Wednesday, February 11, 2009

Effective Human Network

Now, this guy from NYSE, the one with the Ponzi scheme, he never asked me to invest with him. Did he ask you? No one ever told me about his “system”, if he had one. Nobody tried to sell me. In fact, I suspect that none of his investors even know me. I suspect that none of them even know my name. I'm hurt. Oh, well, I guess I'm better off. ... Besides, I never knew who he was either.

But here's my question. What happened to the Kevin Bacon idea? I know who KB is. I don't think he knows who I am – even though I would be a whole lot more likely to invest with him than with the other guy. But the point is, according to common wisdom, we're connected. And still, he's not soliciting anything from me. He doesn't know me, he doesn't want to know me. He's not interested in my ideas, and sour grapes aside, the feeling is more or less mutual.

Lots of people know who KB is, and some even know every piece of public information that can be found about Kevin Bacon. He's a celebrity, above the crowd, but I have it on good faith that there are six (or maybe seven) degrees of separation, at most, between him and every other Hollywood actor, no matter how ordinary. We could broaden the terms of the connection game and make it between him and everyone else. Do we have to make it eight degrees to get everyone in? The game depends on the fact that connections between Hollywood people are part of the public domain, but for every publicly known connection, there may be thousands of anonymous connections.

Now, if he has kids, maybe they played soccer. My kids played soccer. Their coaches knew a bunch of grown-up soccer players and the league officials (and a bunch of sales reps I bet). Those guys probably know some state officials, who probably know some national officials, who probably know some officials in California, and so on. I'll bet there exists a four-link connection somewhere between my kids and KB's kids. Besides, I know the coaches myself, and I'll bet he does, too. That makes us almost buds. (Coincidence: I saw one of the coaches just this week crossing the street. I also learned a lot about him by seeing where he was. If I pass on that information, he will have no idea where it came from.)

Does Kevin Bacon go to church? Does he travel on commercial airlines? Does he know his congressman? I know a congressman. Hey, I'm not a particularly social person, but I'll bet there are lots of ways, pretty short connections, between me and Kevin Bacon. So, what's the deal? How come he's not hooking me up with a bunch of these Hollywood people?

There are probably short link-chains between just about everyone and everyone else. The world is pretty tightly connected, as the annual wave of flu epidemics should tell you. The flu bug, however, can take advantage of these connections in ways that you or I cannot. Jokes maybe, jokes and flu, that's all we share. The problem is that these networks have rules and filters that prevent and control the traffic. The reason for that is that many people are, like the flu bug, willing to abuse these networks for less than altruistic motives. Memes and viruses can have sharp teeth, and lots of people want to enlist you into their personal Ponzi schemes. If you have anything, or care about anything, then you have to protect yourself, ensconce yourself. So the tribal impulse constantly reasserts itself, which is not necessarily a bad thing. Nevertheless, if you want to solve big problems, the barriers are counterproductive.

Fred Smith, who delivers groceries to old ladies three days a week, has some interesting ideas that, believe it or not, would be useful to several congressmen, a certain epidemiologist, and even Bono. Fred has a creative mind but sort of puts people off with some of his outlandish malarkey. In fact, no one but the old ladies will listen to him any more, and they don't really understand him despite their best efforts. The problem is that the filters placed on Fred are pretty suffocating, unforgiving, and judgmental. Since Fred doesn't know how to explain himself, and no one who could explain him is taking the trouble, then those ideas are just going to have to wait until someone more respectable manages to think of them. I know -- innovation is 99 percent perspiration -- but it is also 90 percent luck and it appears in the oddest places. ... Maybe I could get Kevin to introduce him to Bono.

Fred is not the only one who gets marginalized. My daughter tells me about a young lady who invented and patented the windshield wiper. Her invention was never used, as such, and the implication seems to be that automobile makers could not believe or accept that a woman, much less a young woman, was capable of making such a contribution. Or maybe they just didn't like the idea of sending royalty checks to a woman. She wasn't inside their meme-walls. At any rate, we know what they should have done. They should have pursued the concept and the person. It was a good idea, and more importantly, an original mind is not usually limited to a single idea.

So here's my idea. We want networks that help us build collective intelligence and collective wisdom, that allow us to aggregate the strengths and knowledge of all, but that's not what we have now. Our networks are effective as peer-links. They are capable of passing certain types of information, especially among peers, but they are very poor at distributing trust. They are poor at concentrating veracity and intelligence; they are poor at recognizing the potential of specific pairings, and they are poor at producing enforceable collective decisions.

So I suggest that we redesign our networks to remediate those failings.


There are many organizations that are effective at concentrating power, which is both a strength and a failing. The caucus is one such organization. Despite great pains taken to prevent what they considered the evil of factionalism, the Framers of the Constitution were unable to prevent the development of groups in Congress from caucusing together. What happens is that members agree to hash out their differences and then vote as a bloc. The normal mode of independent voting would dilute their power by allowing minor differences to erode the collective impact. By committing to unanimous voting within the caucus, the members are insuring that their influence, and the influence of like-minded individuals, becomes predictable and potent.

