Identity
Filter
Visitors
(1) (0) (0) (0) (1) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (61) (0) (0) (0)Subscribers
(0) (0)
Links
Blogs
Democracy
Search
[top-politics] Re: Lemming-DP
-
Lomax wrote: >Mark wrote: > >Lomax wrote: > >[…] > > >L: Mark expects the world to beat a path to his door. He bought that > > > mousetrap myth. Build a better mousetrap and starve, unless you learn > > > how to market it or engage someone to market it for you. > > > Mark, we are lemmings! Get it? > >-M: We are systems designers.
>L: Oh, I am what profession I wannabe?
-M: Its a hobby for now.
> >I expect there to be few lemmings among us.
>L: We are all lemmings, Mark. That you think you are not — and you do > think you are not — shows that you have bought your own exalted > image of yourself.
-M: I do have an exalted image of myself. I rock. How am I a lemming?
>L: And all of us can break out of leeminghood for times. We can’t do it > more than occasionally without the people in white coats taking us > away. Been there, done that, by the way.
-M: I hope that you were the one with the white coat. I want to be permenently out of lemminghood.
> >-M: My way, everyone is represented, always.
>L: Whether they want to be, or not. You will be represented. > This, by the way, is the status quo. Representatives in the U.S. > House are apportioned according to population, not according to > votes. The population includes many non-voters, including > non-citizens, children, etc.
-M: That isn’t necissarily representing me. By contrast, with SD2-S, all voting power(unless looped), becomes a quantified part of the trust network. It doesn’t get thrown out just because its not part of some majority.
> > > If a member is not represented, so? > >-M: Then they aren’t providing useful proxy information. > >The voter is the data input for the c-algorithm.
>L: Garbage in, garbage out. If the voter is assigned a proxy, then > something provided by the system that assigns the proxy is adding > weight to that proxy without any input from the voter.
-M: In that case, its being added by other voters. With your option, where no proxy is provided, its still other voters’ voice being provided instead(that is all that is counted). The only difference is that my approach attempts to smooth the rank distribution curve by giving bonus votes to underdogs. This makes the rank structure less clustered and entrenched.
So SD2-S actually does something with the voter’s proxy power instead of ignoring it.>L: This is not obvious to Mark because he does not respect the voter in > the first place.
-M: You are the one who wants to ignore the voter’s proxy power instead of doing something constructive with it.
>L: He’d rather design a system that, instead of amplifying the signal in the voter noise, would substitute another signal provided by the system
-M: No, I am trying to amplify the signal in the voter noise, and my system cannot create substitute signals. It sees only the rank distribution curve, and looks for anomolous gaps to fill. This also makes the ranks more differentiated, and therefore the data more useful.
>L: We want the choice of a proxy to be by conscious, voluntary choice > of the voter. The voter is one of the muscles of society. We want him > to function by his choice. Not because the system or some system > designer says he must.
-M: Yes, with SD2-S the voter is given the choice of real people or defaults. With your preference, the voter is given the choice of one person or being ignored
>L: This, among other things, means that it must not be a power structure, it must be about communication, voluntary communication. Adding inputs that don’t come from the citizens just adds, at best, noise, and at worst, it adds bias.
-M: “inputs that don’t come from the citizens” - Again, you are not describing SD2-S.
> >-M: “Automate member participation” vs. “participation to be a free > >choice”. Did it occur to you that the automation of member participation would > >be free choice participation?
>L: No. Because it would not. Automation, perhaps, of the potential for > voter participation, perhaps. In other words, the system could make > it easy. But if it is too easy, one might wonder if it is inviting > unconsidered input. It should be easy. But the system should not make > it very easy, as, for example, by presenting the voter with a > couple of choices for proxy automatically. It is possible that there > are ways such a process could be done without introducing bias.
-M: Your example here would be ‘too easy’ and would add the bias that you describe.
SD2-S would not add recommendations unless they are calculated to unentrench the rank structures.>L: But doing it by PageRank, if I understand this correctly, will merely > amplify the existing bias in the organization, instead of adding the > signal from the voter to that bias.
