|
Yiggy posted:I was getting a serious, "Imagine four balls at the edge of a cliff.." vibe reading that. In all seriousness though, we're eventually going to have to figure out how to deal with a post-scarcity world, as more unskilled jobs become automated and jobs become highly specialized. I'm thinking that's going to look more like Star Trek socialism than any sort of attention-based economy.
|
#
¿
Jan 26, 2012 01:25
|
|
|
|
| # ¿ Apr 26, 2024 06:49 |
|
|
Eripsa posted:Not that I disagree with, but that is incompatible with solving the coordination problem. That money has not solved it does not mean that it is not capable of doing so. Whether it is likely to do so is another debate altogether. I really don't think the issue of poverty is a sheer function of a system, but rather a consequence of the fact that humans run these systems. I mean, assuming that you have benevolent robots running your theoretical system, OK, but that means that its completely impractical right now. Also, you're trying to objectively measure subjective qualities.
|
|
#
¿
Feb 1, 2012 06:00
|
|
|
Eripsa posted:The whole point of my system is to crowdsource those decisions, so you don't concentrate power into the hands of a few people. My system is an anarchist system, so there is no monopoly on authority for petty, malicious weasels like yourself. Except that crowds don't somehow arrive at better or more fair decisions simply due to the nature of them being made by a crowd. Let's take a look at any of the monthly Reddit witch hunts, for example, or decisions that its OK to force kids to pray in school because we're all good Christians in this community. Fact of the matter is that a lot of people are unable or unwilling to think critically, and the remainder tend to be more concerned with their own lives than fairly handing out judgements on toothpaste. The very nature of the system is that you're going to get busybodies who think its super important that you have the correct amount of toothpaste. See homeowners associations for an example of this. You've basically just handwaved the problem away as, "Well it will be crowd-sourced so everyone has a say and the correct decision will be reached." That's not how it works in practice. Large groups can make terrible decisions too, and in some cases make *worse* decisions than a small elite, informed group. In other words, evilweasel's concerns are completely justified, and I'll have to echo those concerns. Even more so because you apparently haven't given any real thought to the actual pragmatic results of this system, but rather prefer to stay in idealism-land.
|
|
#
¿
Feb 1, 2012 18:44
|
|
|
Eripsa posted:My idea is that if you distribute the authority among the crowd, and if you keep the system inclusive and open access, then it won't be as easy as you suggest to take over the Toothpaste Tribunal or whatever. That assumes that enough people care about the Toothpaste Tribunal to prevent power from being concentrated. It also assumes that consensus arrives at the correct decision. Consensus does not necessarily arrive at the correct decision, and now you're subject to the tyranny of the majority. And you're giving this group power over basically every facet of my life. Eripsa posted:I mean this is just a fact about networks. These sorts of questions suggest a fundamental misunderstanding of the digital paradigm, and do not indicate a problem with my system. And there are serious arguments concerning Wikipedia about crowd-sourcing overriding the input of experts with incorrect conclusions. And tyranny of the majority locking down pages. Wikipedia created admins for a reason. The internet is "built" by the population at large, but most of the technical decisions and software infrastructure was built by a subset of the greater population that use the internet. An elite. Your average person, given power to make decisions on how to run the internet, would invariably make terrible decisions on how to do so because they do not understand it to a good enough degree. Eripsa posted:Nothing in my system prevents people from doing lovely things. Nothing in my system prevents people from being people. But I don't accept the implication from "people do lovely things" to "we need to control their behavior by incentivizing their labor with money". Money doesn't necessarily result in people top-down controlling others' behavior. Given the money I have, I have great leeway in how I want to live my life. Under your system, my behavior is micro-analyzed by your twitter councils. I am far less free in your system than I am under the current system. You've replaced an oppressive capitalist system with a more oppressive democratic system. Eripsa posted:I've been describing the mechanisms whereby a system can self-organize to achieve those ends, instead of having that organization planned in advanced. But I can't tell from your posts whether you just don't understand what it means to be self organized, or if you think that people are incapable of self-organization, or if you think a self-organized system will always devolve into witch hunts, or if you think self-organization will never solve the coordination problem. The point of free markets is that they aren't organized in advance. Unless I missed some part of the discussion where we're comparing your hypothetical system to some other hypothetical centrally-planned system. Eripsa posted:Do you think it is at all possible that a perfectly open system would not immediately devolve into petty witch hunts and fetishization? If you think it is inevitable, then I'm guessing you think that any attempt at self-organization will fail. No I don't think that this system will devolve into witch hunts in all cases necessarily. I do think that it will empower people seeking to start witch hunts and I think it will empower busy-bodies far more than the system we currently have. This is my experience based on these sorts of community-organized systems that grant far-reaching powers into ones' personal life. Eripsa posted:That suggests to me that your view is really that people NEED TO BE TOLD WHAT TO DO and NEED TO BE THREATENED WITH THEIR LIVES to motivate them to follow their orders, and that if they don't follow those orders that they SHOULD NOT BE ALLOWED TO LIVE. That's an absurd and insulting strawman of evilweasel's position. Shame on you. Eripsa posted:I've been very clear why I think this is false, and you have not given any arguments against my proposal here except to say it is loony. I don't know how else I am expected to respond. I don't see any problems with their argument. On the other hand, I do have a problem with your insistence in handwaving concerns away by assuming that micro-managing peoples' lives based on community-organized consensus will somehow result in fewer problems than the alternatives.
|
|
#
¿
Feb 1, 2012 19:06
|
|
|
Let me put it another way. Assuming that your idea is fool-proof and correct, you can't even drive a consensus of random goons on a forum to agree that your idea is correct. Why would you ever believe that this could scale to a level required to manage an economy? It has failed on its most basic test.
|
|
#
¿
Feb 1, 2012 19:14
|
|
|
The Duke of Ben posted:Actually, I think that this brings us to an actual thing to discuss about alternate economies. It's my belief that in order to get them to work, we would need to intentionally use propoganda on ourselves and especially our children in order to change the fundamental ways in which we view the world. More relevant I think is that we are propagandized to think that capitalist or social democratic capitalist systems are the best possible systems. So to some degree I tend to agree with you, as that seems to be how it works today. The alternative is to allow people to self-select into the systems they think will work, which fundamentally requires far more open borders and far more variety in economic systems than we have today. Of course, the caveat with the propaganda not being necessarily true is that its not necessarily false. The issue here is that in order to test new economic systems you run the risk of dramatically lowering peoples quality of life and introducing oppressive political systems (since one tends to follow the other in real life).
