the beginning of the twentieth century, cooperation between specialties had become a prerequisite to the march of science and technology, and today it is difficult to think of any technical or scientific advance that is not the result of cooperative work between many engineers and scientists.

As a matter of fact, it has been argued that technology is being held back by a lack of communication between all the specialized branches of science and technology. If we could communicate better between all the branches it would be easier to combine advances into new and useful products.

There are even indications that the Internet has become a major facilitator in communications between specialists, and that the improvements in data retrieval technology on the Internet will result in a new wave of technological breakthroughs.

This need for technological cooperation takes place at any level.

At the start of the century one engineer could design a car, at the end of the century it requires at least 10 engineers. Engineers have not become dumber; cars have become more complex and reliable, and are being manufactured for lower relative cost, which requires more complex technology.

At the start of this century, sewage technology consisted of a toilet and a pipe to the nearest body of water. Today the connection to the body of water is hard to find because of all the technology that sits between the toilet and the clean water discharge.

Both trends also cause us to have ever wider horizons. There was a day not too long ago where I did not care whether there were lions in Africa or not. I only cared about the lion at the local zoo, since that was where I would go to see one. Now technology and the trend to a single world have made it possible for me to see lions in their natural setting and now I need to cooperate with people in Africa, who I do not even know, to ensure that there will be lions there by the time I get there.

Will this ever-increasing need for cooperation ever end? Not for the time being, but there are some interesting technologies that have and will reduce the need for human-to-human cooperation.

The Internet and Cooperation

Not too long ago I would shop for a new refrigerator in a store and decide on the model I would buy with the help of a refrigerator salesman. The last time I needed a new refrigerator I went to the local discount store and asked the salesman to help me select a refrigerator. He flat out told me he had no time for that, and that I should make my selection based on information available on the Internet. Once I had picked a refrigerator he assured me that he would sell it to me for the lowest price in the area.

This shocked me and I went to another store that is known for its customer service.

I told the sales man about my experience at the other store and asked him for help. But he was not helpful either! It took me a while to figure out that he was afraid that he would provide the selection assistance, and that I would then buy the refrigerator at the other store for a lower price.

Not until I assured him that I would not do that, and reinforced this by pointing out that my brother also had bought appliances from him recently, could I get him to help me.

In those days I did not have a high speed internet connection, and I did not relish the idea of having to spend hours downloading pictures, reviews and specifications.

Ironically with my new high-speed connection, today I would probably select my refrigerator on the internet and shop for the lowest price.

---------------------------------------------

The internet is just the tip of the iceberg. While doctors do not like to admit it, disease diagnosis is very adaptable to expert computer systems. Such systems would reduce the need for contact with medical specialists.

Probably by the time this book is read, the world will have become familiar with computer programs that are based on Cyc, which is a database of common knowledge.

This database has been under development for over two decades and includes commonsensical rules such as:

When a person is dead he stays dead.
A water glass is carried with the open end up.
People who are claustrophobic might not like enclosed spaces.
A tunnel is an enclosed space.
Anthrax is a disease
Anthrax is a rock band

Through software these rules can be interconnected to come to conclusions. Some will argue that it will never be true common sense, but then again who has “true” common sense anyway?

However, such a tool with a useful interface will reduce the need for cooperation. Robinson Crusoe certainly could have used such a tool.

So in the future the need for cooperation could lessen, and we might have a modest return to rugged individualists assisted by very able computers.

Is this bad? Probably not. Cooperation is time consuming and exhausting, and therefore best used on levels where there are no alternatives.

In other words; let’s save cooperation for our marriages, families, school boards, and natural disasters, and provide streamlining for situations where cooperation can be replaced with alternatives such as the internet.

What is cooperation anyway? How does it come to be?

The Science of Cooperation: A Brief History

History is a lesson in cooperation, but in the early eighties a group of scientists stumbled on a set of tools that provided a much higher level of understanding on cooperation.

For many years a classic dilemma had provided scientists, game players and philosophers with much food for thought.

This dilemma was often referred to as the prisoner’s dilemma. It assumes there are two people who have been arrested as suspects in a crime. It is suspected that the suspects had been working together and the authorities are trying to have them confess by interrogating them separately hoping that one will testify against the other.

For the suspects this poses an interesting dilemma. If one confesses and testifies against the other, his punishment would be light while the other’s punishment would be severe, and vice versa. If neither confesses, they will probably go free. If they both confess they both will get punished, but probably not as severely as the person who refuses to confess.

If you are one of the suspects what would you do?

Upon some reflection one would conclude it probably pays to confess, that would ensure that at least you will not get the severe punishment. However, if you really knew the other suspect, and are comfortable that he will not confess, you will not confess either. But, you are not in contact with your partner and the interrogators might be able to convince you that your partner confessed.

It would also make a difference if the type and severity of punishment is varied in this dilemma. If the death penalty is the worst option, and confession would prevent it, confession is almost a certain option. In the end there are a thousand variations.

One variation is to have to players play the prisoner’s dilemma a number of times.

How does the decision your partner made on the prior turn affect your decision on the present turn?