-M: WTF? All PageRank sees are the voter’s inputs and nothing else. And it is accumulative from proxy-to-proxy to an unlimited depth.
>L: There will always be bias, but we want it to be the general bias of the society. Mark might call this “lemmingism,” but that is only a problem when “lemmingism” prevents the consideration of new ideas.
-M: ‘we want…the general bias of the society’. Me to. And I want this measured by PageRank.
Gotta better idea?
If you want to measure this general bias,
then you must nail a c-algorithm.> >So if the voter gave the free choice not to select proxies, that the > >computer would automaticly select underdogs as proxies.
>L: Those are not proxies. Those are appointed voters. They have been > given added power without the excuse that this power was voluntarily > and knowingly assigned by the voter.
-M: Simple. Tell the voter: “If you don’t assign proxies, your proxy endorsement power will go to filling gaps in the rank distribution curve.”
Now its knowing and voluntary.>L: If the system selects “underdogs,” then it is introducing counter-bias of a kind, a signal > different from that which, on average, would be presented by the voter. It is, again, noise.
-M: No, it is a slight instability that can keep the rank structure from becoming too entrenched.
>L: There is no “underdog” problem with FA/DP. Underdogs simply organize > themselves and can be as large a force as they can muster. They will > be heard. But the system will not balance out underdoggedness such > that it becomes equal to topdoggedness.
-M: Nor will SD2-S. I simply want to smooth the curve, not level it.
>L: It leaves people as they are, but provides them with means of expression and an assurance that they will be heard. The automatic assignment of proxies does not set up the relationships of trust that we consider essential to the long-term effects of FA/DP. If Mark wants to offer automatic assignment tools to organizations, fine. But I think most of us won’t want that crap.
-M: Your system works as automatic assignment by ignoring the lacking proxy.
The result is the same as if an automatic assignment is assigned proportionally to those who did provide proxies. You aren’t making a comparison.>L: He is designing in bells and whistles that his potential customers — at least this > one!, and I might have some influence — consider more of a problem > than a solution. He is one of that class of abstract designers of castles in the sky, > utopian fantasies. And he refuses to connect with the real world. > Which is … lemmings. According to him.
-M: Doink! Lomax, calm down. I am only conditioning the data.
>L: He hates this. Why? Because they won’t listen to him. If they were lemmings following him, would he mind so much? Perhaps. And perhaps not.
-M: Do you expect people to follow your opinion here? If so, then following isn’t so bad, is it?
If not, then are you being pointless?> >Low-order DP is populistic. > >High order DP (four or more proxy chains deep) allows for better > >filtration and peer selection.
>L: This is an illusory distinction. The only thing that would make > Delegable Proxy “low order” as he defines it is a small size of the > organization. Large-scale DP will be automatically high order. That > is, the average chain length from individual member to top proxy will > be more than four. We can easily calculate this.
-M: I am not talking about the data itself. I am talking about how the data is processed,
because this it what will yield a decision.>L:[…] actual direct proxy count of twenty, then > with N = 4, we’d have 160,000 members. Now, what in the world does > this have to do with “better filtration and peer selection?” > It simply scales the organization.
-M: It matters how this hierarchy is generated. If it is generated locally at each layer by counting, then this is populistic.
If it is generated locally and across layers, both top-down, and bottom-up, then there is more peer-filtering.> > >L: You can open the doors, but unless people come in, you don’t have a > > > party. You’ve got some open doors. > >-M: It could be marketed as being competitive.
>L: Marketed by whom? By you?
-M: Yes. “Democratic decisionmaking without being burdened by the ignorant. Give SD2-S a try.”
I want to market to the intelligent and knowledgable, most of whom find the lemmings to be a burden.> >People’s general ranks could be posted. > > > > >L: Where is one? Show me! > > > > > > > >-M: I think that I have provided more specifics for SD2-S than you have > > > >for specific DP systems.