|
|
#
¿
Feb 1, 2012 19:29
|
|
|
Eripsa posted:I don't think it needs to be centralized, I don't think anyone can or should have the power to decide on their own. I think you give the people all the information they need, and you have some coordinating principle that allows them to search through that information as they see fit such that the best ideas for a solution rise to the top, and then you empower the people to act on those best solutions. All without centralized coordination or executive decisions. How do you handle any decisions that need to get made that require expertise? You are making the extremely bad assumption that all decisions can be properly made by the collective assuming the proper information is given. This is not the case. For example, a crowd of people given access to Web MD may give the wrong diagnosis whereas a doctor would give the correct one. Another example: software engineers and system administrators make better decisions on how to run network infrastructure than average people. Information often takes expertise to properly analyze. Actual real-world example: Reddit had severe issues with running on AWS. The average redditor's response was that AWS was bad. The real response from technically literate people with domain knowledge of the problem was that the particular service they were using, EBS, was unsuitable for running a database, and that they need not abandon AWS as a whole. In other words, the crowd made a bad and inaccurate decision. Even assuming that they had access to the raw data, they don't have the expertise required to interpret that data. Another real-world example: Raw global warming data is often analyzed by laymen and their conclusions are used as evidence that global warming is fake. In reality, it takes someone knowledgeable in climate science to understand what the data is telling them. This is why we don't defer to the broad consensus of the population on whether global warming is actually real. You don't address this at all. In the real world, decisions that require some level of expertise are far more common than decisions that require none. Deferring to consensus would lead to suboptimal results. This is fine if you are able to opt out of these sub-optimal decisions, but what you are proposing is, in effect, tyranny of the majority. Under your system, I am subject to the decisions of people who likely do not have the qualifications to be making proper decisions. This is a serious problem, and one that you have repeatedly ignored.
|
|
#
¿
Feb 1, 2012 21:05
|
|
|
Eripsa posted:But I've given examples of how to address it already. Expertise (which is really just a matter of having paid extended attention) ultimately provides weights to the attention economy, which is used to evaluate the decisions being made. This point is flawed. The Google page rank algorithm originally worked on such an idea: pages with more links tend to have better result. Turns out that this is easily gamed, and future versions of their algorithm moved away from using this as the basis for ranking. Another example is movie stars. Their views get a lot of attention, but they are often not well informed on what they are talking about, nor experts. Attention does not necessarily correlate with expertise. Eripsa posted:Your concern is that Art, who has spent the most time at the market, and knows best how it works, will have a judgment that ultimately has more merit than a random person with no attachment to the market at all, but that my system weights them equally. Except that it frequently does not work this way, and large groups of people who dissent are often not capable of recognizing their lack of capability. There is actual scientific literature along this effect. See Dunning-Kruger. My global warming example, in particular, demonstrates this. Quite a lot of lay-people think that global warming is made up and that they are clearly able to see how the experts are wrong. These people are incorrect. But under your system, their policy decision wins over the experts. Likewise, this is a persistent problem with Wikipedia where groups of people will disregard experts' information because they think they know better or have a better understanding. A little digging brings up plethora of examples, from the anti-vaccine movement, to SOPA proponents, to armchair economists. In other words, your version of events is highly idealized and not likely to work out that way in practice. Eripsa posted:But not necessarily. Art might suddenly go mad and start demanding that they sell the beating hearts of little children at the market. And though Art has put in a lot of effort into the market, you would hope that the crowd is capable enough at recognizing how bad an idea that is, that they won't let Art's popularity outweigh the clear insanity of his view. Not a good example. You've used an extreme example where the crowd is clearly able to tell what the correct answer is. A more poignant example is if I, a fairly skilled software developer, came to a group and said that I think we need to replace a sub-system with a rewritten alternative. A group might find my explanation reasonable, but they actually have no ability to rationally gauge my points other than relying on reputation. On the other hand, a skilled developer might analyze my position on its actual merit and decide that I am making a mistake. In other words, you still have problems unless you specifically exclude people that do not have the capability to analyze decisions. This is generally how peer review in science works -- papers are submitted to a smaller community that has the proper credentials to provide input. I'd encourage you to actually think deeper about some of the claims you are making. In particular, you claim that the internet itself is an example of how community driven consensus works, but on deeper inspection it really isn't. In the few examples of consensus working, its because you have alternatives and the ability to walk away from decisions you don't like and defer to experts. In your system, I am not given that option. Another aspect here is I very much dislike the fact that you have given no room for privacy in your system, which is a non-starter. I do not wish to make all information public. In fact, your system comes off strongly as totalitarian, because if I disagree with baring my life to the community and acquiescing to their communal decisions then I face negative consequences, even if I have valid reason to go against the community's wishes. You seem to think that making decisions communal fixes this problem -- it does not. It makes it worse. Eripsa posted:Again, I'm resting on the empirical evidence from the internet as a test case. The internet proves quite conclusively that if you grant open access, a lot of people will do lovely things with no value, but enough people will collect around the important things with a lot of value that the shittiness is far outweighed by the good. In my estimation, the good of the great open projects on Internet (like Wikipedia) so far outweigh the bad of lovely youtube comments, that the former easily makes up for the latter, even though there are more people who contribute to the latter than the former. Well the problem is you're resting on a far-too superficial analysis of the internet. A deeper examination of the the evidence shows the opposite - crowd sourcing usually does not lead to good results unless you specifically limit the crowd to people who have qualifications to make a decision. You've taken your own broad sentiment of the internet and enshrined it as fact without actually making a deeper examination of that sentiment.
|
|
#
¿
Feb 1, 2012 22:46
|
|
|
evilweasel posted:You might, if you were utterly ignorant of the history of pre-Nazi Germany and the rise of the Nazis, and merely were babbling based on a vauge theory of what happened. Or Japanese Americans in WW2. Or homosexuals. Or black people. Tthere are tons of forms of bigotry and oppression that operate with the approval of the majority. How many people thought that homosexuality was a choice? How many still think that? I take it as a point that your system is not implemented, so we don't really know if a proper completely democratic system would have stopped these things, but I think you're being a bit naive by assuming that the populace would have stopped these clearly wrong things if they had direct power instead of proxy power via representatives. VVVVVVVVVVV Best Friends posted:What I really love here is the supreme irony that Eripsa's plans are failing the only metric Eripsa recognizes as valid - crowdsourcing. Eripsa takes this to mean that in this and only this instance, crowdsourcing itself is showing the deficiencies of the crowd. Crowdsourcing remains, of course, the optimal method for distribution of toothpaste. I brought this point up before, and so far he has completely ignored it. If a small group of goons can't arrive at the "correct" decision after nine pages of discussion, what makes him think that this system will scale?
