Simplicity, complexity, apathy.
Algorithmic Progress in the Everyday
Hmm, I was fascinated by the post of Xystum on What Is Progress Anyway? Guess I'll walk it down a tangent, and given that Xystum's way beyond my expertise when it comes to algorithms, they may feel free to redirect or laugh as needed. You, the reader, may too.
Progress strikes me as procedural enhancement and/or expansion of a system where benefit results. In other words, both streamlining and branching of a process can benefit a system, and streamlining the interaction of processes, while increasing the overall complexity of a system, can result in faster and more efficient operations. I don't say that faster and more efficient is of necessity progress - instead, I would analyze cost and benefit across the whole procedure and seek to maximize benefit while minimizing cost.
Basic economics, basic runtime streamlining for database work, basic project management - it seems to me that all of these touch on the everyday, and can be implemented not only in a computer or pure maths environment but also in, "Wake up, make coffee, have shower, brush teeth, add deodorant." Linearly, these steps take time, and doing them one after another lengthens time to finish - which might in turn be paid off with the enjoyment of smelling the coffee brew, sipping that first cup, and then getting clean.
So, when building an algorithm, conflicting benefits need to be costed out or deemed acceptable. Tossing coffee in the brewer, brushing teeth in the shower while coffee brews, adding deodorant while drying off, drinking coffee on the way to work - endless variations are possible, and in a human system not only do the benefits of the procedure itself need to be determined, but humans get bored. Same thing every day results in lower ability to function outside known procedures, dulls the consciousness. Thus, adding burn in to the procedure and compensating for it may well decrease efficiency for the specific procedure, but increase overall efficiency of the human and have the added benefit of being fun.
That's more or less how I live my life, and similarly how I mess myself up. I am a creature of habit, and when I'm burned in, I'm very likely to miss fundamental steps. I don't have the math to determine how burn in would be calculated for a learning machine, for example, but I can envision a system where reinforcement of positive results could easily lead to optimizing for a single solution in a case A where, with the addition of variables, might still result in a path being used as optimal for desired result A which inadvertently degrades the path to result B.
Shower, brush teeth, shave. It could be a disaster if the process entry point for shave is carelessly applied, especially with an electric razor - and yes, that's a blunder I've achieved, cooking my rechargeable.
I believe that recognizing patterns and solving for fun, safety, efficiency, and minimization of harm to others can be done across a broad spectrum of environments - and should be. As a troubleshooter - my only actual official consulting title - my job is to pin down points of failure, be it web work, fitting stuff into a car, cleaning, or running safety procedures and teaching people how not to die in confined spaces. I'm not specialized at anything other than making that leap into:
Process A needs to do this. These means exist.
Process A, if modified, could be more efficient given current parameters.
Process A, either modified per B or as stands, can be unsafe if interacting with Process C.
Solving for process integration is art, pure and simple - but I'd also wager that the grasp I have on this, whatever it is that lets me look at something and just know: This is dangerous, this is a shorter path, this is a point of failure - could be achieved with a computer, or simply by training a person in what to look for.
Now: I am constantly stunned by people's inability to grasp basic critical thinking. That includes my own failures, much to my chagrin, and I've trained in this stuff until it's reflexive. I am baffled by people drinking and driving, for example, or teachers who force their students to learn using only one method of teaching, regardless of who the target audience is. I believe it's essential in seeking progress to trap errors in our ways of thinking as much as possible, and a basic understanding of algorithms could be a wonderful starting point.
For example, determining the why of OUI could be as simple as a difference in costs and benefits between the drinker and the portion of society that's negatively effected - including the drinker, for most OUI cases don't specifically think, "What if?" When taking to the road.
I would go so far as to say that nearly every conflict in society could be solved for procedurally, and that the main sticking point would be benefit. In other words, when I ask, "What's progress?" I have to ask, "What's benefit?"
So, what is benefit? The ability to do as you will but harm none seems a good place to start, as does prevention of harm. So does promotion of fun, yet even at the outset there are some issues, both with differences in religious/moral constraints and in perception of fun, that conflict.
For example, take the case of pain. Some people like it, some people find it sexy, some people are just horrified by it. Globally applying a set of values to operations involving pain is thus inefficient because no single path renders optimum result for all participants. In such situations, solving per case becomes essential - and doing as you will has to account for varying flavors of harm none.
Perception of harm becomes a problem when one person, viewing something they find awful, interferes in the activities of another who perhaps finds benefits - perhaps even no cost at all - in a given solution.