>L: We have very small systems in place. Because they are Free > Associations, with specific rules as determined by the members, the > specification is so simple that Mark doesn’t even recognize it as such. > (1) Define an organizational purpose. > (2) Set up a proxy list. This is just a list, it is not an automated system. > (3) Invite members to join and participate. > (4) According to standard process (I recommend Robert’s Rules as > adapted for on-line meetings), conduct polls. A poll result, again, > creates another list. Voter ID, vote. > that’s it. The rest is what the members do voluntarily. If they need > more advanced tools, they will find, build, or buy them.
-M: OK, but it would help you out if you had a turn-key system.
>L: If Mark has tools ready, I’m sure they would be considered. But, I can tell you, > if they research his writings and find all this “lemming” nonsense, > they will turn away, perhaps needlessly.
-M: ‘…“lemming” nonsense…’
- I have seen you agree with most of my lemming points, infact I have seen you have a willingness to call more people ‘lemmings’ than myself.
>L: Mark has understood a great deal, certainly about DP. He understands > some dimensions of the problem, and this understanding is unusual. > But he is also a lemming himself, and his lemmingness keeps him from > seeing beyond certain limitations. He can’t see how an organization > that is, in itself, powerless, could radically alter the balance of power.
-M: LOMAX! Power does NOT equal powerlessness.
>L: He does not see the power of advice and information. Neither of these > has power in itself.
-M: If it makes change, then it is power, regardless of how its being done.
An advisory organization can have more power than a command-and-control hierarchy.
This means that dirt-bags will infiltrate a powerful advisory organization just like they would a more centralized organization with the same power.>L: He also does not see the danger of centralization.
-M: I do see the dangers of centralization, this is why I want to do it better.
By contrast, you don’t see the essential nature of centrality, so again you aren’t making a comparison. You try to do last minute centrality as if this will magicly improve things.>L: His opinions about this, about power in general, are actually the “common wisdom.” His > dislike for direct democracy is a very old idea, including his belief > that most people are merely sheep. Or lemmings, his preferred term. > People are lemmings, yes. It’s a protection and a strength. But it > does introduce certain problems, and we believe that the kind of > communication that FA/DP should foster will solve those problems.
-M: “Communication will solve world problems.” Like it is something new?
> > > Mark behaves as if what is > > > being compared is a clearly established system, and what others are > > > proposing are mere theories. > >-M: Yes, my system is established more clearly than others.
>L: No. It is not established at all. It is a design. A design is not > “established” until it is in use. One design may be more practical, > though less completely specified, than another which has been even > completely specified.
-M: There are degrees of establisment.
>L: A completely specified error is, what?
-M: Does this apply? Where is SD2-S in error?
>L: Superior to a design approach that is, necessarily, open and therefore not completely specified?
-M: I’d say open and completely specified(SD2 and SD2-S) is superior to not completely specified. When something is nailed, it can be tested and optimized.
> > >L: Dysfunctional, again. I do not recall the specifics Mark is > > talking about. > >-M: Then use the search option. I found the input field by searching > >"input field".
>L: Whatever time I’m spending here is essentially stolen time. I’m > already spending too much. Do not expect it to continue.
-M: I am busy too.
> >-M: You are assuming that I can’t develop an incorruptable filter. > >You are being disrespectful and presumptuous.
>L: Yes. I am assuming that, because I consider it an extremely > difficult problem. I don’t think Mark has solved it, at all. But if > he is correct, surely his system will gain notice from the millions > of people who are searching for one. Or are there millions? > Probably not. There may be three or four. I’ve been one of them. I > abandoned the search, because I don’t think it is possible.
-M: Larry Page developed PageRank as a highly incorruptable ranking algorithm for webpages. And instead of using PageRank for webpages, by seeing the links, SD2-S uses it for political rank, by seeing the votes.
Almost all the work has already been done for me by the founders of Google.
And I have used it on a meatspace human network in Pakistan.>L: […] Mark is assuming that he can develop a centralized but incorruptible > system. It has not been done, it has never been proven.
-M: It has already worked for another human network, the WWW. Ever use the Google search engine? Does it give good results?
>L: My assumption, that he cannot, is highly likely to be correct. It is not > impossible that he could succeed, but I would not bet the time of day on it.
-M: Lighten up. This will work.