|
|
#
¿
Feb 1, 2012 23:13
|
|
|
How much do you want to bet that he claims that no-one has brought up relevant arguments against his system again, despite having posted a multitude of examples of crowdsourcing failing and explanations as to how it can lead to suboptimal results? I mean, his argument in favor of crowdsourcing is basically, "Hey guys, the internet has some cool things that are the result of people working together to contribute," which is all well and good, but its completely devoid of any deeper inspection of how well those things work, or whether its extensible to more important subject matter.
|
|
#
¿
Feb 1, 2012 23:28
|
|
|
Even setting that aside, how is a large group dictating things like how much toothpaste I am allowed not tyranny? Are we supposed to just deal with the fact that our freedom is severely limited for the greater good?
|
|
#
¿
Feb 1, 2012 23:56
|
|
|
DMBFan23 posted:If by "counterproductive" you mean "limiting the amount of racism, misogyny, and rape jokes", and by "not require moderation" you mean "will be a festering pool of poo poo and rear end for lack of it", then yes I think I see your point. Reddit is a particularly terrible example, because the noise to signal ratio is so high, even if you discount bigotry. Useless memes and pictures are upvoted, and intelligent discourse is not. What few articles meet the outrage bar enough to get to the front page are usually accompanied by hilariously inaccurate headlines and analysis. The prevailing wisdom, even on reddit, is that you have to remove all the popular subreddits and search out the niche subreddits to make it worthwhile. That's basically a vindication of the view that crowds really suck for high-quality judgements, and that smaller, selective groups can provide better decisions. The OP might want to avoid trying to use Reddit to prove his point. Also, pretty much every social media aggregation site has this problem. Slashdot had it (despite ostensibly being for smart people). Digg had it. Reddit has it now. Hacker News is starting to show signs of this. Something Awful forums also ostensibly have this problem. If you want me to use social networking and social aggregation sites as the benchmark for plausibility, then my reaction is that this system is likely to end poorly.
|
|
#
¿
Feb 2, 2012 01:00
|
|
|
Eripsa posted:The social networks you describe do face a number of problems, and those problems do require solutions. One solution is to restrict who can participate network. One solution is the appointment of moderators. One solution is to crowdsource the moderation and let the crowd up/down vote to adjust what is seen by the typical user. These methods will all work or fail to work in different contexts and depending on the situation, but the point is that there ARE solutions to the moderation problem (aka the problem of lovely Human Beings), and we know those problems don't require solutions that violate the basic digital paradigm. Crowdsourced solutions can solve the moderation problem; and the problem itself doesn't necessitate setting up a centralized body with a monopoly on power. And our point is that they don't solve the problem. In reddit you have actual moderators and crowd-source moderation. The result is poor. In Slashdot, you had a crowd-sourced meta-moderation group that was selected out of the general users. The result was still poor. On Something Awful, you have appointed moderators. The result is often poor. In other words, moderation and the problem of lovely human beings is unsolved. That's fine for a website, because you can opt out of visiting the site. It's not good for an economic system that you cannot opt out of. Eripsa posted:Just real quick, I want to point out that I've never denied the problem of lovely Human Beings, and in fact I have often conceded it is a real problem, and that (just like the internet) my system isn't structured to overcome that problem. I make that concession as a matter of realism, to show that I'm not a wide-eyed idealist about human benevolence, and that I don't think my system is meant to be the salvation of humanity. I have only claimed that my system will solve the coordination problem, and although it accounts for lovely Human Beings it doesn't have any built in protections against them. The problem is that you're 1) underestimating the average human's capacity for being lovely and 2) giving lovely Human Beings undue power without proposing any realistic way of handling this. It's one thing to have lovely human beings in a system where you can walk away or otherwise work around them to some degree. Its another to have them in a system where they can control the minutiae of your life. And yes, under the current system, there are certain areas where lovely human beings have undue control over peoples' life. Your solution is to give them more control, and gee-golly hope that crowd-sourcing takes care of it. Forgive me if I'm not too enthusiastic about that option. Eripsa posted:In any case, I think it is completely obvious that there ARE decentralized solutions to the moderation problem, and though there are also centralized solutions I think there are some reasons to prefer the decentralized ones. It is not completely obvious, and you need to stop claiming it is. You're dodging the argument. You have not proven that distributed moderation systems can lead to good enough results to justify granting them high levels of control. Eripsa posted:I do find it quite amusing how many people are quick to jump to shocked accusations of "tyranny of the majority" when I suggest distributed solutions, as if the system we have isn't tyrannical in so many more extreme ways. You haven't established that the tyranny of the current system is worse than the tyranny of the proposed system. We have outlined several ways in which your proposed system is worse. Please stop claiming that your system is better until you address this. Eripsa posted:I was chastised for suggesting that there should be some public check on abuse of resources, with people essentially arguing that they should be able to waste resources without any checks on their abuse whatsoever, just like we do in a proper free country like America! You were chastised for proposing a system that gives undue power for busy-bodies looking to pry into peoples' personal life. Excessive resource usage is undesirable, but eliminating all semblance of privacy and giving people excessive power is worse. Eripsa posted:Never mind the fact that our government can steal you from your home in the middle of the night, lock you in an overseas prison, and throw away the key and that not only can they but we know they have done it and will continue doing it and that we have literally no legal protections or any expectation that justice may prevail in such a system now or at any point in the future. But heavens to betsy someone cares if I'm wasting toothpaste why that's wurse than a goolog. You act as if the people complaining about your system do not also have problems with the current US government's actions. That's called a strawman. We're specifically saying that your solution doesn't make anything better, or at least you haven't made a convincing argument that it does. Its not as if a crowd-sourced solution is going to necessarily be better, given that the current population seems to be largely OK with these things, so long as they're done to "them" and not "us." Again, keep in mind that we have practical examples in the current system. Any new, hypothetical system has incredible potential to be abused and lower the standard of living. Thus your system cannot simply be hypothetically better, but you must present a very strong argument that it will be better in practice. You have utterly failed to do so. And no, I don't really think that people micro-managing my toothpaste is going to be an improvement. Please quit assuming people agree with your unsupported assumptions. Eripsa posted:Right now, right this very loving second, your freedoms are severely limited because that's what the wealthy capitalists would prefer. I'm suggesting we take this power out of the hands of the wealthy elite, and putting that power entirely in the hands of the people. Well, I have a retirement account, so that makes me a capitalist, although I'm not wealthy. In both systems my freedom is limited, but in your system my freedom is limited more than in the current system. This is not that hard to understand. If I think I need something, then I can go out and buy it in the current system. No-one is going to come in and tell me that I can't buy it because they don't think I need it. And yes, this assumes that I have money, but surprisingly enough a decent number of people who are not rich still have money, even in this crappy economy. I can purchase things, change my living quarters, move about, find new employment, and pursue my interests without having my decisions meta-analyzed by Redditors or Twitterers. And yes, the poor are severely limited, but that's why I advocate for strong social safety nets. Coincidentally, that still leaves me with the ability to decide for myself what I should be able to buy and use. Bringing up capitalists is a dodge. The fact that a decision is arrived at by consensus doesn't mean that decision is somehow automatically fair. The capitalist system isn't perfect and needs serious reform or perhaps even replacement, but not by a system which removes any and all choice in how I can manage my life. Eripsa posted:"Let all the people, together, decide what is best, instead of the wealthy few" "The current system isn't perfect and your freedom is limited in part by your wealth. Let's just take everything away and force you to submit to the will of the majority, regardless of whether its actually good for you. See, it will be good decisions because we crowd-sourced it!" "Yeah, so you're going to reduce my freedom and hand my personal life to the internet. Not a very good deal for me, even with the current system screwing me over. Also, you're a tyrant."
