Introductions, as common as they are, are mostly a waste of time. Instead of introducing, let's explain a difficult subject, and connect the dots to figure out what this blog is all about.
Ask yourself this question: have you ever been in a social situation where there was not an explicit or implicit hierarchy? At work you have titles, at home you have seniority, and with your friends you have an alpha, beta, etc. Not only does a situation have a hierarchy, but the same two people can have two different rankings, depending on the situation.
Tim and Huang are on the football team and the math team at school. On the field, Huang is the first string quarter back and the team captain, so Tim defers to him there. On the math team, however, Tim is the captain and tutors many of his teammates, including Huang, to compete for scholarships at math competitions.
So, Tim and Huang's behavior to one another changes depending on the environment, due in large part to a difference in rank. So, who has the higher social status? You might say football is regarded more highly than mathematics, so Huang is the answer, but what if Tim is better looking and has a powerful father? What if Huang is also considered very popular, funny, and kind?
The more activities in which each person is highly ranked, the more social status they have, but this seems to be very inexact. Social status seems to be up to personal preference, which is not necessarily wrong, but given how important status is in society, this seems too trivial.
Imagine overhearing a conversation between A and B, and you know nothing about them that could help you determine rank. In the conversation, A is nice to B, but B is not nice to A. Instead of retaliating, A continues to be nice, and B continues to be negative. Who has the higher status?
One thing that makes understanding behavior difficult is that we think on the object level, not on the data level. A peach is a peach, not a sensory input signature. Tim and Huang are their rankings, not loosely associated sensory input values that we used to determine rank. Let's think on the data level, not on the object level, to get our heads around why we know B has a higher rank than A.
I stated that A was nice to be, but B was mean or negative to A, but I didn't have to explain what nice, mean, or negative are. We interpret sensory data as negative or positive feedback without really thinking about it. That data can be social interaction, compliments or insults, winning or losing money, professional criticism or praise. After we measure the feedback value of data, we use a feedback differential between people to determine rank.
In the conversation we overheard, B was receiving a net gain in feedback, because A was giving positive feedback while receiving negative. B was giving away negative and receiving positive. If A's positive feedback was worth 1, and B's negative feedback was worth -1, A is left with -1 + -1, and B is left with -(-1) + 1. This implies that, in their relationship, there is a feedback differential, and the one coming out ahead is ranked higher.
This gives us a way to rank social status. Measuring total feedback that everyone gets away with, in every hierarchy they inhabit, gives us a total feedback differential we can use to measure a social status. Though still somewhat fuzzy, because how do we equate feedback on a math team and on a football team, we have made progress in our attempt at understanding.
So, where do we go from here? We go backwards, and ask why is there a feedback differential, and why is everything ranked.
What work has to do with rankings is explained by answering why Google search results ranked. Why can't the list of links provided by a query put into a search engine be of arbitrary order? Assuming that we are solving for our benefit by searching, it's important to read and interpret each link. That takes work. The farther away a link is from the first link, the more work it takes to get there.
Similarly, the way we learn behavior is by copying the behavior of someone else. The copy may be imperfect by accident, or it may be deliberately optimized, but none the less learning behavior takes work. If we can interpret solving for our benefit as increasing our ranking, we should naturally copy the behavior of the highest ranked individual we have access to.
So, we rank one another based on a work gradient, where copying behavior from the highest ranked person helps us to increase our own rank, but why is there a feedback differential exactly? Why do we know B is ranked higher than A because of meannness?
Remember that at the heart of why we rank people is our drive to solve for our benefit, and we do so by copying behavior of those we think are highly ranked. Meaning, the benefit of highly ranked people is they teach us how to improve our own rank.
To know what our own rank is then, we have to know how much exactly others benefit from us. What information would we have to put out to get the information of our own ranking? What information can we trust to get back? Can we believe people's compliments? In solving for their own benefit, other people are capable of deception in order to benefit from us.
If other people have to solve for positive feedback, then they wouldn't tolerate any negative feedback you give to them, unless they benefited from you such that there was still a net gain for them. In other words, that people accept negative feedback from you is trustable data which lets you know they think they benefit from you.