If your partner squealed on you the last time you probably will squeal on him this time. What if you had a chance to discuss your partner’s last decision? What if he convinced you that he would never squeal again?

This game is easy to play on paper and has been played on paper many times. After playing this game, game theorists have come up with interesting and useful conclusions.

They discovered that if people started by cooperating they will probably continue to cooperate. If people did not cooperate at first, but at some stage started to cooperate they will probably continue to cooperate.

If over a long number of turns you want to finish with consistent cooperation it is better not to let the players know when the last turn is, because it is very tempting to confess on the last turn.

The success of strategies can be measured by assigning points to each outcome. For example if you squealed on your partner and your partner did not, you get 5 points and your partner gets nothing. If you both squeal each get 1 point. If neither squeals you both get 2 points.

If the object is to get as many points without regard for the other, there are certain strategies that might be most successful depending on the number of turns there are in the game.

The more turns there are, the more likely it would be that the players start to work together, but if there is only one turn there is little benefit for not squealing on your partner.

Many of these things make sense. It sort of fits the Teddy Rooseveld line: “Speak softly and carry a big stick”. In other words; if there is a consequence on the long term, one is likely to cooperate.

But it gets better than this.

In the early eighties with the advent of personal computers, some scientist decided to have a prisoner’s dilemma challenge. What this meant was that any person could submit a computer program for a prisoner’s dilemma strategy and those programs would run against each other to determine which strategy would be the most successful.

In the first competition there was one clear winner, which quickly was named “tit for tat”. This strategy was only a few lines of code.

First cooperate (do not squeal)
Check what the opposite player’s last move was
If the last move is cooperate, cooperate on the next move
If the last move was not cooperate (squeal), also do not cooperate (squeal) on the next move.

What this program does is cooperate until the other side does not cooperate, then stops cooperating until the other starts to cooperate again.

This simple program ran against many very complex programs, but simply came out on top in most cases and was never found to be a bad loser.

This result was so startling that they ran another competition. Again “Tit for Tat” was the big winner. This resulted in a big pause for thought. What if the rewards were changed? Again “Tit for Tat” was the real winner. It was possible to write a special program to beat one “Tit for Tat” player, but in random settings “Tit for Tat” won.

Further study determined that there were not that many strategies. Some were friendly, some were selfish and some were rugged, most strategies were just variations and combinations of these themes.

Then it was decided to have a slightly different competition. What if all the different programs ran randomly against each other? Sometimes Tit for Tat would run against a friendly program and sometimes against a selfish program. If one of the strategies was winning in a match up within the simulation universe, the other strategy defaulted to the winning strategy.

This is getting to be a rather complex simulation. Very difficult to do in real life, but in the computer it works pretty well. When running these simulations again there was a clear winner.

Over time, after many moves, most of a simulation universe would consist of “Tit for Tatters”.

But there was a surprise: In most cases there would be a small sprinkling of selfish strategies that survived. No matter how long the simulation ran these selfish strategies persisted.

Another pause for thought. Why did those selfish strategies hang on so persistently?

---------------------------------------------

Let me switch to a real life story.

What I discovered as a young engineer: My Life and Cooperation

When I was a young engineer I was very fortunate to work for a yacht designer who was designing a boat for the 1987 America’s Cup.

My boss was a marvelous man and very quickly loosely designated me as his chief engineer. What this meant was that it was my job to evaluate all the outside research projects that were taking place in support of our campaign.

The America’s Cup is the holy grail of sailing. It is somewhat like Formula 1 car racing, but takes even more people to pull off successfully. Instead of one or two racecar drivers there are as many as 25 sailors in the team. There is a very long period of design and testing followed by training and on the water testing.

Unlike Formula 1 racing which is heavily commercial, there was a large amateur component in America’s Cup racing. The research, design and testing phase is supported by a lot of in kind donations from research institutions, corporations, universities and private individuals.

To a large extent my job was to sort the crazy offers for assistance from the useful offers for assistance.

To my amazement some really famous engineers and scientists were offering free support, and as a young man I was thrilled to sit at the feet of these demigods of technology.

At the same time there were also offers from private individuals who had no prior record in the engineering world.

I tried to be nice to everybody and made a real effort to carefully listen to everybody’s ideas. This is a rather time consuming process, and often can be somewhat humorous.

At times a person would call and claim they had an idea that was sure to win the America’s Cup. I would arrange a face-to-face meeting with the person. The person would come to our office, generally carrying a box, with some type of show and tell thing. It could be a type of cloth that is stronger, or lighter or smoother than any other cloth in the world, or a jar of paint that will make the hull more slippery, or even a black box with electrical leads on it that could predict the wind speed two miles ahead of the boat.

Some of these people could be very mysterious and it would be difficult to determine if their idea was real or not, and some ideas were plain silly. But the vast majority of them were bright well-meaning people who thought they could help, and once we took into account the rules of the game and our budget some of the ideas were gratefully incorporated into our design.

After a few private individuals, it became time for me interact with some of the heavy hitters. These people were chief structural engineers of major engineering and manufacturing firms, or deans of engineering of the best universities in the country, or chief aerodynamicists of major aerospace corporations.