> > > >L: You are assuming that SD2-S is corruptable simply because it is > > > >centralized. Yet you don’t understand it, do you? > > >L: No. I’m basing this in an inference. It could be wrong. Show me. > >-M: OK – inference, which is uncertain.
>L: Yes. Now, show me a deductive proof that SD2-S is uncorruptible. In > the absence of that, induction is all we have.[…]
-M: I’ll go with real world data of similar systems.
> >L: I do have some ideas about what SD2-S is, but I have not seen sufficient > > > definition to claim to understand it. > >-M: Look at the above input field.
>L: I did. First thing I see is “manditory representatives.” Mispelled, > by the way, Mark should learn to use spellcheck.
-M: Spellcheck is not an option in the window that I am using. So you are my spellcheck.
>L: What do I do if I am undertain[uncertain?] about joining this organization and I see that? > I go somewhere else. I don’t like being subject to mandates, period.
-M: Then I will just call them ‘representitives’ or ‘general trustees’.
> >The general trustees stay the same, but the specialists change for each > >issue.
>L: This is utterly impractical. The “lemmings” cannot manage complex > issues outside their individual areas of focus, they simply do not > have time. Even to name specialists.
-M: Fine. I only need a small threshold for a decision to be made. This would work.
>L: […] We wish to keep the proxy system extremely simple, and this requires single proxy, with default proxies, those active in the absence of the single proxy, coming from proxy assignments of the absent proxy, thus concentrated.
-M: If you have more than one proxy-to-proxy layer, then these default proxies should already be active, regardless of the absence. With SD2-S, these layers are unlimited, and one’s proxy aquired vote is proportional to one’s rank. If a proxy doesn’t vote, no whup, the proxy chains are simply followed to tally up the issue votes of those who have voted.
>L: Encouraging the designation of alternate proxies destroys this simplicity.
-M: Alternate proxies:
- add redundency to the network
- keep proxy endorsements from being too concentrated
- allow a Markov algorithm to fork around loops (this gives deeper data)
- gives greater voter choice
>L:[…] Mark has got a few aspects of this vision, and in particular, much of > the reasoning behind DP. But he does not know how to get from here to > there, yet he is trying to design a system that will be used when we > are “there.”[…]
-M: We have been ‘there’ forever. Communication and polling is nothing new.
Its time for the next step, SD2-S.> >Now imagine the proxy chains to have unlimited depth.
>L: We imagine proxy chains to have unlimited depth, it goes without > saying in delegable proxy. If proxies are delegable, there is no > limit to the delegation, until it loops back. Theoretically, if > everyone names a proxy, there is a limit to the chain depth, it is > the human population. But, of course, information would move very, > very slowly across that “network.”
-M: So what if it has unlimited depth? How are you going to use this depth without a Markov algorithm? (And there are two networks, the one that is real and the one that is represented. Just knowing that a network has unlimited depth won’t automaticly connect this with application.) shanti
Mark, Seattle WA USA—~—~————-~—~——~——————~———-~—~——~ You received this message because you are subscribed to the Google Groups “top-politics” group. To post to this group, send email to top-politics@googlegroups.com To unsubscribe from this group, send email to top-politics-unsubscribe@googlegroups.com For more options, visit this group at http://groups.google.com/group/top-politics?hl=en -~—————~——~——~——~———~——~———~—~—-
+1
-
At 07:09 PM 11/4/2006, Mark wrote:
> >L: We are all lemmings, Mark. That you think you are not — and you do > > think you are not — shows that you have bought your own exalted > > image of yourself. > >-M: I do have an exalted image of myself. >I rock. How am I a lemming?
You rock.
You might be shocked to realize how much of what you have said and written consists of ideas that you have absorbed and repeated without original analysis.
Let’s start with language itself.
> >L: And all of us can break out of leeminghood for times. We can’t do it > > more than occasionally without the people in white coats taking us > > away. Been there, done that, by the way. > >-M: I hope that you were the one with the white coat. >I want to be permenently out of lemminghood.
You want to be dead. That’s what you’d be.
