From: Jeff Bone (firstname.lastname@example.org)
Date: Sat Dec 08 2001 - 17:15:13 MST
"Eliezer S. Yudkowsky" wrote:
> Perhaps. Understand that I do not have "faith" that immortality is
> possible. I am simply stating that before we get all emotional
Hey now... I may be stating things forcefully, but I'm pretty emotionally detached
from this whole thing. :-)
> about this
> issue - that is, before we begin making value judgements or philosophical
> assumptions based on it -
Let's understand that there's a difference between those things. IMO, the issue of
"value judgements" is a huge one, but from my perspective it's the notion (as it's
been advanced so far) of Friendliness that's entangled with a priori value
assessments. Philosophical assumptions, well, in order to argue logically you've got
to ground out in some set of assumptions. I try to make that set as small as I can
in my arguments, but it's a challenge when dealing with issues this broad.
> we should remember that the model the prediction
> is based on is a model which historically has often changed and currently
> is still in flux.
Yawn. That's a very anti-scientific argument, Eli --- indeed one I often hear
religious folks (like the Creationist crowd) use to undermine the credibility of
arguments from a scientific bias. Let's keep in mind that the model that we have now
--- while *certain* to change and improve --- is now yielding very fine-grained
predictions with accuracy improving at a rate that outstrips our ability to verify
them. Those that we can measure are turning out to be very, very precise. (I'm
speaking specifically about some of the predictions coming out of QED.) That aside,
it's the best model we've got to make predictions from --- and given that all this
stuff is highly predictive, we should use the best predictive models and
understanding that we have.
> Well... this is something I happen to disagree with, because personally
> the second law of thermodynamics strikes me as being statistical in
> nature, and often rather fragile.
Go read up on it --- it's probably the best law we've got, and has very, very deep
implications for the most essential workings of spacetime. Yes, it's statistical,
and yes, there are issues with scope (i.e., clearly life violates 2LT on a very local
scale, but then the law doesn't talk about those scales.) But it's still got some
amazing properties relative to many other "fundamental" physical laws.
> I would not be surprised to find out
> that it is utterly impossible to exceed the speed of light within a given
> gravitational frame of reference, forever and amen. 2LT seems to me to
> have more the character of a guideline than a rule.
2LT doesn't predict the impossibility of exceeding c. And IMO, you're right in your
bias: I'm not convinced of the inviolability of c --- or even its constancy. OTOH,
only 2LT gives reason to many of the other assymetries and observed peculiarities in
our reality, and we have no better explanation of one very real phenomenon: the
directionality of time.
Go read _Order Out Of Chaos_ by Ilya Prigogine. It's a good (if pop) introduction to
the topic from back in the mid-80s, by a Nobel prizewinner (and local-to-me --- he
lives in Austin too!)
Aside: I find it amusing and appropriate that you should take issue with the
validity of 2LT, given the Maxwell's Daemon-like nature of your Sysop in some
possible interpretations. :-)
> We can never prove (epistemological impossibility) that any given
> > set of laws, no matter how accurate the predictions they yield, are the best or
> > even only model of how the world works.
> I think perhaps our species is too young to know that as well.
Really? You're really close to arguing that logic itself is tenuous, and other
similar epistemological problems -wrt- the nature of knowledge. There are some
things that are either knowable or unknowable independent of the characteristics of
the knower, and there are things that can be known to be knowable without actually
knowing them. This isn't a matter of philosophy, it's a matter of math and logic ---
and unless you're up for taking on some of the brightest minds in those fields then
there's no argument to be had. (BTW, these aren't my arguments, rather me
paraphrasing and synthesizing Curry, Church, Godel, Turing, Russell, and a host of
> I can
> conceive of a model which starts at the First Cause
What if there is no First Cause? Many currently fashionable reinterpretations of the
origin of the universe actually imply that there is no timelike lower bound to the
universe, i.e. that timelike boundary is asymptotic, and there is no initial causal
> > IMO, any system of "rights" in practice actually results in unresolvable
> > inconsistencies and paradoxes.
> "Unresolvable?" That sounds pretty strong. Can you name a single
> unresolvable inconsistency or paradox?
Sure. Alice and Bob both have a "right to life." For whatever reason, a situation
that arises in which Alice can persist but Bob must perish, OR Bob can persist but
Alice must perish. (Imagine that a Sysop is charged with guarding the physical
inhabitants of some region which is facing some catastrophe, and time or resource
constraints make it impossible to save everyone in peril.) More mundanely, your wife
and your daughter are crossing the street, unaware of an oncoming bus --- you have
time to push one or the other out of the way but not both. More generally, if two
actors have the right to pursue happiness, but happiness for one results in inability
to pursue happiness for the other, then you've got an inconsistency.[Image]
Etc. etc. these are of course contrived, but it's generalizable and formalizable,
and if it's formalizable it's decidable. If we want to go down that path, I can give
you the formal notation. I cannot (yet) prove that *any* system of rights results in
unresolvable inconsistencies, but I can prove that *many* (and common) systems of
assumed rights result in unresolvable inconsistencies. My conclusion from this is
that rights are not and cannot be considered to be innate, and that rights are not
absolute but must be contextual within some system of values.
> > It may be that the optimal system for allowing independent actors to achieve
> > optimal balance of competing self-interests is not a system of axiomatized
> > rights coupled with protective and punitive measures (a "legal" system, or a
> > Sysop) but rather a kind of metalegal framework that enables efficient
> > negotiation and exchange of consensual, contractual agreements.
> The two main problems with this are as stated earlier: First, the
> possibility of a universe in which offense beats defense; second, the fact
> that a simulated, enslaved citizen has no position from which to
Two problems with the problems: first, you can't *ever* prove that any given defense
against some threat is infallible; there's an old logical argument to that effect,
and even more recent quantitative defense analyses that underscore and even reinforce
the principle. Second, how did simulated, enslaved citizens work their way into the
argument? Nothing I said requires or even implies that.
> A metalegal framework might be superimposed on an intelligent
> substrate, but it still requires an intelligent substrate to ensure that
> no being is stripped of citizenship rights.
Again, rights are a very superstitious, nebulous, and anthropomorphic concept which
*may* in fact *impair* civilization's ability to achieve desirable outcomes such as
longevity or Pareto optimal happiness of individuals.
> > > Eliminate negative events if possible.
> > But "negative" has many dimensions, and most of those are subjective...
> ... he argued, correctly assuming that the audience would perceive
> "subjectivity" as a negative quality with respect to principles intended
> for a Friendly AI.
Not really my assumption, but it doesn't surprise me that you would assume that. ;-)
Believe it or not and dispite your skepticism of philosophers, Eli, this kind of
discourse and moral calculus does indeed exist outside the scope of the discussion of
Friendliness. :-) You might want to check out  below for some contemporary
discussion along these lines. Bit of a tangent and not necessarily an argument I
agree with, but interesting nonetheless.
BTW, this discussion doesn't offend me at all, and indeed there's a long tradition
along these lines, going back at least to the Greeks. It should be understood that
I'm merely arguing along the lines of Epicurus, liberally seasoned with Lysander
 The Dialectical Necessity of Morality : An Analysis and Defense of Alan Gewirth's
Argument to the Principle of Generic Consistency by Deryck Beyleveld, Alan Gewirth
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:37 MDT