The relationship between A and B then is made more clear to us: A takes some abuse from B because A benefits from B, and now we have a sort of reason why we think B is higher ranked. Sort of.
There's problems with our model. Because we are trying to explain behavior, we can look at observed behavior for support. Does everyone who benefits those around them dish out abuse? Do mean individuals benefit society? Are we really always fighting for a higher rank, for benefit? How do we explain most people not being hugely ambitious?
We talked about the work of solving for our benefit, and that we cannot escape the work of getting that benefit. Furthermore, we rank people according to who benefits us the most, and we know who benefits from us because they accept negative feedback from us. By observing feedback gradients, we can measure who gets to dish out the most negative feedback, and assign that person a high status. We do this to know who whose behavior we should copy so that we can improve our own status.
Our model breaks down, however, when we try to use this to predict whose behavior we should copy. Copying the behavior of the meanest person you know is not likely to have a positive result for you. So, what exactly is missing? We've been talking about work, but let's get specific on what work is in this case.
Alex Honnold is a famous rock climber who makes his work difficult by not using equipment. TJ Hollowachuck is a famous programmer whose output ranks among the highest in the world and is used to build many of the web apps we use every day. What do these two have in common? Almost no one in the world in their respective fields can match what these two have accomplished. Why?
If climbing the side of a cliff can be described as a series of specific actions in a specific order, then the work of climbing can be described as searching for that particular sequence of actions out of the total number of possible sequences. We can describe this as finding a needle in a haystack. Given that some climbers use safety equipment, what this does is effectively increase the number of needles, or sequences, they can use to successfully achieve a climb. Meaning, their search is less difficult and takes less computation. Their job being easier, there's more of them, and each one stands out less from the crowd of climbers than Alex.
TJ Hollowaychuck, author of the Node Javascript library and the Express framework, outputs lines of code in a specific order that achieves a specific technical task. Figuring out the code and the order to achieve that task can also be described as finding a needle in a haystack. Many people have written many popular libraries and frameworks, but TJ found his needle very quickly compared to them. The task could be done by many people, if given enough time, but doing it quickly made it harder, and so he stands out from the crowd, just like Alex.
The search for the needle in the haystack is a good general description of the work it takes to get a benefit. Let's try to apply it to more behavior to make a better model.
If we are always solving for our benefit, then we can say we do the work of creating and maintaining relationships because they benefit us. In choosing to do that work, what behavior do we look for that helps us to decide who to have a relationship with? If solving for benefit means improving social status, then we can say we keep people around when they treat us as if we have high social status.
If you show up to a job interview wearing clothes that are comfortable, not formal, and slouch in the chair in the most comfortable position, and greet your potential employer in the same way you greet your friends, casually, what outcome can you reasonably expect compared to someone who does a lot of work to act politely?
An employer can make the assumption that, if this person doesn't do the work of being polite, then why would they do the work I assign them in the job?
Polite behavior can be expressed as behavior that is more difficult than impolite behavior. If you want to pay someone a compliment, saying something that sounds nice but is not crafted for that person can come off as insincere. Inversely, figuring out what positive traits a person sees in themselves, and using that information to make someone feel good can have a positive outcome. But remember, we had to figure that out. Saying the right words in the right order requires us to search for, you guessed it, a needle in a haystack.
So, what behavior do we look for in people to make the work of worth while? We look for proof of work they do on our behalf.
Given that maintaining a relationship is work, and we look for proof of work someone else does on our behalf in order to make the work worth while, we can say that there is a work exchange happening in all of our relationships. Thinking back to our original interpretation where we said a hierarchy was simply a feedback differential gradient, our model now seems so complicated.
An exchange of work sounds simple enough, and makes sense. You do something for me, I do something for you. But why does it happen? What makes a work exchange possible. Why is there sometimes no work exchange?
A fruit tree in a forest does work: it takes things from the ground, and makes a fruit, which contains the seed that makes the next tree. However, it cannot make the seed travel away from it, where child won't compete with parent.