No, I was not wearing a white coat. Probably jeans.
> >L: Garbage in, garbage out. If the voter is assigned a proxy, then > > something provided by the system that assigns the proxy is adding > > weight to that proxy without any input from the voter. > >-M: In that case, its being added by other voters.
In which case the proxy is just added power for those other voters. It’s noise, taking one subset of the data that is missing and filling it in from what surrounds it. It makes for a seamless picture, but not for a full picture. It’s pretend data.
>With your option, where no proxy is provided, its still other voters’ >voice being provided instead(that is all that is counted).
Yes. So what’s the difference? Mark below acknowledges that he distorts the representation:
>The only difference is that my approach attempts to smooth the rank >distribution curve by giving bonus votes to underdogs. This makes the >rank structure less clustered and entrenched.
It makes it less representative.
>So SD2-S actually does something with the voter’s proxy power instead >of ignoring it.
It steals the voters proxy power and assigns it to someone.
The voter has power directly in standard DP and may choose to (1) exercise it, (2) delegate it, or (3) abstain.
SD2-S excises the first and last option. It reduces the freedom of the voter. On the argument that the voter is a “lemming.” (Or is probably a “lemming.”)
But true lemmings aren’t going to vote directly, most of the time. Indeed, the hardest thing is going to be to get them to join, unless most of their circle has joined. If they join, they will probably delegate or abstain, most of the time….
>L: This is not obvious to Mark because he does not respect the voter in > > the first place. > >-M: You are the one who wants to ignore the voter’s proxy power instead >of doing something constructive with it.
If I do something constructive with someone else’s voting power, I’ve stolen their vote. Unless, of course, they authorized me to do it. In other words, they named me as their proxy.
Very much I don’t want to see this done automatically. It defeats the major feature of Delegable Proxy, which is personal relationship and accountability.
Faagh! Enough.
I didn’t go over the rest of the message except for a very cursory glance. I’m sure there is some value in there, Mark is one of the relatively few people who have some understanding of delegable proxy. So if I’ve missed something that others think I should address, by all means, tell me. That is, let the others tell me, not Mark. Though I suppose he could briefly ask for a specific response to a specific question, if he has any.
—~—~————-~—~——~——————~———-~—~——~ You received this message because you are subscribed to the Google Groups “top-politics” group. To post to this group, send email to top-politics@googlegroups.com To unsubscribe from this group, send email to top-politics-unsubscribe@googlegroups.com For more options, visit this group at http://groups.google.com/group/top-politics?hl=en -~—————~——~——~——~———~——~———~—~—-
+1
-
Lomax wrote: >Mark wrote: > > >L: We are all lemmings, Mark. That you think you are not — and you do > > > think you are not — shows that you have bought your own exalted > > > image of yourself.
> >-M: I do have an exalted image of myself. > >I rock. How am I a lemming?
>L: You rock. You might be shocked to realize how much of what you have said and > written consists of ideas that you have absorbed and repeated without > original analysis. Let’s start with language itself.
-M: I think that I have thought deeply about all issues that I have discussed.
> > >L: And all of us can break out of leeminghood for times. We can’t do it > > > more than occasionally without the people in white coats taking us > > > away. Been there, done that, by the way.
> >-M: I hope that you were the one with the white coat. > >I want to be permenently out of lemminghood.
>L: You want to be dead. That’s what you’d be. No, I was not wearing a white coat. Probably jeans.
-M: I realize the group think is a very human characteristic, but I don’t always equate this with lemmingism.
Lemmingism is correlated with suicidal group think. Those who do group think within authoritative tradition aren’t necissarily lemmings.> > >L: Garbage in, garbage out. If the voter is assigned a proxy, then > > > something provided by the system that assigns the proxy is adding > > > weight to that proxy without any input from the voter.
> >-M: In that case, its being added by other voters.
>L: In which case the proxy is just added power for those other voters. > It’s noise, taking one subset of the data that is missing and filling > it in from what surrounds it. It makes for a seamless picture, but > not for a full picture. It’s pretend data.
