From: Strata Rose Chalup (strata@virtual.net)
Date: Wed Aug 16 2000 - 20:50:02 PDT
Brian Atkins wrote:
> [...]
> I don't want to shoot you, but I want you to read this:
>
> http://www.singinst.org/tmol-faq/meaningoflife.html
>
> and give a response.
Brian,
Wow.
Whoever wrote that FAQ is obviously a deeply passionate and committed
individual who wants very much to help some combination of
humanity/theworld/theuniverse. They are also obviously letting their
enthusiasm for the process keep them from digging into the memetic
affordances expressed in the document. The future they would build has
enough elements left to chance, and enough stacked *against* an outcome
that most people, even transhumanists, would consider "benevolent", that
I am highly concerned.
If I were to drop everything else for an entire weekend and do nothing
but annotate the FAQ document, that would be a good start at
communicating some of the issues. I won't have time to tackle this for
at least 4 weeks without neglecting other things. My estimate of the
likelihood of "results" coming to fruition for the project within the
next few weeks is low, based on what I know about the project. If I'm
incorrect, I hope Brian or someone else will tell me, and I will bump
the priority WAY WAY up.
If this is a serious project, and has access to serious funds and
resources, it is (IMHO) begging for some serious intervention.
If you would like to start without me, here is a suggested roadmap.
Please start with a read of _The Structure of Magic_ by Bandler &
Grinder, followed by _Patterns 1 & 2_, ditto. Go back and analyze the
FAQ, especially the section on goals.
Analyze the primary modalities of the document, and identify the
modality from which the writer is primarily speaking.
Reason: there will be value in determining which physical sensory mode
seems to predominate in setting the values in the document, whether the
AI builders choose to deliberately ignore or to deliberately promote
setting similar structures in the AI mind.
Determine whether the modality shifts, either seemingly randomly, or
predictably, when describing the various pieces of the project.
Reason: as previous, plus assistance in identifying parts of the
document that may have different authors or ideas originating outside
the author (he/she will want to check to make sure he/she is aware of
such antecedents), plus data on shifting modalities consistent with
topic is statistically correlated with emotional import for most authors
(he/she will want to be aware of places where he/she is presenting
technical information from a place of emotional response).
Re-map each sentence to fill in any places where concepts have been
elided or indirect referents used, and identify the place where the most
likely referent appears in the text.
Reason: identify any conflating of ideas and factually or technically
inaccurate borrowing of affordances between ideas/concepts which are
only semantically related by sentence structure and have little or no
relation in the outside world, plus change in form makes it easier to
spot simple logic errors, most common of which in this kind of writing
is "Joe is from Rome, Romans are from Rome, therefore Joe is a Roman".
Read or skim Lakoff & Johnson's _Metaphors We Live By_ in order to
sharpen awareness of some of the hidden contexts that may be in use.
Create an index of annotations of comparators used as if they were
expanded, e.g. "better" without a definition of "better than what".
Reason: further identification of idea/concept conflation.
Recurse on that index and identify imprecise or vague definitions in
who/what/where/when/why/how space, e.g., "better than the current
condition" without specifying in what way.
Reason: identification of value set of author being implicitly taken to
be value set of reader/goalset of reader/both of project; may match up
when everyone has a chance to look, but needs to be laid out carefully
so that it is clear what is being proposed and how it is proposed to be
accomplished.
Recurse on that index and identify a scale, context, or fully described
value-set for things which seem to be describing variable quantities
within a framework, e.g., "having more freedom than America in the late
1990's" -> "having more ability to speak out, change jobs, travel within
and outside of state & country borders, [...]".
Reason: clarification of proposed value set.
When elaborating possibilities, if all evaluators adopt the convention
of ordering choices by their individual beliefs about what may be meant,
analysis of the annotations can help the evaluators understand where
their assumption and belief/goal sets mesh and where they are disjoint,
as well as clarifying differences in vocabulary usage.
Very simple example:
Did I mention that doing this right would take a lot of work?
Transforms to:
Did I mention that [[this] -> [analyzing the FAQ at
http://www.singinst.org/tmol-faq/] || [creating an AI without
guaranteeing an outcome seen as non-benevolent by Strata Rose Chalup]]
[[right] -> [according to personal standards of correctness of Strata
Rose Chalup] || [producing a benevolent outcome] || [carrying out
procedures according to a commonly agreed upon (need WHO referent)
standard of scientific exploration principles] || [in a way that will
satisfy the author of the FAQ] || [according to personal standards of
goal-desirability of the author of the FAQ]] would [[take] -> [require
from participants]] a [[lot] -> [a minimum of 100 man-hours if outcome
is FAQ analysis] || {much clarification needed!}] of [[work] ->
[[effort] -> [reading, writing, informal discussions via a range of
communications media, study of related works, research to discover
related works]]] [[?] -> [statement, rhetorically asked as a question]]
That is a simple example, and would benefit from a lot more thought on
my part but I really need to get back to work.
My belief is that the Singularity Institute has the potential to help
bring about a singularity, but I see a miminum of evidence (in the
online materials that I have examined) that the preceptual framework
which they are in the process of constructing would NOT lead to some
type of event, singularity or merely something very noticeable, that
would adversely affect myself and various aspects of the physical and
cultural environment in which I enjoy living.
Phil Agre has spent a lot of time doing analyses of technological (and
other) writing, and would probably consider my methods above as stone
knives and bearskins. I encourage you to contact him about the
Singularity Institute, but have no idea whether he would be interested.
_SRC
PS- This is a prime example, BTW, of the kind of thing I mean about
wishing that the Universe had me on retainer of some sort. I doesn't
have to be me, for sure. I selfishly wish that there was someone
involved with the project who is sufficiently "me-like" for me to feel
reassured. I don't mean that as an arrogant statement, but I don't know
quite how else to phrase it. I don't think that the Universe requires
my intervention in the Singularity Institute to keep functioning for a
majority of humanity, I just think that things might suck for ME if they
get where they are pointed to go based on what I read online! If that
makes the rest of the Universe happy, by accident or choice, well,
that's just tough luck for me. I hope I'd be a nice enough person to
take the hit if it seemed that way.
PPS- I do this kind of thing in microcosm as part of Project Management,
and it makes most projects go a LOT more smoothly with a lot less time
spent chasing tails/butting heads due to mutual confusion with the team,
the manager(s), and the client(s).
-- ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Strata Rose Chalup [strata@knownow.com] | strata@virtual.net, KF6NBZ Director of Network Operations | VirtualNet Consulting KnowNow, Inc [http://www.knownow.com] | http://www.virtual.net/ ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
This archive was generated by hypermail 2b29 : Wed Aug 16 2000 - 20:50:28 PDT