A forest elephant spends the energy it gets from fruit to find more fruit. It can't make fruit, but it has mobility. After it eats the fruit, it passes the seed and drops it, along with some fertilizer, far away from the parent.
How did the tree get the elephant to do the work of spreading its seed? By doing the work of making the flesh of the fruit. From the elephant's perspective, its seed is the work the tree did to make the fruit, and its fruit was the work it did to spread the tree's seed. We, though neither tree nore elephant, also get others to do work for us by wrapping our seed in flesh, and the exchange of work can be described as a fruit.
Unlike the elephant and the tree, which cannot decide to not be their species, we can learn to do a variety of work. Meaning, we can escape competition by diversifying our output such that, as input to others, it increases their output. An elephant can travel to distant lands to escape competition from its peers, which also escapes competition for the tree. If an elephant cannot change the work it does, then can he ever increase the size of his seed, or the work others do on its behalf?
If the work others do on our behalf is our seed, and the work we do on behalf of others is the flesh we wrap our seed in to make a fruit, then we can describe social status as a ranking according to seed size. There is of course the size of the fruit, we can ignore this to reduce the complexity of the problem. Getting others to accept a small amount of flesh for a large sized seed is work in itself.
So, going from simple, to complex, and back to simple, we have a found a way to measure what we call social status that is really hard to argue with. In order to get there, we described feedback as general, meaning it can be money or praise, and we are beginning to prove the reasons that human society works, or how we escape competitive relationships. We can call this the emergence of complexity, and this is a subject which will be at the core of future posts on this blog.
[go into three branches: how evangelicals get rich {fruit analogy, human instruction sets}; how do we explain complexity of behavior and form; how currency works {why we trust data, why risk is a form of currency, how intelligent social engineering can work like currency if enough work backs it up}]
Ask yourself this question: have you ever been in a social situation where there was not an explicit or implicit hierarchy? At work you have titles, at home you have seniority, and with your friends you have an alpha, beta, etc. Not only does a situation have a hierarchy, but the same two people can have two different rankings, depending on the situation.
Tim and Huang are on the football team and the math team at school. On the field, Huang is the first string quarter back and the team captain, so Tim defers to him there. On the math team, however, Tim is the captain and tutors many of his teammates, including Huang, to compete for scholarships at math competitions.
So, Tim and Huang's behavior to one another changes depending on the environment, due in large part to a difference in rank. So, who has the higher social status? You might say football is regarded more highly than mathematics, so Huang is the answer, but what if Tim is better looking and has a powerful father? What if Huang is also considered very popular, funny, and kind?
The more activities in which each person is highly ranked, the more social status they have, but this seems to be very inexact. Social status seems to be up to personal preference, which is not necessarily wrong, but given how important status is in society, this seems too trivial.
How we see rank
The question we haven't answered yet is, if Tim and Huang didn't have explicit "captain" statuses, how would we know their ranks in the first place?Imagine overhearing a conversation between A and B, and you know nothing about them that could help you determine rank. In the conversation, A is nice to B, but B is not nice to A. Instead of retaliating, A continues to be nice, and B continues to be negative. Who has the higher status?
One thing that makes understanding behavior difficult is that we think on the object level, not on the data level. A peach is a peach, not a sensory input signature. Tim and Huang are their rankings, not loosely associated sensory input values that we used to determine rank. Let's think on the data level, not on the object level, to get our heads around why we know B has a higher rank than A.
I stated that A was nice to be, but B was mean or negative to A, but I didn't have to explain what nice, mean, or negative are. We interpret sensory data as negative or positive feedback without really thinking about it. That data can be social interaction, compliments or insults, winning or losing money, professional criticism or praise. After we measure the feedback value of data, we use a feedback differential between people to determine rank.
In the conversation we overheard, B was receiving a net gain in feedback, because A was giving positive feedback while receiving negative. B was giving away negative and receiving positive. If A's positive feedback was worth 1, and B's negative feedback was worth -1, A is left with -1 + -1, and B is left with -(-1) + 1. This implies that, in their relationship, there is a feedback differential, and the one coming out ahead is ranked higher.
