[Altruistic-Economics] Building on AE framework

Wed Mar 23 14:42:45 PDT 2011

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 20/03/11 01:22, Robin Upton wrote:
> Long time no traffic on this list, so I'm forwarding a recent enquiry:
>> Date:     Fri, 18 Mar 2011
>> From:     John Preston
>> >> On the subject of AE, am I correct in thinking the basic idea is that
>> >> for each person, there are a set of `sympathy functions' (one for each
>> >> commodity), each of which defines how the resource gain by the receiving
>> >> party varies with the resource loss of the giving party?
>> > Sympathy is one the most important examples of the general item which is
>> > exchange rates,
>> > all presumed to be piecewise linear functions. The other major use is to
>> > express capability
>> > (e.g. can convert wood and time into chairs etc.) The sympathy functions
>> > define interpersonal
>> > resource exchange. e.g. I am prepared to forego X so that you can have Y.
>> It seems to me that in some respects AE is a higher-level abstraction of
>> traditional economics, with an emphasis on mutual satisfaction. However,
>> instead of promoting a particular economic model, may I propose an
>> (higher-level and slightly more distant) alternative? A network of
>> persons (i.e. individuals, collectives, companies, etc.), linked by
>> their satisfaction with historical transactions. For example, a computer
>> hardware manufacturer (H) and a computer retailer (R) might be two
>> persons. If R wishes to obtain hardware from the H, R must approach H
>> and perform some kind of transaction. They will either come to some form
>> of agreement, or the transaction will fail. If the transaction succeeds,
>> then each will have a level of satisfaction with the transaction (e.g. a
>> score out of 10), and this can be made available publicly (without
>> exposing any information about what the transaction was for, etc.) in
>> the form of a network (a graph where vertices are people and
>> satisfaction is edge weighting) connecting the two people. As more
>> transactions occur over time, the overall satisfaction between the
>> parties (e.g. it could be computed as the mean average of all
>> transaction scores) will change a little, but will probably stay
>> more-or-less the same. From this information, we know the compatibility
>> between two parties when they perform transactions, and this information
>> is completely separate from how the transactions take place. The parties
>> may reach their agreement by bartering, exchange for currency, gifting,
>> etc. each one causing a different level of satisfaction for each party
>> (altruism would be the desire for someone to maximise the satisfaction
>> at both their end and the other person's end). Suppose now some other
>> person comes involved, let's say an individual consumer Alice (A). A can
>> either obtain hardware from H or from R, and let's say she wants to
>> maximise her own transaction satisfaction. A has a friend Bob (B) who
>> has previously obtained hardware from both H and R. A and B have also
>> had transactions in the past, and have a high transaction compatibility
>> (i.e. both have high satisfaction scores with respect to each other),
>> and so it is likely that A will have good transactions with the people
>> that B has had good transactions with (because as their transaction
>> compatibility is high, they probably have similar economic thinking).
>> Therefore, A can use the transaction satisfaction of her compatible
>> neighbours, and their compatible neighbours, and so on, to find people
>> with which to have optimal transactions. In a network before any
>> transactions happen, the graph will be completely disconnected, and it
>> would build up from there by people trying transactions with strangers.
>> Naturally the network would be in machine-readable format, so that
>> people can search for desired goods/services and order the results by
>> estimated compatibility (for strangers) and by overall satisfaction (for
>> `friends', defined as people you have had at least one transaction with).
> This would be one example of something that could be built on top of AE,
> couldn't it?
Rather, I'd say that AE is an economic model that can be used on top of
this (as this is a technique for evaluating past transactions), although
there are some ideas in AE that could be used in the construction of my
suggestion (is everyone OK with calling it ``Historical Transaction
Satisfaction Evaluation'' or HTSE?)
> I refer to this as 'metaevaluation' in http://altruists.org/ae12 for
> more on this topic.
> I'm assuming that a simple but flexible standard would exist as the
> bottom layer,
> that allowed people to do different things with it to suit their styles.
This is where I'm thinking AE could help, as it has a lot of reusable
implementation.
> 
> AE10 has more on that standard, which is basically just a set of (units,
> amount) pairs,
> so that people can express their feelings as a basket of currencies - no
> need to
> lump apples with oranges if people define them separately.
People should be able to make available as much (or as little)
information as desired. The cool thing is that we can say that any large
transaction can be broken down into its smaller components (e.g. into
two transactions, one of time and one of money), each with their own
satisfaction ratings. Therefore I think the most we need to do is
provide support for one transactions in the form of a 5-tuple: (donated
resource type, donated resource amount, received resource type, received
resource amount, satisfaction). Satisfaction is the only mandatory
field, resource types are arbitrary and are chosen by the user (e.g. it
may be time, money, food, televisions, etc.).
> 
> Robin
> 
> 
> _______________________________________________
> Altruistic-Economics mailing list
> Altruistic-Economics at lists.gifteconomy.org
> http://lists.gifteconomy.org/listinfo.cgi/altruistic-economics-gifteconomy.org

- -- 
Yours Sincerely,
John Preston

Please note, that this message may be signed and/or encrypted with PGP
(available from http://www.pgp.com/) or GPG (available from
http://www.gnupg.org/). You can locate my PGP/GPG Public Key on my
website at:

http://geek.shell.la/~gizmoguy/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
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=14dy
-----END PGP SIGNATURE-----