So: Add a couple of operators to the system. Consent and right to consent.
Much of this can be seen across the American spectrum of social interaction. Hetro/homosexual conflicts, religious imperatives against buggery and states of mind -Messy, but don't think for a minute that only religious and sexual conflicts get tricky. Anything from drug use and recovery to operating a motor vehicle within the bounds of law can be as ugly, and I'll state here that responsibility becomes the key.
In attempting to solve algorithmically for human interaction, one needs to solve both for direct harm - hitting someone with a car - and indirect harm, such as knowing a person is homosexual and having a problem with it. By direct, I'm meaning to imply that by a specific action, a person is causing a harm another person cannot avoid, and is thus responsible for fixing that harm. By indirect, I mean to imply cases where harm is conditional and avoidable - where, for example, did I not know a person was homosexual, or a free time drug user, I would have no problems with them.
Notice I'm using extremely charged examples here, and that in cases of indirect harm I'm going to specifically say that ANY indirect harm is by it's nature the business of the person indirectly harming, not of the person being harmed. Now, in the case of drugs, indirect harm moves to direct harm where, as a result of impairment, damage is done - but in the case of use itself as an isolated event, only indirect harm exists.
Take that a step further: If I am bothered by something that I can simply walk away from, something which doesn't directly interfere with my freedom to safely live my life, that's indirect harm. I can choose not to take damage from it, to keep going on with my life. If I instead choose to work against that indirect harm, I may myself be doing direct harm. I can think of very few people who would say, "No fun sex for you, ever." Or, "You can't be Jewish, Hindu, Christian," Isn't direct harm.
Now, I'm possibly being unnecessarily open-minded here. I know people who like to be tortured. I myself love to fight competitively. An infinity of indirect harm cases and consensual direct harm cases exist. I would say, providing the harm is indirect or consensual, there's no problem with it.
Adding responsibility to the mix - not shagging or fighting in public, where someone might make a bad assumption and be directly harmed - is something that I have to account for. In presenting this algorithmically, it seems best to allow for multi-path, multi-target processes. Additionally, since some regions of process seem likely to conflict, allowing for privacy (separation of conflicting states) seems essential.
I have to admit, I have the advantage of being bothered by almost nothing humans do for fun, providing it doesn't directly harm an unwilling party. This means that I'm not bothered by suicide beyond the burden in places on those who must clean up. Neither sex not religion - any flavor of either - seems a problem within the confines of it's reasonably closed environment. In fact, I can think of nothing so outlandish that it cannot be dealt with in terms of direct and indirect harm.
Taking modern society and optimizing our system - both as a whole and as a series of patterns and operations upon patterns - is tricky, however. Everyone wants to be right, everyone wants to be safe, everyone wants to have power. Applying a fair harm based algorithm to the entire society seems extremely likely to border on anarchy, which, if it interferes with infrastructure and commerce, results in direct harm.
I think it could be done, but I think it would have to be done in small groupings at first, and I think it would have to begin with education. Remember the bit about teaching each according to the process by which they best learn? I would propose that by standardizing our entire system, we're creating the greatest harm possible, not solving for minimization of harm. The predication of fairness in such a system is flawed, both logically and ethically. In other words, from the front to the finish, it seems to be that from the perspective of both progress and benefit, our entire society and culture is working against what it claims to be working for - and unintentionally at that.
The idea of using algorithms to achieve this excites me a great deal. The point of failure however, is intolerance and masking indirect harm as direct harm to back an argument. We're a long way from getting beyond that - and I don't think it'll happen in my lifetime. On the other hand, I don't see any reason not to try, even if it's something as simple as taking your coffee in the shower and brushing your teeth when done.
I know, I've wandered some in this - but how could I not? The complexity of life isn't a trivial thing, and I know I've avoided touching some subjects while dwelling on others. It's unavoidable. Instead - because I really do think it's worthwhile to analyze and solve for problems - I'd close with this.
Go take a look at your life and make it better. Figure out what harm you're doing and minimize it. Figure out what you've got problems with and whether those problems are direct or indirect harm - and how to avoid both if you can. Don't be shy about accepting hard topics and dealing with them, because the painful areas of life don't just go away. Do as you will and harm none almost works - it really does - but it needs a third part. Don't let someone who isn't being harmed claim they are. If you do, you might as well hand them your fear and your heart, because anything, anything at all, can be claimed an indirect harm.