|
|
#
¿
Feb 2, 2012 01:59
|
|
|
Has he dealt with the crowdsourcing? I feel he hasn't. He has yet to address how to properly moderate the crowdsource to prevent sub-optimal decisions or ethics violations. I mean, unless you count "assume it will all turn out OK" as addressing it.
|
|
#
¿
Feb 2, 2012 04:04
|
|
|
I find myself agreeing with most of those points, except when I get to this part:Eripsa posted:And maybe when all the ore that all the willing people have mined is used, and there is still demand for ore, then maybe, just maybe, we will go without ore. The problem with that statement is that it comes with the implicit assumption that we can go without that ore that we demand. If it means that we don't build a super yacht, then I think that most people won't feel that is too much of a loss. If that means we don't have sufficient iron for girders to build a school or hospital, maybe less so. There's the underlying assumption that when the willing people stop, then that is how much is "enough," which doesn't follow from the premises. Edit: Er, also missed this: Eripsa posted:Maybe when all those inefficiencies are ironed out, we'll have a task that is manageable enough that a small group of passionate and experienced people can approach it with the support of their technology and for the benefit of the crowd. Yeah, I disagree with that too. There are some jobs out there that people simply don't want to do, no matter how easy you make them. For lack of passion, you have to provide some other incentive. You lost me at the important part.
|
|
#
¿
Feb 2, 2012 04:26
|
|
|
Eripsa, the problem I have is that you're assuming that sustainability must come at the cost of privacy via your system. My preferred method would be to strongly monitor the resources at the input. In essence, I don't believe that natural resources should be owned privately but in aggregate for the reasons you outlined. I also don't think that I have a "right" to consume without limit. However, within reason, I would like the luxury of consuming how I choose, without having to make my consumption public. If that is simply impossible, then so be it, but my argument is that you have not made a case that it is. There is a huge gap between "everyone is raping the planet," and "everything is collectively owned and shared," with a lot of potentials in between. To answer your hypotheticals -- no I don't believe that first world consumption patterns are sustainable, and I am willing and able to lower my quality of living to match that. But you make the assumption that the drop in quality of life under your system is slight. It very well might (and I'd argue its likely) be massive. If you wish me to take a massive hit in quality of life, then you have to make a drat good case that its necessary and that other, better systems cannot be used. Moreover, I don't buy the appeals to populism. To me, it matters not if my actions are being controlled by a cabal of elite or a collective - it impacts my life in the same way. And I have no more chance to change the decisions of a democratic body than my elected representatives. People who promote this style of collectivism, as it seems you do, always seem surprised that we don't see this as an unmitigated good. I like my individuality. Being that I'm an introvert, it seems like hell to toss aside my privacy and submit to the collective in such a way. If its absolutely necessary for the human race, then well so be it, but we're also a very very long way from proving that it is necessary. And really, that's not even touching on the feasibility of your system which you've still failed to prove. Your pithy comment that you don't buy that people cant share isn't really based about any sort of reasoned analysis, but around how you feel. That's a terrible metric. We can see how crowd-sourced systems work -- they often do so relatively poorly. We have a decent understanding of human psychology. If you take a really hard and honest look at the systems you hold up as examples, you'll find that they are not particularly promising for solving difficult problems. Maxwell Lord posted:What're people's thoughts on distributionism in general? I only know the basics- everyone owns shares in the means of production and receives dividends- and it seems like it has some upsides (the modern corporation is a little oligarchy to itself most of the time) and downsides (some industries may benefit from not having a democratic process resolve every market decision.) Ah, I've had this idea tossing around in my head lately. It's nice to know that there's a name for it. Require that public companies give a certain percentage of their equity (a controlling share at least) to the workers, and also require that they pay a certain percent of the profit in dividends. Hopefully, such a system should cause profits to go back to the workers. The fact that they're paid on dividends may prevent the "growth at all costs," issue. Still, I think its not a really strong solution to the sustainability problem. And the big problem that I can see is that, yes, democratic decisions are often not good for making market decisions, for reasons outlined in this thread. They require expertise. But hopefully, the workers could elect someone to make these sorts of decisions on their behalf and fire them if they fail to make good decisions. Also, I haven't read a lot of Marx, so could someone chime in whether such a distributionist system is compatible with his ideas?
