I am afraid of bad drivers. When I am in the car of a bad driver I always think he could do something stupid. The only good thing about bad drivers is that they think the rest drive like them and are therefore careful.
I am afraid of good drivers too. They think the rest drive like them, therefore they are not too careful. A good driver tends not to keep a safe distance from the vehicle in front and is not troubled if the driver of the vehicle behind does the same. This is based on the wrong assumption that should something happen they (the good one and the one behind him) will be able to stop in a moment.
I am also afraid of some good thinkers, who think that all people think like them. Many of those thinkers specialize in game theory, so let me take for example a popular story with a game-theoretic flavor.
There is a faraway island inhabited by 10 people, 2 of whom are blue-eyed and 8 of whom brown-eyed. There is a taboo so they do not use mirrors and do not discuss eye colors. Any islander who by chance discovers his true eye color must commit suicide the same day.
One day an old sailor arrives on the island and oblivious to the local custom greets the islanders by saying how glad he is to see another blue-eyed person like him for the first time in years. What happens then?
Based on those assumptions only, many smart game theorists imagine the following. Nothing happens on the first day as any islander knows there is another one who is blue-eyed. On the second day one of the blue-eyed islanders (let us call him A) is surprised that the only blue-eyed islander he knows (B) did not commit suicide on the first day. To A this says there is another blue-eyed one, whose existence prevents B from realizing he (B) is the only blue-eyed one and kill himself. As A knows that all other islanders are not blue-eyed, he concludes he himself is the blue-eyed one and commits suicide. B thinks the same way and kills himself too. On the next day the rest realize they are brown-eyed (as all blue-eyed ones committed suicide on the second day) and commit mass suicide, leaving the island to the sailor.
I claim that the above assumptions are not enough to justify this particular end of the story. There are other (unstated or even unrealized) assumptions:
1. All islanders are as smart as game theorists. I claim that there is evolutionary explanation why people are not that smart, i.e. why game theorists are not that many. If islander A were smart but B (and all others) were not, then A would have reached the right conclusion but B (and the others) would have not. Therefore, A would have killed himself and B (and all others) would have not. Ask yourself whose grandchildren will inhabit the island.
2. All islanders are as honest as game theorists. I claim that there is evolutionary explanation why people are not that honest, i.e. why game theorists are not that many. If islander A were honest but B (and all others) were not then all would have reached the right conclusion but only A would have killed himself. Ask yourself whose grandchildren will inhabit the island.
3. All islanders are traditionalists. I claim that there is evolutionary explanation why people are not such traditionalists and break with tradition whenever possible. What if A realizes there are better ways to risk his life than commit suicide and commits a successful coup d’etat, resulting in abandoning the stupid taboo among other things? Ask yourself whose children will live in the ruler’s palace?
4. The sailor speaks the language of the islanders (or they speak his). This is a common assumption game theorists, logicians, politicians and businesspeople make. In the real world the islanders would have eaten the sailor long before he had learnt their language. If the sailor was American or English, they do not even need to eat him. He will never speak their language as he will expect they should learn his instead. That is how English became the language of the world – English speakers travelled more than anyone else and never cared to learn foreign languages.
5. The islanders and the sailor share common color perception. Had the sailor been color-blind and called blue-eyed those 2 islanders who in fact were green-eyed, all islanders would not have believed him and would have stayed alive (having him punished for lying). Had they (instead of the sailor) been color-blind, they would have done the same, as they would not have known they are color-blind, i.e. they will have perceived themselves as normal and him as color-blind.
Conclusion
What if the sailor was (or was not) color-blind and some of the blue-eyed and brown-eyed islanders were (or were not): color blind and/or honest and/or smart and/or English speakers? We are entering the Garden of Forking Games. Let us stop and ask the following questions.
Can game theorists control their assumptions or is it that they just do not care? As long as a work based on assumptions like those (that all people are smart, honest and English speaking and are not color-blind or just blind) can earn one a living or even a Nobel Prize in Economics, why care if those assumptions are wrong or insufficient?
