Re: FAI: Collective Volition

From: Metaqualia (metaqualia@mynichi.com)
Date: Tue Jun 01 2004 - 12:36:37 MDT


----- Original Message -----
From: "Eliezer Yudkowsky" <sentience@pobox.com>
To: <sl4@sl4.org>
Sent: Wednesday, June 02, 2004 1:26 AM
Subject: Re: FAI: Collective Volition

> Norm Wilson wrote:
>
> > Eliezer Yudkowsky wrote:
> >
> >> Self-determination is not the only criterion of an acceptable outcome.
> >
> > Is self-determination a criterion of an acceptable outcome? Sounds like
> > speculation about the outcome to me :)
>
> True enough. Sigh. That's the problem with allowing myself speculation,
> even if it isn't at the expense of helping; it's too easy to lose track of
> the distinction between that and reality.
>
> >> It seems to me that an FAI can make huge improvements to background
> >> rules before that starts interfering with self-determination.
> >
> > Who are we to assume that an FAI will value our self-determination at
> > all, unless we make it an invariant?
>
> Conceded, and thanks for reminding me.
>
> > For that matter, the idea that collective volition is a correct path to
> > morality is arguably itself a moral judgment. By making it an
> > invariant, you run the risk of imposing moral content on the FAI?s
> > morality-seeking structure.
> >
> > While I like the idea of specifying friendliness "structure" over its
> > "content", I think it will be difficult to introduce invariants that
> > don't themselves have implicit moral content.
>
> There's a *lot* of implicit moral content in collective volition - why,
> half the paper is about the implicit moral content of collective volition!
> How else would I select that solution from solution space? The point is
> that it's rewritable moral content if the moral content is not what we
> want, which I view as an important moral point; that it gives humanity a
> vote rather than just me, which is another important moral point to me
> personally; and so on.
>
> --
> Eliezer S. Yudkowsky http://intelligence.org/
> Research Fellow, Singularity Institute for Artificial Intelligence
>
>



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:47 MDT