-M:‘…not for a full picture…’ – compared with what? The other option is missing data, which is not full picture either. So I make a seamless picture instead of an incomplete picture.
> >With your option, where no proxy is provided, its still other voters’ > >voice being provided instead(that is all that is counted).
>L: Yes. So what’s the difference? Mark below acknowledges that he > distorts the representation:
-M: Yes, representation is distorted, but:
- distortion is the only option (it occurs with missing data too).
- representation isn’t the final goal, republicanism is.
> >The only difference is that my approach attempts to smooth the rank > >distribution curve by giving bonus votes to underdogs. This makes the > >rank structure less clustered and entrenched.
>L: It makes it less representative.
-M: The parallel is administrators and legislators appointing others below them.
More often than not these appointees do not have the votes of others. By contrast, the SD2-S underdogs so have significant votes of others, and the SD2-S system is giving them small and temporary amounts of bonus points to rise and challange potentially entrenched rank clusters.
Again, the goal of representation is a small goal, (but one that SD2-S does better than competing systems) compared to the primary goal of principled governance(republicanism).> >So SD2-S actually does something with the voter’s proxy power instead > >of ignoring it.
>L: It steals the voters proxy power and assigns it to someone.
-M: No, it takes freely given proxy power and makes constructive use of it.
>L: The voter has power directly in standard DP and may choose to (1) exercise it,…
-M: SD2-S does have this DD priciple, but it is for:
- limiting decisions for deliberation
- setting decision thresholds
>L: (2) delegate it…
-M: With SD2-S, delegation is manditory - the idea being that government is always done by others, therefore a voting system should be RD to reflect this fact. And proxy power that isn’t used isn’t merely ignored, but is reused to condition the voting data.
>L: …or (3) abstain. SD2-S excises the first and last option.
-M: SD2-S uses all these options, but tries to use abstained data instead of ignoring it.
>L: It reduces the freedom of the voter.
-M: No, with SD2-S, the voter has the utmost freedom. He/she is given close to 20 vote option combinations.
>L: On the argument that the voter is a “lemming.” (Or is probably a “lemming.”)
-M: The lemming argument is only to justify RD, not specificly SD2-S or its features.
>L: But true lemmings aren’t going to vote directly, most of the time. > Indeed, the hardest thing is going to be to get them to join, unless > most of their circle has joined. If they join, they will probably > delegate or abstain, most of the time….
-M: Any combination of lemmings and non-lemmings will render a:
- popular vote
- generalist vote
- specialist vote
SD2-S will see this, and the decision will be made by the specialist vote(default 60%) as allowed by a threshold of a popular vote (default 40%).
> >L: This is not obvious to Mark because he does not respect the voter in > > > the first place.
> >-M: You are the one who wants to ignore the voter’s proxy power instead > >of doing something constructive with it.
>L: If I do something constructive with someone else’s voting power, I’ve > stolen their vote. Unless, of course, they authorized me to do it. In > other words, they named me as their proxy.
-M: If they are using the voting system, they have authorized the voting system to do something with their missing proxy data.
>L: Very much I don’t want to see this done automatically. It defeats the > major feature of Delegable Proxy, which is personal relationship and > accountability.
-M: Again, Lomax, your ignoring of the missing proxy data still does something automaticly with it by having the effect of redistributing to those who have voted for proxies. By contrast, I do something more useful with it by challenging rank clusters.
>L: […] I’m sure there is some value in there, Mark is one of the > relatively few people who have some understanding of delegable proxy. > So if I’ve missed something that others think I should address, by > all means, tell me. That is, let the others tell me, not Mark. Though > I suppose he could briefly ask for a specific response to a specific > question, if he has any.
-M: How about just conceding to my points?
shanti
Mark, Seattle WA USA—~—~————-~—~——~——————~———-~—~——~ You received this message because you are subscribed to the Google Groups “top-politics” group. To post to this group, send email to top-politics@googlegroups.com To unsubscribe from this group, send email to top-politics-unsubscribe@googlegroups.com For more options, visit this group at http://groups.google.com/group/top-politics?hl=en -~—————~——~——~——~———~——~———~—~—-
+1
-