I am afraid of good drivers too. They think the rest drive like them, therefore they are not too careful. A good driver tends not to keep a safe distance from the vehicle in front and is not troubled if the driver of the vehicle behind does the same. This is based on the wrong assumption that should something happen they (the good one and the one behind him) will be able to stop in a moment.
I am also afraid of some good thinkers, who think that all people think like them. Many of those thinkers specialize in game theory, so let me take for example a popular story with a game-theoretic flavor.
There is a faraway island inhabited by 10 people, 2 of whom are blue-eyed and 8 of whom brown-eyed. There is a taboo so they do not use mirrors and do not discuss eye colors. Any islander who by chance discovers his true eye color must commit suicide the same day.
One day an old sailor arrives on the island and oblivious to the local custom greets the islanders by saying how glad he is to see another blue-eyed person like him for the first time in years. What happens then?
Based on those assumptions only, many smart game theorists imagine the following. Nothing happens on the first day as any islander knows there is another one who is blue-eyed. On the second day one of the blue-eyed islanders (let us call him A) is surprised that the only blue-eyed islander he knows (B) did not commit suicide on the first day. To A this says there is another blue-eyed one, whose existence prevents B from realizing he (B) is the only blue-eyed one and kill himself. As A knows that all other islanders are not blue-eyed, he concludes he himself is the blue-eyed one and commits suicide. B thinks the same way and kills himself too. On the next day the rest realize they are brown-eyed (as all blue-eyed ones committed suicide on the second day) and commit mass suicide, leaving the island to the sailor.
I claim that the above assumptions are not enough to justify this particular end of the story. There are other (unstated or even unrealized) assumptions:
1. All islanders are as smart as game theorists. I claim that there is evolutionary explanation why people are not that smart, i.e. why game theorists are not that many. If islander A were smart but B (and all others) were not, then A would have reached the right conclusion but B (and the others) would have not. Therefore, A would have killed himself and B (and all others) would have not. Ask yourself whose grandchildren will inhabit the island.
2. All islanders are as honest as game theorists. I claim that there is evolutionary explanation why people are not that honest, i.e. why game theorists are not that many. If islander A were honest but B (and all others) were not then all would have reached the right conclusion but only A would have killed himself. Ask yourself whose grandchildren will inhabit the island.
3. All islanders are traditionalists. I claim that there is evolutionary explanation why people are not such traditionalists and break with tradition whenever possible. What if A realizes there are better ways to risk his life than commit suicide and commits a successful coup d’etat, resulting in abandoning the stupid taboo among other things? Ask yourself whose children will live in the ruler’s palace?
4. The sailor speaks the language of the islanders (or they speak his). This is a common assumption game theorists, logicians, politicians and businesspeople make. In the real world the islanders would have eaten the sailor long before he had learnt their language. If the sailor was American or English, they do not even need to eat him. He will never speak their language as he will expect they should learn his instead. That is how English became the language of the world – English speakers travelled more than anyone else and never cared to learn foreign languages.
5. The islanders and the sailor share common color perception. Had the sailor been color-blind and called blue-eyed those 2 islanders who in fact were green-eyed, all islanders would not have believed him and would have stayed alive (having him punished for lying). Had they (instead of the sailor) been color-blind, they would have done the same, as they would not have known they are color-blind, i.e. they will have perceived themselves as normal and him as color-blind.
Conclusion
What if the sailor was (or was not) color-blind and some of the blue-eyed and brown-eyed islanders were (or were not): color blind and/or honest and/or smart and/or English speakers? We are entering the Garden of Forking Games. Let us stop and ask the following questions.
Can game theorists control their assumptions or is it that they just do not care? As long as a work based on assumptions like those (that all people are smart, honest and English speaking and are not color-blind or just blind) can earn one a living or even a Nobel Prize in Economics, why care if those assumptions are wrong or insufficient?
RSS Feed