|
|
#
¿
Feb 2, 2012 07:27
|
|
|
Eripsa posted:You said this earlier, and I didn't call you on it. Where the hell did this come from? Every one of the internet communities you mention are thriving, flourishing communities that have sustained an active population of regular contributors for years and in some cases over a decade. That's an amazing success, given that before these genuine digital communities existed it was an open question if anyone would even bother to use the internet. That's an incredibly superficial metric of success, and I have called you on that before. Do the sites given accurately summarize and analyze the articles they post? Do they often post articles with a high quality and low noise? Do the highly-rated commenters read and properly understand the articles? Does bad analysis get properly downvoted or left behind? The answer is no for all these sites. In all them, summary and analysis is terrible -- misleading headlines, biased sources, and commenters who likewise promote their own bias instead of a nuanced understanding. Group-think and knee-jerk reactions are upvoted. In many of these sites, noise greatly overwhelms signal in terms of quality. Posters contributing bias instead of reasoned analysis are more highly regarded than posters taking a rational look at the issues presented. Yes, in regards to growing the sites and drawing people in, they are successes, but in regards to promoting greater understanding of issues and quality discussion, they are failures. This is what I meant by you giving superficial platitudes with regards to the internet's success. You have not properly analyzed any of the things you use as examples. Eripsa posted:I also have no idea where this came from. I've cited actual psychology that is incredibly optimistic about our ability to share and cooperate, and to be motivated by the goodness of the work we do over the material value of the reward we are given. Unless your pessimism is anything other than anecdotal, this is also unwarranted. Scientific research also shows that people are more interested in having more than their neighbors than they are having an equal number of things in common with neighbors. Also, people are easily gamed and irrational. And they tend to organize hierarchy and like conformity. They do have capacity for sharing, but they also have capacity for greed, and for being manipulated. Humans are a mixed bag, but you can't just reach in and only take the candy you like. If you don't address basic human nature at all, and just hope the good parts shine through, then your system won't work. It's just that simple. I mean, how do you reconcile the Milgram experiments with human benevolence? Assuming we recognize experts or authorities on subjects, then that becomes a problem. And we also have the bystander effect too. Eripsa posted:Yes, slavery seems like a good strategy. It works! I think we have an obligation to do better than loving wage slavery. Jesus loving christ. I'm really not sure how people can equate working for a living to slavery. Remember that in the absence of society and any sort of effort, the default state is for you to be dying. Entropy is a bitch. Continued life is the result of effort on someone's part towards that goal. Wages are just an abstraction of an abstraction of an abstraction of the idea that one has to put effort in to stay alive, and that people should contribute effort towards that goal.
|
|
#
¿
Feb 2, 2012 16:53
|
|
|
rudatron posted:No, it's not. attention doesn't measure anything than what you are looking at/thinking about. Things that you don't currently see, that could still be in use, aren't going to get your attention. Are you looking at your refrigerator right now? If not, then how is 'ma' going to be able to continually feed it power? It doesn't solve the problem of how to bring about improvements that people do not understand or know that they need. Would you be able to get attention for a car in a world of people that own horses? After you've built the car, the advantages become obvious. But history is replete with people building things that people did not know would become valuable at the time of their invention. See transistors, for example. The Duke of Ben posted:Eripsa, you've repeatedly held Wikipedia up as an example of crowdsourcing effectively. Does it occur to you that Wikipedia has a leadership (government)which makes decisions? This includes adding and removing admins (top down decisions, denying access to certain users), deleting unwanted articles (denying access, silencing voices), and generally policing (law enforcement) the site? Wikipedia added these things because crowd-sourcing failed for moderation of their community.
|
|
#
¿
Feb 2, 2012 16:57
|
|
|
Eripsa posted:I'm sure life is great for the middle class of America, but that poo poo is unbelievable luxury for the majority of the world, and I'm sorry you won't get to keep your comfortable middle class life after the revolution. I almost feel bad for you, really I do. And I'm done. See you guys later.
|
|
#
¿
Feb 2, 2012 18:02
|
|
|
The Duke of Ben posted:In Capitalism, you either work (for money) or die. In Attention Economy you either work (for no personal gain), or everyone dies. Just popping back in here that my pointing out that life doesn't come "for free" and requires effort, and that wages were an abstraction of that, was met with venom, open hostility and Erisa wishing me ill, while at the same time making broad assumptions about my background. Something to think over.
|
|
#
¿
Feb 2, 2012 18:45
|
|
|
I know I said I'd stay away since Erispa seems to care more about calling me selfish and hoping I become poor than addressing my concerns, but I figured I'd try to throw this in here because it is relevant to a field in which I am experienced. You have a weighted graph of "attention" that we want a computer to process to find optimal production. I understand the problem correctly, then this is a variation of the Travelling Salesman problem. This problem is NP Complete. What this means is that finding a solution scales very poorly. What that further means is that you're unlikely to be able to scale this to an entire economy. Now, you may want to simply put up with sub-optimal solutions. This will make solving the coordination problem harder. Lower efficiency means that more inputs need to be put in to make sure everyones' needs are met. Also, I'd like to point out that his system is the very definition of the Free-Rider Problem: http://en.wikipedia.org/wiki/Free_rider_problem I mean, if you had a bunch of game theorists sitting around trying to come up with an idealized example of the free-rider problem, it'd probably end up looking something like this system.
|
|
#
¿
Feb 2, 2012 21:41
|
|
|
Eripsa posted:Good. Perfect. I think you are right that this will scale poorly. It needs massive computational power. We would have to take all those computers running the stock market and use them to model the actual economy. It will be computationally expensive. Well its a start, but I doubt that the computers running stock markets would be enough. Of course, in such a system you'd probably also throw in all the hardware modelling the stock market in an effort to make profit. It's still not going to be enough, though. You over-estimate how much hardware goes into the stock market and dramatically underestimate how hard NP-complete problems are to solve. The realistic solution will almost certainly be to put up with sub-optimal resource distribution, which is what we have now, and the only variance is whether we can make it more equitable. evilweasel posted:Let's add NP-complete to the ever-growing list of things you don't understand, as we ponder what you just said. I'm going to be completely fair here and say that I'm not 100% sure that its NP-Complete because I haven't bothered to sit down and do the necessary transformations to Eripsa's system to prove NP-Completeness (or NP-Hardness either), so its entirely possible that I'm wrong here. Also, I haven't done that sort of computer science for 6 or so years. Are there any other computer scientists here that are willing to fact-check me?
|
|
#
¿
Feb 2, 2012 22:07
|
|
|
So my memory was fuzzy and basic Travelling Salesman (where one finds the optimal path through the graph) is only NP-Hard and not NP-Complete (which is a version where you are given a length and asked to determine whether any path is shorter than it). To illustrate, http://en.wikipedia.org/wiki/Travelling_salesman_problem#Exact_algorithms According to Wikipedia, the latest solution for the problem had 85900 points and took 136 computer years. There are 7 billion people on the planet. Keep in mind that by definition this does not scale linearly. OatmealRaisin posted:What the gently caress He's already explained that we should not expect privacy in his system, so I don't see why this is surprising. Well, unless you're concerned with the technical feasibility of tracking 7 billion people in real-time, which is a good concern.