This gives us a way to rank social status. Measuring total feedback that everyone gets away with, in every hierarchy they inhabit, gives us a total feedback differential we can use to measure a social status. Though still somewhat fuzzy, because how do we equate feedback on a math team and on a football team, we have made progress in our attempt at understanding.
So, where do we go from here? We go backwards, and ask why is there a feedback differential, and why is everything ranked.
Everything is ranked because the work of solving for our benefit can never be escaped
There is one presumption to this explanation, and that is that we are always solving for our benefit. I can't prove it, but I will say that any branch on the human evolutionary tree that is very poor at solving for a benefit is likely terminal. More than that, benefit is never free. It always takes work. We can support this by reminding ourselves that benefit is something we compete for, and if something proves to be very beneficial, getting it is hard, or in other words will take a lot of work.What work has to do with rankings is explained by answering why Google search results ranked. Why can't the list of links provided by a query put into a search engine be of arbitrary order? Assuming that we are solving for our benefit by searching, it's important to read and interpret each link. That takes work. The farther away a link is from the first link, the more work it takes to get there.
Similarly, the way we learn behavior is by copying the behavior of someone else. The copy may be imperfect by accident, or it may be deliberately optimized, but none the less learning behavior takes work. If we can interpret solving for our benefit as increasing our ranking, we should naturally copy the behavior of the highest ranked individual we have access to.
So, we rank one another based on a work gradient, where copying behavior from the highest ranked person helps us to increase our own rank, but why is there a feedback differential exactly? Why do we know B is ranked higher than A because of meannness?
Feedback differential as a consequence of information gain
If we begin describing behavior as a consequence of improving our rank, then we have to know what our current rank is to see if our behavior resulted in a change, and then use this as feedback for our behavior model.Remember that at the heart of why we rank people is our drive to solve for our benefit, and we do so by copying behavior of those we think are highly ranked. Meaning, the benefit of highly ranked people is they teach us how to improve our own rank.
To know what our own rank is then, we have to know how much exactly others benefit from us. What information would we have to put out to get the information of our own ranking? What information can we trust to get back? Can we believe people's compliments? In solving for their own benefit, other people are capable of deception in order to benefit from us.
If other people have to solve for positive feedback, then they wouldn't tolerate any negative feedback you give to them, unless they benefited from you such that there was still a net gain for them. In other words, that people accept negative feedback from you is trustable data which lets you know they think they benefit from you.
The relationship between A and B then is made more clear to us: A takes some abuse from B because A benefits from B, and now we have a sort of reason why we think B is higher ranked. Sort of.
There's problems with our model. Because we are trying to explain behavior, we can look at observed behavior for support. Does everyone who benefits those around them dish out abuse? Do mean individuals benefit society? Are we really always fighting for a higher rank, for benefit? How do we explain most people not being hugely ambitious?
How expensive is work really
We talked about the work of solving for our benefit, and that we cannot escape the work of getting that benefit. Furthermore, we rank people according to who benefits us the most, and we know who benefits from us because they accept negative feedback from us. By observing feedback gradients, we can measure who gets to dish out the most negative feedback, and assign that person a high status. We do this to know who whose behavior we should copy so that we can improve our own status.
Our model breaks down, however, when we try to use this to predict whose behavior we should copy. Copying the behavior of the meanest person you know is not likely to have a positive result for you. So, what exactly is missing? We've been talking about work, but let's get specific on what work is in this case.
Alex Honnold is a famous rock climber who makes his work difficult by not using equipment. TJ Hollowachuck is a famous programmer whose output ranks among the highest in the world and is used to build many of the web apps we use every day. What do these two have in common? Almost no one in the world in their respective fields can match what these two have accomplished. Why?
If climbing the side of a cliff can be described as a series of specific actions in a specific order, then the work of climbing can be described as searching for that particular sequence of actions out of the total number of possible sequences. We can describe this as finding a needle in a haystack. Given that some climbers use safety equipment, what this does is effectively increase the number of needles, or sequences, they can use to successfully achieve a climb. Meaning, their search is less difficult and takes less computation. Their job being easier, there's more of them, and each one stands out less from the crowd of climbers than Alex.
