[ExI] size of polities

Lee Corbin lcorbin at rawbw.com
Tue Feb 19 04:53:16 UTC 2008


Jef writes

> [Lee wrote:]
>>>> One obvious idea that emerges from this is that most nations
>>>> are too large to properly reflect the desires of an individual
>>>> citizen.
> 
>>> A broader view might restate your assessment in terms of information
>>> technology..., recognizing that while historically "obvious" that a nation
>>> could not effectively represent, let alone model, the fine-grained
>>> values-complexes of its constituents, we have now at hand [a new]
>>> technology...
>>
>> Well, yes, but this perhaps does not address the main point that
>> no matter how accurately we aggregate desires or values, the
>> larger the number of inputs, the further will be the expected
>> outcome from the desires or values of any particular person.
> 
> Funny, I thought I was addressing just that point, placing your claim
> within a broader context of increasingly effective technologies
> modeling values over increasing scale, and further that it is
> characteristic of this technological trend that our values are modeled
> in ever **finer** detail, such that salient features are preserved and
> available to influence decision-making over increasing scope, rather
> than being lost in the aggregate, as was historically (but not
> fundamentally) the case.

Bringing this down to the concrete, perhaps an example would
help.  You see, I'm concerned that any kind of *aggregation*
of desires---which is what we see in conventional elections
today---results *necessarily* in less custom-tailored effects
(solutions) pertaining to an individual.

Okay, thanks to the following explanation you provide next,
I gather now that you *are* interested in dispensing individually
tailored effects:

> I mentioned amazon, lastfm, pandora {.com} as present examples of
> increasingly fine-grained knowledge of individual values-complexes,
> over much greater than individual scale, facilitating prediction of
> future states likely to be found desirable despite being presently
> unknown and therefore undesired.
> 
> In case this still isn't clear, please note that the nature of these
> information-centric processes is that the knowledge of the system is
> --and must be -- greater than the knowledge of any of the individuals
> it serves, and that with increasing scale these systems deliver
> increasingly effective personalization.

But recall that the general context of the discussion heretofore
was how laws, for example, fail to take into account local
information. Perhaps you are suggesting that an incredibly
complicated law could apply differently to each individual.
Well---it would be much, much simpler to have smaller
polities, and let freedom rule.

Lee

> Perhaps I should have been more explicit in the extension of this
> thinking from its (to be expected) roots in entertainment and commerce
> to virtually all domains of social choice.  In this regard I thought
> my final paragraph, on the power of prediction markets, and their
> moral meta-benefit, would have been sufficient.




More information about the extropy-chat mailing list