|
|
#
¿
Feb 2, 2012 22:22
|
|
|
Eripsa posted:Are you suggesting that the only thing holding back the creationists from taking over everything is money? Are you really suggesting that capitalism alone is the last barrier between us and an onslaught of pseudoscientific powergrabbers? No, the thing holding back creationists is representative government. In particular, the judiciary. Creationism has passed popular vote before. A large number of people believe in Young Earth Creationism. Its now under 50%, luckily: http://en.wikipedia.org/wiki/Level_of_support_for_evolution#United_States However, 43% is still enough to influence a system in which votes in which attentiveness is used as a weight.
|
|
#
¿
Feb 2, 2012 22:30
|
|
|
Eripsa posted:Hold on a second. You don't need to solve TSP to track all the data. My only claim here was that it would be computationally complex to track all the data, not that the computers on their own could optimize it a la the TSP. You can collect the data, yes, but at some point you have to analyze it. If you want to find an optimal solution, then it mimics TSP and is NP-Hard. If you're content with sub-optimal solutions, then you can design sub-optimal algorithms that get good-enough results. But that means that you are going to have inputs into your system that you don't need and are wasted due to inefficiency. Also keep in mind that "good enough" has to be real-time. That's real-time in the computer science sense, not the layman sense. Eripsa posted:The optimization part isn't solved by the computers themselves, but by people looking to optimize the system with the benefit of the data and the computer models. Humans aren't guaranteed an optimum solution either, and I've repeatedly said my system does not guarantee optimality. So no where have I claimed that I have a TSP solution. This is pretty much meaningless. If humans could optimize the system they would be doing it already. If it can't be handled by a computer, then you're unlikely to have humans able to optimize it either. The "free market" works sort-of like evolution in that it says "There's no way in hell we can optimize this. Everyone should do whatever, and more optimal solutions will win over less optimal ones." Its basically genetic programming applied to economy in a non-centralized manner. There are severe problems with this approach. I'm pretty sure you agree with me on this point. I'm not saying that your system can't use such a sub-optimal system, but you seem to be hanging your hat on efficiency and solving the coordination problem, which highly implies that you need something closer to an actual optimal solution.
|
|
#
¿
Feb 2, 2012 22:41
|
|
|
So, I'm going to point out that I was wrong. I believe that Eripsa's attention graph would be fundamentally directed, whereas the Travelling Salesman problem involves undirected graphs. That means that the Travelling Salesman problem is not transformable to Eripsa's problem, and Eripsa's system is not necessarily NP-hard. The actual computational complexity of this system is still left as an exercise to the reader. Sorry for the mistake everyone. I'd like to point out a caveat that "solving" graph problems are still pretty hard, which is why companies like Google and Facebook require top talent.
|
|
#
¿
Feb 2, 2012 22:52
|
|
|
Eripsa posted:But I think that some luxuries that Suburban Americans have need to be stopped. For instance, I think the luxury to waste resources without any check on that waste is bad. Starbucks throws away bags of bread each night because it can afford to, and this is a luxury I don't think they should have. Ryde or whoever is proud of having this luxury, but I think it is intolerable. That's a pretty gross and insulting mis-statement of my position, Eripsa. My position was pretty much the opposite of what you claim it was. I explicitly stated, several times that I would be willing to lower my quality of living in order to support other people. I disagreed with your solution to the problem, and you twisted that into a moral failing on my part. Shame on you. Eripsa posted:My argument has been that money does not let people do whatever, and so we should free them up to pursue their own interests and let the optimal ones win out, and that we should let the crowd select the criteria for optimality. It actually does work that way in theory. You can do what you want, so long as it is legal. If other people find value in what you do, you can receive compensation for it. If other people do not find value, you do not receive compensation. If you do something better than someone else, you receive compensation rather than then. Thus, everyone does what they think is valuable and the overall system will replace less optimal choices with more optimal choices. Putting aside that the other reasons the free market does not work that way in practice, the main problem with this system is that if you are not doing anything someone finds valuable, then you receive no compensation and you die unless you are self-sustaining. The reason this problem exists is that the system does not have any built-in concept of intrinsic value of a person. If you do not have a value to society greater than the cost to keep you alive, then you're on your own. And people tend to die unless effort is put in to keep them alive. In other words, the intrinsic value of human life is not a variable in the system, so it is not optimized for. Most reasonable people myself included, feel that this is a bad thing, and that's why pretty much every first world country, even the USA, has systems to mitigate this effect. That's why I don't advocate pure capitalism. I'm only a capitalist via technicality (the fact that I own equity via retirement savings). If your system did a better job of ensuring that everyone got what they needed, then I'd support it. But I don't think it does.
|
|
#
¿
Feb 3, 2012 00:08
|
|
|
OatmealRaisin posted:What on earth are digital values anyway? Raging at some girl because she won't gently caress you? Voting RON PAUL? Looking at pictures of cats? He posted them earlier in this thread. Something along the lines of broad involvement, inclusion, equal voice, and distributed contributions.
|
|
#
¿
Feb 3, 2012 00:20
|
|
|
Achmed Jones posted:They are values that are binary: either you agree with a science fiction techno-fascist dystopia or you literally think food wastage and slavery are good things. Welcome to the future beep boop. Yeah, this is apparently what it means in practice. It would be nice for Eripsa to come back and admit that he mis-represented my position, especially since I stated that I am unequivocally not in favor of endless consumption or people not receiving basic needs.
