Re: This is interesting

From: Keo Ormsby (chor02@xenomexico.org)
Date: Mon 13 Dec 2004 - 23:51:49 GMT

  • Next message: Van oost Kenneth: "Memetic Perspective"

    ----- Original Message ----- From: "Alan Patrick" <a.patrick@btinternet.com> To: <memetics@mmu.ac.uk>; <bafuture@yahoogroups.com> Sent: Monday, December 13, 2004 6:58 AM Subject: This is interesting http://www.wired.com/news/culture/0,1284,65317,00.html

    This is indeed very interesting. Let me yet again tiresomely ponder on and on, and risk an unreplied post.

    >From a traditional Western view of rules and laws, what the
    Southampton team did was "legal cheating", i.e. finding a loophole in the rules. This would be unremarkable (we see it happen in courts everyday) if it were not for the fact that here the rules were unbelievably simple, and seemingly would answer the "intended" question posed by the competition's organizers
    (namely, is tit-for-tat the best strategy for the iterated prisoner's dilemma?) If we stay with the traditional view of rules, the answer to this problem is to be very clever rule makers and make more rules that fix this (amendments). But I will bet you good money that someone will come up with another strategy that "bypasses" the original intention without actually breaking the rules. So we will end up with an arms race between rules and competitors, and hence in no time we will have a 1000 page rulebook for the iterative prisoner's dilemma. But what boggles my mind in this case is the inability to propose an airtight rule system even in a situation as simple as this one. What hope do we have for a reasonably good law system, if even scientists cannot come up with a simple solution to a simple rule problem? Lawmakers are notoriously biased and in many cases dim witted, so, are these the people that will write airtight corporate laws on complex financial systems? That the arms race has started a long time ago is obvious, just walk in any legal library. Most western people are proud of the intricacies and complexities of their legal system, and consider it civilized to have a rule for every single human act, human possession, or even animal or mineral matters. That we have gone from "thou shalt not steal" to thousands of pages of laws on fraud, money laundering, insider trading, etc. in my mind is not progress, its just testament of a poor strategy (an arms race) to achieve justice.

    So, before I get chastised by our moderator (just joking), let us go to the memes. In the competition case, from a memetic view it makes a lot a sense that this is happening. You define your system as comprised of entities (programs, strategies) that have to have a certain effect on the environment (made of other programs and judges), and must succeed (win in most matches). I will grant that this is not a strict memetic situation, since the favored strategy is not rewarded by replication, but by $50, but I believe an evolutionary approach is still warranted (just substitute the 50 clams for "leaves offspring"). The point is that as in all evolutionary systems, the end justifies the means, so obeying the spirit or intentions of the rules may be more, less or equally rewarded as loopholing (or even outright cheating). The important part here is the reward part. The organizers wanted to answer the question posed above, so they set a rule to themselves where they would give the prize to ANY strategy that WON the most number of matches without braking the RULES. Sound sensible, but we saw what happened. So if instead of making amendments to the RULES, you can change the definition of WINNING. If the reward is given to the strategy that best answers the question (the most illustrative), not the one that wins most matches, we are back on track. Of course this begs the question on how to evaluate the "illustrativeness", without human bias, but that is something I think scientists can work out easier, instead of devising more and more rules. A rule is a suboptimal straightforward partial solution to a game theory problem. If we see how evolutionary scientists try (albeit not as successfully as we would like) to solve problems in biology (epidemics, breeding, stopping extinction) without the ability to set rules
    (imagine something as ridiculous as: "by order of health authorities, it is forbidden for influenza viruses to mutate more than one base pair per 100 replications") we may gain insights on how to play the memetic game (life) fairly for all contestants.

    Keo Ormsby

    =============================================================== This was distributed via the memetics list associated with the Journal of Memetics - Evolutionary Models of Information Transmission For information about the journal and the list (e.g. unsubscribing) see: http://www.cpm.mmu.ac.uk/jom-emit



    This archive was generated by hypermail 2.1.5 : Tue 14 Dec 2004 - 00:06:28 GMT