TJ Hollowaychuck, author of the Node Javascript library and the Express framework, outputs lines of code in a specific order that achieves a specific technical task. Figuring out the code and the order to achieve that task can also be described as finding a needle in a haystack. Many people have written many popular libraries and frameworks, but TJ found his needle very quickly compared to them. The task could be done by many people, if given enough time, but doing it quickly made it harder, and so he stands out from the crowd, just like Alex.
The search for the needle in the haystack is a good general description of the work it takes to get a benefit. Let's try to apply it to more behavior to make a better model.
If we are always solving for our benefit, then we can say we do the work of creating and maintaining relationships because they benefit us. In choosing to do that work, what behavior do we look for that helps us to decide who to have a relationship with? If solving for benefit means improving social status, then we can say we keep people around when they treat us as if we have high social status.
If you show up to a job interview wearing clothes that are comfortable, not formal, and slouch in the chair in the most comfortable position, and greet your potential employer in the same way you greet your friends, casually, what outcome can you reasonably expect compared to someone who does a lot of work to act politely?
An employer can make the assumption that, if this person doesn't do the work of being polite, then why would they do the work I assign them in the job?
Polite behavior can be expressed as behavior that is more difficult than impolite behavior. If you want to pay someone a compliment, saying something that sounds nice but is not crafted for that person can come off as insincere. Inversely, figuring out what positive traits a person sees in themselves, and using that information to make someone feel good can have a positive outcome. But remember, we had to figure that out. Saying the right words in the right order requires us to search for, you guessed it, a needle in a haystack.
So, what behavior do we look for in people to make the work of worth while? We look for proof of work they do on our behalf.
The flesh of the fruit of our labor
Given that maintaining a relationship is work, and we look for proof of work someone else does on our behalf in order to make the work worth while, we can say that there is a work exchange happening in all of our relationships. Thinking back to our original interpretation where we said a hierarchy was simply a feedback differential gradient, our model now seems so complicated.
An exchange of work sounds simple enough, and makes sense. You do something for me, I do something for you. But why does it happen? What makes a work exchange possible. Why is there sometimes no work exchange?
A fruit tree in a forest does work: it takes things from the ground, and makes a fruit, which contains the seed that makes the next tree. However, it cannot make the seed travel away from it, where child won't compete with parent.
A forest elephant spends the energy it gets from fruit to find more fruit. It can't make fruit, but it has mobility. After it eats the fruit, it passes the seed and drops it, along with some fertilizer, far away from the parent.
How did the tree get the elephant to do the work of spreading its seed? By doing the work of making the flesh of the fruit. From the elephant's perspective, its seed is the work the tree did to make the fruit, and its fruit was the work it did to spread the tree's seed. We, though neither tree nore elephant, also get others to do work for us by wrapping our seed in flesh, and the exchange of work can be described as a fruit.
Unlike the elephant and the tree, which cannot decide to not be their species, we can learn to do a variety of work. Meaning, we can escape competition by diversifying our output such that, as input to others, it increases their output. An elephant can travel to distant lands to escape competition from its peers, which also escapes competition for the tree. If an elephant cannot change the work it does, then can he ever increase the size of his seed, or the work others do on its behalf?
If the work others do on our behalf is our seed, and the work we do on behalf of others is the flesh we wrap our seed in to make a fruit, then we can describe social status as a ranking according to seed size. There is of course the size of the fruit, we can ignore this to reduce the complexity of the problem. Getting others to accept a small amount of flesh for a large sized seed is work in itself.
So, going from simple, to complex, and back to simple, we have a found a way to measure what we call social status that is really hard to argue with. In order to get there, we described feedback as general, meaning it can be money or praise, and we are beginning to prove the reasons that human society works, or how we escape competitive relationships. We can call this the emergence of complexity, and this is a subject which will be at the core of future posts on this blog.
[go into three branches: how evangelicals get rich {fruit analogy, human instruction sets}; how do we explain complexity of behavior and form; how currency works {why we trust data, why risk is a form of currency, how intelligent social engineering can work like currency if enough work backs it up}]
Comments
Post a Comment