|
|
#
¿
Feb 3, 2012 00:26
|
|
|
Eripsa posted:I interpreted this as making the following two claims: What I was saying is that people should not have the ability to dictate my purchases, and that I do not think that dictating my purchases is the only way to solve poverty. This is in no way saying that I believe in endless consumption. Its possible to have a checked level of consumption that still allows me the freedom to choose within those confines. Whether I have the freedom to make purchasing decisions is orthogonal to whether I have the freedom to over-consume resources. Put another way, I can buy one to ten jellybeans, and I don't have to have my purchase meta-analyzed by twitter. But I am unable to buy ten million jellybeans. How do you prevent me from buying ten million jellybeans without actually having veto power over my purchases? Well, that brings us back to the discussion now, doesn't it? You have your system, which I do not agree with. Theoretically, it could also be done by limiting my purchasing power for that item. If a jellybean is 50 cents, then I'm not going to spend 5 million dollars purchasing jellybeans. I don't have a formal solution for this, but my intuition is that it would start by holding natural resources in common rather than privately and properly distributing externalities throughout the system. Put in that light, its more a system of equitable wealth distribution combined with internalizing externalities. No privacy violations needed. So let me bring up this point: You ask for people to read your statement's charitably, but you fail to do the same for others. I actually specifically stated I don't support endless consumption but you read over that part in an effort to paint me with your wide-rear end privileged-middle-class-American brush. So lets go over my consumption habits. I don't own a car. I walk to work. I live in a 700 sqft apartment. My power bill is $70 a month at most, and $30 at least. I give to charity. My only vice in terms of consumption is electronics. I'm a software developer, so I have a lot of gadgets. But if sustainability demands that I give up some of these I am more than willing to do so. I don't think I have a right to consume to the point that other people suffer. I just don't like your system. I don't think it will work. That doesn't make me selfish. You need to quit painting disagreement with your system as moral failing. Eripsa posted:But things don't get done by money. Money doesn't do anything but work as a coercive agent to keep the people toiling. What gets things done is people with skills and talent and knowhow who are willing to put in the work and time and effort to get things done, and those people aren't going anywhere and will still be around when we blink money goodbye. And what others have been trying to tell you is that this is not an inherent property of money. This is an inherent property of entropy, and the physics of the universe. If someone does not expend effort to continue a human's life, either that person themselves or society, then that person will die. This is true in your system as much as it is in a capitalist system. This is why everyone in this thread is advocating for enough of a safety net to keep people alive. The difference is that money can still exist as an incentive above peoples' basic needs. If everyone stops working, then people start dying. Reality is inherently coercive, and its just a matter of how many layers you throw over it.
|
|
#
¿
Feb 3, 2012 00:48
|
|
|
Cefte posted:Capitalism isn't a globally mandated system with universal information access, universal propagandizment and no opt-out. To mix a metaphor, if when a system goes wrong, it goes wrong in exactly the same way for everyone, I don't want any truck with it (company store, teehee). Actually, I have this problem with a lot of economic systems. Its better in Europe because you can pick up and move a country over. This is sort-of why I trend anti-federalist for most government spending, because I'd really love to see different US states with different economic systems and adopt the ones that work well. The flip-side is that spending by a large, central institution benefits from economies of scale whereas a myriad of smaller entities does not.
|
|
#
¿
Feb 3, 2012 01:09
|
|
|
paumbert posted:I noticed you haven't addressed this. I'm not trolling: this is a foreseeable and serious concern that your dictatorship of the loletariat creates, and one that isn't implicated by other concepts of planned economies. I think that the twitter council is only intended to discern who is consuming the "appropriate" amount of resources or not, and for directing economic production, in which case such a decision would fall outside their purview. In other words, they don't have the authority. At least that's my understanding. Edit: Of course, economic power often translates to actual power, so who knows what mischief the twitter council could cook up. Edit: According to Wikipedia, the Asymmetric TSP problem, which uses directed graphs, counts, so we're back to being NP-Hard!
|
|
#
¿
Feb 3, 2012 02:14
|
|
|
Eripsa posted:Take the internet itself. Billions of end users, and only so many servers to serve all those users. A computationally complex problem! But surely we know there are methods for load balancing and optimizing the flow of traffic across the network. That's not to say it isn't a difficult problem, or that the system can't be further optimized, but the complexity of finding the optimal path through the network doesn't itself imply the impossibility of orchestrating a global telecommunications network at all. We know we can do it. It takes a lot of computers, and you have to let the (possibly private) data flow freely across the network for it to work, but it works well enough. I'm sorry, but are you actually a software developer or computer scientist? Because these two paragraphs really make me think that you are not. In particular, things like routing and load balancing are not even the same order of magnitude of complexity as solving TSP, and neither is Facebook. You're applying some pretty broad generations about these things that make it seem like you don't know what you're talking about. This goes back to the whole experts thing I harped on earlier. Edit: Also, I want to make it clear that I'm still not 100% sure that your system decomposes to TSP, so the actual computational complexity may be far less than I am estimating. However, "Look at the internet!" is not a rigorous engineering argument for whether this complexity can be effectively managed. Your resource system does not behave like the internet. Let's do some actual back-of-the-napkin calculations. What kind of data gets stored in your attention system and whats the resolution?
|
|
#
¿
Feb 3, 2012 04:26
|
|
|
Eripsa posted:That seems like a decent starting point. Its not really. IPv6 has enough addresses for every atom in the universe, so its overkill. I'd assume a 64 bit identifier which provides for 4.6 × 10^18 possibilities, and is a fundamental data type in modern computer languages. So you have 16 bytes for a minimum of saying that X is paying attention to Y. That means 112 GB for one frame at the minimum. So then, the questions you have to answer are, "What other types of data are required?" and "What is the necessary resolution required to solve the coordination problem?" Keep in mind that lower resolution is still going to increase capacity. Assuming that a person pays attention to something different once every second, that means you're storing 403 terabytes per hour. If you want to encode things like "computer that Eripsa is looking at sits on a desk, and Eripsa is eating a bagel," then you're going to start getting more and more combinations to add to this. 403 TB/hr is likely a gross underestimate, because we're not including metadata nor are we including overhead. Eripsa posted:When did I suggest it was? I said that we can create and manage a complex network (and the internet is definitely a complex network in the sense of complexity relevant to the P?=NP problem) without solving NP class problems like TSP. Managing and optimizing the flow of data across the network is not the same problem as finding an optimal route through it. The computational complexity required for running the Attention Economy is a problem closer in scale to "managing the flow of network traffic" (with just a lot more nodes for all the tubes of toothpaste), and not a problem like TSP. Routing is not NP. I'm not entirely familiar with the overall system, but assuming no optimizations and a basic search algorithm, you're looking at O(log(n)) to O(n), if I remember correctly. In practice, we have things like routing tables. I would like to actually examine whether your system is an instance of TSP (up until now, I've only assumed so based on cursory examination, which is not a rigorous proof), so it'd be nice if you could roughly explain how the underlying computer algorithm that transforms attention graphs into a resource usage plan would work.
|
|
#
¿
Feb 3, 2012 04:40
|
|
|
Eripsa posted:Can you explain the math here? Where did you get 112 GB? My phone does not need 112 GB of data to track its location over the course of a day. Every person or object is identified by 64 bits, which is 8 bytes. "Attention" is defined by a graph edge, which can be defined as source and target. That's 16 bytes per attention edge. Every person in the world is paying attention to something. That's 7 billion times 16 bytes, which is 112 GB. Now we have to decide what the appropriate resolution is (how frequently we're going to collate and publish), roughly how many things on average people pay attention to in that time frame, plus any additional data required to encode your system (metadata and overhead.. for example is your graph weighted? How many possible weight values?) I assumed that someone pays attention to one thing per second, which leads to 403 TB per hour. That's still within the realm of feasibility for transfer and storage if the network is highly distributed, but I'm not so sure about computation. It depends on the underlying algorithm and whether we can throw away any of this data. Also, what your phone requires is irrelevant. We're not encoding the same information as your phone. You didn't answer my question: Are you a software developer? Are you a computer scientist?