Hmm, I was fascinated by the post of Xystum on What Is Progress Anyway? Guess I'll walk it down a tangent, and given that Xystum's way beyond my expertise when it comes to algorithms, they may feel free to redirect or laugh as needed. You, the reader, may too.
Progress strikes me as procedural enhancement and/or expansion of a system where benefit results. In other words, both streamlining and branching of a process can benefit a system, and streamlining the interaction of processes, while increasing the overall complexity of a system, can result in faster and more efficient operations. I don't say that faster and more efficient is of necessity progress - instead, I would analyze cost and benefit across the whole procedure and seek to maximize benefit while minimizing cost.
Basic economics, basic runtime streamlining for database work, basic project management - it seems to me that all of these touch on the everyday, and can be implemented not only in a computer or pure maths environment but also in, "Wake up, make coffee, have shower, brush teeth, add deodorant." Linearly, these steps take time, and doing them one after another lengthens time to finish - which might in turn be paid off with the enjoyment of smelling the coffee brew, sipping that first cup, and then getting clean.
So, when building an algorithm, conflicting benefits need to be costed out or deemed acceptable. Tossing coffee in the brewer, brushing teeth in the shower while coffee brews, adding deodorant while drying off, drinking coffee on the way to work - endless variations are possible, and in a human system not only do the benefits of the procedure itself need to be determined, but humans get bored. Same thing every day results in lower ability to function outside known procedures, dulls the consciousness. Thus, adding burn in to the procedure and compensating for it may well decrease efficiency for the specific procedure, but increase overall efficiency of the human and have the added benefit of being fun.
That's more or less how I live my life, and similarly how I mess myself up. I am a creature of habit, and when I'm burned in, I'm very likely to miss fundamental steps. I don't have the math to determine how burn in would be calculated for a learning machine, for example, but I can envision a system where reinforcement of positive results could easily lead to optimizing for a single solution in a case A where, with the addition of variables, might still result in a path being used as optimal for desired result A which inadvertently degrades the path to result B.
Shower, brush teeth, shave. It could be a disaster if the process entry point for shave is carelessly applied, especially with an electric razor - and yes, that's a blunder I've achieved, cooking my rechargeable.
I believe that recognizing patterns and solving for fun, safety, efficiency, and minimization of harm to others can be done across a broad spectrum of environments - and should be. As a troubleshooter - my only actual official consulting title - my job is to pin down points of failure, be it web work, fitting stuff into a car, cleaning, or running safety procedures and teaching people how not to die in confined spaces. I'm not specialized at anything other than making that leap into:
Process A needs to do this. These means exist.
Process A, if modified, could be more efficient given current parameters.
Process A, either modified per B or as stands, can be unsafe if interacting with Process C.
Solving for process integration is art, pure and simple - but I'd also wager that the grasp I have on this, whatever it is that lets me look at something and just know: This is dangerous, this is a shorter path, this is a point of failure - could be achieved with a computer, or simply by training a person in what to look for.
Now: I am constantly stunned by people's inability to grasp basic critical thinking. That includes my own failures, much to my chagrin, and I've trained in this stuff until it's reflexive. I am baffled by people drinking and driving, for example, or teachers who force their students to learn using only one method of teaching, regardless of who the target audience is. I believe it's essential in seeking progress to trap errors in our ways of thinking as much as possible, and a basic understanding of algorithms could be a wonderful starting point.
For example, determining the why of OUI could be as simple as a difference in costs and benefits between the drinker and the portion of society that's negatively effected - including the drinker, for most OUI cases don't specifically think, "What if?" When taking to the road.
I would go so far as to say that nearly every conflict in society could be solved for procedurally, and that the main sticking point would be benefit. In other words, when I ask, "What's progress?" I have to ask, "What's benefit?"
So, what is benefit? The ability to do as you will but harm none seems a good place to start, as does prevention of harm. So does promotion of fun, yet even at the outset there are some issues, both with differences in religious/moral constraints and in perception of fun, that conflict.
For example, take the case of pain. Some people like it, some people find it sexy, some people are just horrified by it. Globally applying a set of values to operations involving pain is thus inefficient because no single path renders optimum result for all participants. In such situations, solving per case becomes essential - and doing as you will has to account for varying flavors of harm none.
Perception of harm becomes a problem when one person, viewing something they find awful, interferes in the activities of another who perhaps finds benefits - perhaps even no cost at all - in a given solution.