To understand why the members of a caucus are willing to sacrifice their independence and even vote against their own interests, you have to understand that they are participating in a form of iterated prisoner's dilemma. There is a payoff for cooperating: I am with you guys because I trust you, and because my influence is stronger with you than on my own. I trust your judgment and I believe that, though I must sacrifice on occasion to vote with you, sometimes you will sacrifice to vote with me. If you deviate significantly from that covenant, though, I will punish you by removing myself, and though I might lose influence thereby, you will feel some pain yourselves, especially if I ally myself with your enemies. The pain you feel, or the threat of it, might well cause you to consider my positions and my situation more carefully.

I have written in the past about voters' unions. Somewhat like investment clubs, such groups ask each member to contribute research and opinions toward the collective goal of choosing a candidate. They caucus together. Once a decision has been made, the group acts as a unit, casting a unanimous vote.

Each state in the United States is actually a voters' union within the Electoral College. If a candidate wins by one vote within the state, it will be as if the entire state voted for a single candidate. It enhances the power of a state to influence the national election. But the idea can be generalized upward to groups of multiple states, or downward to groups of voters. States today do not cooperate in this fashion. And, except for special interest groups, neither do voters, but they could.

Some couples I know have a sort of voters' union. They argue together about the candidates. If they can agree, they both vote. If they can't agree, then neither of them votes, essentially saving them the trouble of getting to the polls. A group of three [I'm not supporting Big Love here] could settle things more efficaciously, but more coercively by resorting to majority rule rather than consensus. Since they pool their wisdom as well as their power, I would argue that such groups of three, or a hundred and three, are more likely to make the right decisions than a standalone voter, and more likely to influence others.

The catch is that official balloting is secret -- a very good thing in my estimation -- and that there is no way to ensure that the dissenters are cooperating. The conundrum is that small groups could benefit, in terms of power concentration, by committing themselves to a coercive voting protocol, but there seems no way to implement that regimen in the larger arena.


The Shiite and Sunni religious parties are feeling cheated over the January election results, not yet finalized. It was bound to happen. They have a perception of their popularity based on fear-elicited responses. Basically, sing our praises or we shoot you, or at least punish you. This is also why many managers are convinced that their employees love them. They hold power. The secret ballot removes the kind of leverage that such social power generates.

The first response of these parties, shocked by their loss of stature, will be to take control of the polling stations and consciously reinsert their influence. I expect they will be thwarted this time. Their second response will be to start lying about their motives and intentions. Their third response will be to concentrate on constituent services, at which point they will have raised themselves up to the level of Tammany Hall or Hezbollah in Lebanon, though perhaps not as influential. Getting beyond that is not easy. We tend to underestimate our own accomplishments in this regard.

The secret ballot is a boon to liberty but is insufficient for promoting trust. It allows you to vote for whomever your wish, but it doesn't protect you from making a bad choice. For one thing, there is still partisan gamesmanship, slate management, demagoguery and outright deception. The long and narrow conduit between the constituents and the leaders is a tempting target to everyone who seeks more influence than a rightful allocation would indicate. Here's the problem: you still do not know the people you are voting for. You do not know what they really intend because they are motivated to lie to you. Beyond that, they are motivated to avoid saying anything that might lose votes. We are consequently misinformed, disinformed and uninformed, which is why we always seem to be voting for the lesser of evils. We cannot distinguish the aberrant intentions of the power-hungry scoundrels from the rational political responses of the earnest public servant. Politicians routinely and necessarily cater to our ignorance and pander to our mistaken views. Since there is no trust and no truth, nothing gets fixed.

In order to increase trust in our elected officials, we need them to tell us things that we don't want to hear. We need them to not care about our vote. But we can only receive such messages from those we already trust implicitly, close associates, family and friends. Politicians can't, in general, have such a relationship with us, but we could in theory establish a personal connection chain from the retail voter to the power agent, in which each pairwise link is imbued with the assurance of personal trust. I might not trust the politician himself, but I trust someone who trusts someone who does. At the same time, I believe that such a chain could and should be sufficiently anonymous that no exterior force can be applied to the effect of distorting the information flow. A bucket brigade of trust and appropriate political deference could serve to solve problems that have been beyond us in the past.

So how can that be arranged? How can a forest of such trees be encouraged to grow? First, political distance can be minimized by using a pyramid scheme. Second, factual variance can be minimized by means of redundancy.

Pyramid schemes have a bad reputation because they have been typically used to distribute false hopes and illusion for the purpose of making money. That's because the pyramid recruits the soldiers rather than vice versa.

Consider the so-called Barack Obama 2.0. This program has a lot of cachet, but I don't believe it is any different from your standard corporate or military reporting structure. It's a pyramid, if not a scheme. Top down, tight span of control, one way flow of information, central command, leadership of each level selected by the higher levels. The lowest level members have virtually no influence, no information processing capacity, no caucus benefits. There is some redundancy, however, because the Internet allows for a lot of crosstalk. Nevertheless, it really should be called Barry's Battalion.