|
|
#
¿
Feb 3, 2012 04:53
|
|
|
So I'm thinking that TSP is actually not a good fit for the problem, because it seems to me that the algorithm would need to exhaustively visit each node and calculate some sort of value based on the value of the edges, which doesn't match TSP. So I'll have concede the point and lower my claim that "it's obviously not computable" to the less decisive "its very likely not computable." Obviously, its hard for me to form a clear mental model of what the algorithm would look like because its description has been vague and hand-wavy at best.quote:So I pretended like we existed in the Attention Economy and resources were allocated according to the attention and interests of everyone in the world, and the nagging question in the back of my mind was "Who maintains the empire?" Like who is actually interested in mining, manufacturing, installing, maintaining, and recycling the infrastructure that monitors them ceaselessly? I thought the point was that computers would analyze how to turn resources into needed products based on the attention graphs, and that volunteers would step up to do the actual work. This presumes that the volunteers are willing to donate their work for their own personal satisfaction or the good of society, with no tangible benefit to themselves. So basically, the answer of who controls it all would be "the computers" wouldn't it? Of course, you also had the idea of the twitter council thrown around to keep people from over-consuming, meaning that all this data would have to be collected and published on these people.
|
|
#
¿
Feb 3, 2012 05:10
|
|
|
Sir John Falstaff posted:Don't forget that this system would need to track both what each person pays attention to, but also each object that is "in use," i.e. it would also need to keep track of every running refrigerator, each desk that is continuing to serve as a desk, each filing cabinet that continues to hold files, each window that keeps letting light in, etc. Not only that, but it would need to keep track of each component of each of those items. Then, it would also have to track back each of those items through the entire production chain to record upticks in the use of the components and resources used in each item--for example, if someone drives a car it would need to record upticks in the use of the iron that makes up each car part, otherwise the iron miner doesn't get credit for the attention paid to the car. I actually didn't catch that part, but my mind is kinda reeling from the amount of data that would be required for that. Presumably, there'd be some way of compressing or consolidating things like "A car is made up of these parts." in order to simplify things. But you're correct that this would have to be accounted for. The reason I'm asking for credentials is because he's making a lot of very MBA-ey style of observations. "The internet kinda seems like this, and the internet is workable, so this should be workable too." Very generic and non-quantifiable. That's not how it works. If he's an actual developer and I'm just a moron who's not seeing his point, fair enough, but I'd like to know where the score stands. As it stands, I feel like I'm listening to the SOPA hearings about how Google can automatically detect infringing content because "Y'know, we have a lot of smart people and they'll figure it out."
|
|
#
¿
Feb 3, 2012 05:33
|
|
|
Strudel Man posted:Computationally speaking, that doesn't actually sound particularly hard. The number of tasks is only on the order of the number of objects in the world, times their average number of components - which is not a particularly large number compared to the order N! tasks you get with e.g. traveling salesman. Well, plus the humans paying attention to the objects and all their edges. But yeah, I think I was completely incorrect with the TSP comparison. My original thought was that the attention graph would somehow map to a resource graph where edges represented resource usage and you'd have to apply a TSP style algorithm to it to find a way of hitting all the nodes (being required objects) while minimizing the total length (resource usage). But such a graph would be a DAG, and you'd likely just be applying normal search algorithms on it. I also think that the representation is actually not what Eripsa was going for. So I think I mistook the representation of the problem. Does that sound correct to you, Strudel? The actual computational complexity really depends on what we need to do with the data. Assumptions aside, Eripsa hasn't painted a clear enough picture to speculate on that, so I'm trying to get him to act like a project manager and write out some technical specs.
|
|
#
¿
Feb 3, 2012 05:42
|
|
|
Sir John Falstaff posted:I would think one of the critical limitations would be the GPS bandwidth required to attach a constantly transmitting GPS transmitter to every object in the world (even assuming a GPS transmitter could actually track whether, say, a desk was "in use" or not). Well the GPS transmitter would actually be broadcasting the data through something like a cellular network, so assuming its distributed enough it may not be a problem. I'm not familiar with how much data cellular networks can handle, so I have no way of gauging whether its enough. Still, 112 GB per second seems like the bare minimum, right? Strudel Man posted:Well, there's at least two things being talked about, the attention-logging (which seems relatively simple, and is also the only thing Eripsa's talked about in any depth), and the resource-allocation, which has the potential to be either phenomenally complicated or dangerously simple depending on the specifics. The latter, despite being vastly more critical for the functioning of the system, has been treated to a less detailed description, so a lot depends on what we can infer about it. The resource allocation was what I was assuming to map to the Travelling Salesman problem. You're correct that we don't have any information on the actual structure of this problem, so I was grossly premature in declaring it uncomputable. I guess I'm also not sure what he wants to do with the attention graph either. It doesn't seem simple in an absolute sense, but we're still in the land of feasibility, depending on the specifics.
|
|
#
¿
Feb 3, 2012 05:48
|
|
|
|
| # ¿ Apr 26, 2024 06:49 |
|
|
Eripsa posted:I've been talking about rfids. I have no idea why we are assuming this will be centrally processed data. Where would the data go and how would it be used? My guesstimate was based off of people, and neglected objects' interactions with other objects', so I'm not sure if staggering updates would get us below the lower bound I proposed. Strudel Man posted:As I see the attention graph, it's just a long list of object categories and people, where every tick (maybe a day long? doesn't really matter) each person's attention allocation from last tick is added to the appropriate counter based on their actions and/or what they're looking at. So basically, your ticket would just accumulate and send one glob along the lines of "X paid 5 attention to car, 3 attention to computer, and 10 attention to bread" per tick in one go? That would dramatically lower the data requirements and might get us below the 112 GB per second if we're being optimistic. It depends on what we're going to do with the data though. It needs to have enough information encoded to handle the requirements. If its a matter of just generating a list of "These are the top 10 things being paid attention to," then that's not hard from a computational perspective. I do suspect that we'll need more information, such as categories, who is paying attention to what (to allow for expertise), etc.
|
|
#
¿
Feb 3, 2012 05:55
|
|