So: Add a couple of operators to the system. Consent and right to consent.
Much of this can be seen across the American spectrum of social interaction. Hetro/homosexual conflicts, religious imperatives against buggery and states of mind -Messy, but don't think for a minute that only religious and sexual conflicts get tricky. Anything from drug use and recovery to operating a motor vehicle within the bounds of law can be as ugly, and I'll state here that responsibility becomes the key.
In attempting to solve algorithmically for human interaction, one needs to solve both for direct harm - hitting someone with a car - and indirect harm, such as knowing a person is homosexual and having a problem with it. By direct, I'm meaning to imply that by a specific action, a person is causing a harm another person cannot avoid, and is thus responsible for fixing that harm. By indirect, I mean to imply cases where harm is conditional and avoidable - where, for example, did I not know a person was homosexual, or a free time drug user, I would have no problems with them.
Notice I'm using extremely charged examples here, and that in cases of indirect harm I'm going to specifically say that ANY indirect harm is by it's nature the business of the person indirectly harming, not of the person being harmed. Now, in the case of drugs, indirect harm moves to direct harm where, as a result of impairment, damage is done - but in the case of use itself as an isolated event, only indirect harm exists.
Take that a step further: If I am bothered by something that I can simply walk away from, something which doesn't directly interfere with my freedom to safely live my life, that's indirect harm. I can choose not to take damage from it, to keep going on with my life. If I instead choose to work against that indirect harm, I may myself be doing direct harm. I can think of very few people who would say, "No fun sex for you, ever." Or, "You can't be Jewish, Hindu, Christian," Isn't direct harm.
Now, I'm possibly being unnecessarily open-minded here. I know people who like to be tortured. I myself love to fight competitively. An infinity of indirect harm cases and consensual direct harm cases exist. I would say, providing the harm is indirect or consensual, there's no problem with it.
Adding responsibility to the mix - not shagging or fighting in public, where someone might make a bad assumption and be directly harmed - is something that I have to account for. In presenting this algorithmically, it seems best to allow for multi-path, multi-target processes. Additionally, since some regions of process seem likely to conflict, allowing for privacy (separation of conflicting states) seems essential.
I have to admit, I have the advantage of being bothered by almost nothing humans do for fun, providing it doesn't directly harm an unwilling party. This means that I'm not bothered by suicide beyond the burden in places on those who must clean up. Neither sex not religion - any flavor of either - seems a problem within the confines of it's reasonably closed environment. In fact, I can think of nothing so outlandish that it cannot be dealt with in terms of direct and indirect harm.
Taking modern society and optimizing our system - both as a whole and as a series of patterns and operations upon patterns - is tricky, however. Everyone wants to be right, everyone wants to be safe, everyone wants to have power. Applying a fair harm based algorithm to the entire society seems extremely likely to border on anarchy, which, if it interferes with infrastructure and commerce, results in direct harm.
I think it could be done, but I think it would have to be done in small groupings at first, and I think it would have to begin with education. Remember the bit about teaching each according to the process by which they best learn? I would propose that by standardizing our entire system, we're creating the greatest harm possible, not solving for minimization of harm. The predication of fairness in such a system is flawed, both logically and ethically. In other words, from the front to the finish, it seems to be that from the perspective of both progress and benefit, our entire society and culture is working against what it claims to be working for - and unintentionally at that.
The idea of using algorithms to achieve this excites me a great deal. The point of failure however, is intolerance and masking indirect harm as direct harm to back an argument. We're a long way from getting beyond that - and I don't think it'll happen in my lifetime. On the other hand, I don't see any reason not to try, even if it's something as simple as taking your coffee in the shower and brushing your teeth when done.
I know, I've wandered some in this - but how could I not? The complexity of life isn't a trivial thing, and I know I've avoided touching some subjects while dwelling on others. It's unavoidable. Instead - because I really do think it's worthwhile to analyze and solve for problems - I'd close with this.
Go take a look at your life and make it better. Figure out what harm you're doing and minimize it. Figure out what you've got problems with and whether those problems are direct or indirect harm - and how to avoid both if you can. Don't be shy about accepting hard topics and dealing with them, because the painful areas of life don't just go away. Do as you will and harm none almost works - it really does - but it needs a third part. Don't let someone who isn't being harmed claim they are. If you do, you might as well hand them your fear and your heart, because anything, anything at all, can be claimed an indirect harm.