America 2.o

I maintain that information can flow both ways, authority can be delegated upward in a democratic and metered fashion, and trust can be nurtured as a rational outcome of a process rather than an emotional response to charismatic leadership. The secret is to have an upward pyramid. There are three requirements: 1) self-assembling units with regulated size and voting power, 2) multiple leaders elected at each level with instructions to represent the group at separate higher-level units. 3) the willingness of members to assimilate discrepant reports from the multiple representatives and coordinate a unified response.

What I'm proposing is essentially a fourth branch of government, a trust distribution network. In my vision, it works like this: People gather together in groups of choice, say twenty (20) people in each group. These would be friends, or members of a club or congregation -- any collection of folks who know each other and can evaluate each other's strengths and weaknesses. In the course of whatever activity they are engaged in, they also take time to select two (2) representatives on the basis of trustworthiness, reliability, honesty and whatever leadership qualities might be interesting. I think people can do this job pretty well. They know their friends. They might have sent Blogojevich to the State House, but he wouldn't stay on their bowling team for very long. Yes, there are exceptions, but for the most part, the groups will serve as a national filter to concentrate trustworthiness.

The next step is that the representatives should seek out other representatives of the same rank to form new groups at the next level up. Imagine growing neurons, dense and overlapping, each reaching its synapses out in hopes of synaptic intimacy.

One important rule is that two representatives from the same level one group should not be together in the same level two group. That's where the redundancy comes in. If there are enough groups, more and more levels can be built up until the whole country is part of the pyramid. Using the twenty-to-two ratio, each level is 10 percent the size of the previous level. We have, what ... eight degrees of separation between me and the top of the pyramid. Yes? But there is an important point to the rule of two. The length of the link-chain is a problem, but there are multiple link-chains that lead to the top, and each link pair represents a bond of trust, a relationship between two people who understand one another, who have a wideband communication interface. I think that could make all the difference.

There's a lot more that can be said about this pyramid scheme, but I'll let the reader digest what I've said here.

Game theory applied to Obama's Iran policy.

It turns out that Kevin Bacon was inside Bernie Madoff's meme-wall.


Labels: , , , , , , , , , ,

Links to this post

Links to this post:

Create a Link


At Wednesday, February 11, 2009 2:37:00 AM, Blogger jj mollo said...

Here's one of my comments on Voters' Unions. It's in response to a question about "the long tail of politics":

Posted by: jj mollo on September 30, 2006 9:40 PM

The long tail of politics is opinion. What do they say? Everyone has one. The algorithm of US politics is to aggregate these opinions into two piles from which the politicians nibble on the edges. So the edges are where the action is. It's not exactly clear who is the consumer and who is the producer. The long-tail product, I guess, is off-brand policies.

If you have a third party in the election, they often can matter, but only if they are willing to be flexible. Can you see Gore making Ralph Nader the Secretary of the Interior?

Beyond that, swing voters in early primaries and swing voters in swing states could have a influence if they were willing to join together. Voter unions, composed of such swing voters, could propose policies in exchange for bloc vote commitments. It would only work if there were some group discipline, as with a religious group perhaps, but the potential for exotic policy presentation is there.

Sounds like it was very late at night.

Here's a reference to voter unions by someone other than myself, in response to a post about increasing income inequality.

At Friday, February 13, 2009 10:32:00 PM, Blogger Steve said...

Good to see you're posting again JJ. I always enjoy reading what you write despite, or maybe because of, your much greater optimism compared to mine.

Years ago I got this notion that some genius could put together a national issues-debating, position-scoring and consensus-reporting system (I suppose it should have a name, so call it HOS (for Hal or Something)). HOS would be managed in such a way that its credibility was unquestioned and the population's trust in the system would be very high (unlike, say, Diebold voting machines).

HOS would be completely transparent. Open source if you like. HOS would allow anyone to check and see whether any issue they were concerned with had been discussed and studied, with their particular input represented to their satisfaction. If not, HOS would allow the citizen to submit it for proper treatment, and the citizen could be assured that their input was made a proper part of the emergent consensus (which is not to say that their input would necessarily sway the consensus, only that each of us had contributed to a rational process of consensus building).

Whatever else it would take, the trust that you call for would be indispensable. Trust would be a central design element of HOS, scientific, policed, independent. Maybe the fourth branch of government that you mentioned.

As with most things, the image in my mind included an empty box with the words, "Star Drive goes here." (That's from a priceless cartoon I didn't clip while I had the chance, possibly from a 20-year-old New Yorker.)

When you think about the conditions necessary to bring about the ideas you covered in your last two posts, what do you come up with? What would it take? I see society becoming more fragmented, and the edges harder, and I wonder how that fits with your thinking.

Cheers from Phoenix! Steve


Post a Comment

<< Home