Taking into account the lofty positions these people occupied, my expectations were very high and for the most part I was not disappointed. Most of these people were very very bright, and I learned more in the three years I spent on that job than I could have learned in a decade at the world’s best engineering schools.

But one thing absolutely amazed me. While most were very smart and cooperative people, a few in this group were very uncooperative, and not very bright. Actually these few people could only be described as jerks and idiots.

One particularly stunning example had been the dean of engineering of one of the top three engineering schools in the United States. This man came up with all sort of ideas, but none of them were physically possible. This is no problem in an innovative design setting since one ineffective idea could lead to another effective idea. However, this man refused to accept that his ideas violated the most basic physical principals.

As far as he was concerned, only his ideas were valid and worth research dollars and other ideas were considered to be wrong or backwards.

I tried nice, I tried not so nice, I tried having older people talk to him. Nothing worked; he was right, the rest of the world was wrong. I talked to my boss, and explained that this man was completely out of touch with reality. My boss agreed.

The dean would not listen. His ideas received funding from the syndicate, because, after all, this man had been the dean of engineering. Besides the syndicate reasoned, maybe my boss, the other topnotch engineers and I were just jealous that we had not thought of the unusual ideas that the dean had thought of.

We were not too worried, because once tested it would become clear that the dean’s ideas were pipe dreams. Testing started and sure enough the dean’s ideas did not work at all.

We thought that these tests were the end of it, but instead he claimed that his ideas were proven successful, but that all that was required to properly apply them to the boat was some work by the engineers.

We had to admit that we could not engineer it into the boat. I was young and reckless and at that stage I was itching for a fight. My boss was wiser and told me to take it easy; this too would pass.

It did pass, but not until the dean published his “brilliant” findings in the Alumni magazine. In the article he wasted no space in making competent engineers look like fools.

The article hit the Associated Press wire and was reprinted in various papers, and mentioned my boss by name as the designer for the syndicate. My boss still refrained from responding.

I wrote a detailed letter to the editor of the magazine outlining the errors and misrepresentations they had published. The letter was never published. End of story.

But not really. One thought never left me. How could such a fool become the dean of engineering of one of the very top engineering schools in the country?

Where did this guy come from? Then I started thinking about other groups I had worked with. I had worked in shipyards as an apprentice, and while, in general, it was a very nice crowd, in that crowd there also was a certain percentage of jerks and idiots. Same thing in grade school. Same thing with the members of the syndicate. Same thing with the faculty in my engineering school.

And no matter what group you were in, the percentage was always about the same; between 10 and 20 percent.

How could this be? Wouldn’t higher levels of society have fewer idiots and jerks? How come I found that on a percentage basis there were as many bright and cooperative shipfitters as captains of industry? Wouldn’t the bright shipfitters displace the dumb captains of industry? At least there should be more dumb shipfitters than captains of industry.

About a year after the America’s Cup I became the general manager of a small engineering firm. It was a nice firm of about ten people, but one person absolutely was not a team player. This guy was a continuous nuisance, always taking advantage of things first, always undermining earnest efforts by others, just an unpleasant human being. Since I was the general manager I went to the partners and suggested that the guy be fired. They agreed that the guy was nothing but a nuisance, and agreed to his dismissal.

I went home that evening believing that from this moment on life would be much better. Tomorrow I would be managing a company of nine team players and life would be great.

I was right, for about three weeks, but by that time one of the other nine had somehow become the new jerk and idiot. What had happened here?

---------------------------------------------

This brings us back to our computer simulation and what it tells about real life. In both cases it is nearly impossible to make cooperation 100% successful. No matter what you do and where you do it, a small percentage of non-cooperators will persist.

Why is this? The simulation provides an answer. As long as the number of non-cooperators is small the Tit for Tat cooperators will not be affected by them to an extent that the system becomes unstable.

At the same time these non-cooperators can take advantage of the cooperator’s initial urge to cooperate. As long as you don’t have to go back to one cooperator twice, you can do very well indeed by being selfish as long as there are not too many people playing the same game in the same group.

What does this mean? Well, it really means that no matter how hard you try, in general, not all people in a group will cooperate.

Is this bad? That depends on how you look at it. If you are a non-cooperator it is good as long as you don’t have too many people play the same game. What happens if too many people play the same game? It will become less attractive to be a non-cooperator because you start to run into more non-cooperators.

Under certain conditions those non-cooperators will start to realize that it does not pay to be non-cooperators and become cooperators, or non-cooperators are ejected from the system until the non-cooperators fall below the horizon. The reverse of this reasoning would indicate that everybody is a potential non-cooperator, which appeared to be the case in my office and once the prime non-cooperator had been eliminated the most likely potential non-cooperator took his place.

In more extreme cases the non-cooperators will not revert back to cooperation in time and basically the system will self destruct, and will have to rebuilt from a new core of cooperators.

This often happens in real life sometimes on a small scale (families) and sometimes on a large scale (countries)




Home

Last Updated: 2/16/05