X-Message-Number: 9546
From:  (John P. Pietrzak)
Newsgroups: sci.cryonics
Subject: Re: Pietrzak 6
Date: Sat, 25 Apr 1998 15:48:31 GMT
Message-ID: <>
References: <>

Hi again.  I should apologize for dropping out of the discussion for a
few days, haven't had the time to devote to the internet lately that
I'd been used to.  (And, I should apologize in advance for the length
of this post, I've gone on too far again.)

On 21 Apr 1998 15:40:40 GMT,  (Ettinger) wrote:

>2. An important point is that Mr. Pietrzak uses "intuition" in a misleading
>way. [...] Actually, there is an extremely important intermediate case, where
>there is no simple or readily available explicit statistical record, yet the
>facts are so clear that most people will be in reasonably close agreement.

Maybe.  I generally find that, in many cases, people in close
agreement on a particular subject are, in fact, not in such close
agreement when one presses the issue to the point of asking "why" they
agree about it.


>An example is the exponential life parameter which I developed in the essay. It

>employs the "intuition" that moderate values of expected life are a priori more
>likely than extreme values. Who could deny that this is true and thoroughly
>grounded in experience? Who could deny that it is useful information?

I could.  Certainly, the "bathtub" formula is well established in both
industry and biology: the extremes of youth and age are subject to
greater mortality rates than those of "moderate" values (leading to
the distinctive "bathtub-shaped" curve of the function).  (In fact,
there are usually plenty of available explicit datapoints to use, no
need to rely on intuition for that one.)

But, why is this the case?  In industry, there are generally only two
factors involved, errors made when the part was generated (causing the
early failures) and the long-term wear of the part (causing the later
failures).  (Biological lifespan is, of course, much more complex.)
If, for example, the errors dominate here (never allowing the part to
reach the point of wearing out), your intuitive concept of the likely
lifespan will in fact bias your calculation _away_ from the actual
truth!

>In the same context, Mr. Pietrzak uses "objective/subjective" in a misleading

>way. He calls the example above (a priori information about the life parameter)
>"subjective," but, as explained above, it is perfectly objective in the sense
>that it is firmly rooted in experience and that different observers are likely
>to have similar interpretations.

It is not my intention to mislead.  By objective, I mean "data not
relying upon any given person", and subjective therefore being "data
relying upon some person".  You may believe that everyone else shares
your own opinions about the world around you, but I don't.  For me, I
cannot identify a particular element of data as objective unless it
can be verified without relying upon _anyone's_ past experiences.

>3. I also noted that Bayes is more objective than (say) Fisher, in the sense
>that it does not ignore some useful and available data, even though the data
>may be imprecise. If you don't like the word "objective" in this context, then
>substitute "scientific." Certainly someone who looks at the whole picture, and
>factors in everything relevant, is being more scientific than someone who
>doesn't.

(Actually, I always thought that science involved _not_ looking at the
whole picture, instead stripping everything else away to get at the
truth of a particular concept.  You avoid having something not
associated with what you are looking for polluting your results.  This
is generally why controls are used in experiments, I believe.)

>4.  Am I appealing to "modus ponens" or to my general approach to probability?
>To both, of course. Modus ponens appears at some point in almost every
>argument, but is rarely labeled such, because it is so routine.

That's not what I was saying.  You didn't seem to be using anything
_beyond_ modus ponens.  You could ignore the rest of your argument, if
all you really want to do is rely upon that and a series of intuitive
statements.  (Where, in fact, do you even use Bayes or anything else
in the conclusion of your paper?  I assumed you were alluding to them
implicitly, but your argument may not use them at all...)

>The key point, as Mr. Pietrzak says, is whether experience justifies the

>statement that "Goals of technology have a good chance of success." This is not
>a "different" argument, but the main line and (I thought) clearly labeled
>argument.

Yes, I did have this wrong to begin with.  I had originally assumed,
the logical way of generating a probability for success with cryonics
was to have a series of cryonics experiments and use them as
datapoints.  As with the "exponential life parameter", I had assumed
that your intuitive beliefs would bias a function over _actual
cryonics data_.  But your paper requires (and contains) no information
about cryonics at all (beyond the implicit characterization of being
similar to other goals of science/technology).

Obviously, this is what has got my hackles up.  I could substitute
practically any other word for the word "cryonics" and your paper
would, presumably, have the same results.

>Naturally, we must be careful (a) in characterizing "projects of technology"
>and (b) in deciding whether cryo-repair fits in reasonably well.
>
>As to (a), the project should be one which at some stage was thought so
>difficult as to be either impossible or forever impractical; [...]
>It should also be "serious" in that it does not violate "known" laws of
>nature and has at least a few proponents with respectable credentials.
>[...] Finally, the "project" should be one of ends rather than means--for
>example, to build a flying machine, not necessarily to build one that employs
>a specific means, such as flapping wings [...]
>
>As to (b), cryo-repair is certainly serious, and we do not much care whether
>the successful means will involve full-fledged nanotech, nearer term chemistry
>and biology, or a pragmatic mix of approaches.

(As to that final point of (a): what if, for example, the means end up
involving something no-one has thought of yet?  If the repair involves
something you haven't taken into account (say, a full-body molecular-
level scan stored for later use in disentangling the various bits of
stuff in the frozen body) wouldn't this have an affect on your
calculations?)

[On present-day nanotech w.r.t. future cryo-repair]
>This is not a perfect argument, but on the whole one must agree that we are
>more likely to achieve success in an ongoing project than in one not yet even
>conceived.
[and]
>Remember also, once more, that we refer to the likelihood of EVENTUAL success.

>A "failed" project means one which not only has not yet succeeded, but has been
>proven to be almost certainly unachievable.

So, I'm curious.  Exactly what, to you, would be an "unachievable"
goal?  And do you at all care about how long the process takes?

---------------------------------------

Let me close with a statement on time.

I do believe that, eventually, suspended animation will be achievable.
But it will take time to achieve.  Even if frozen, people cannot wait
forever.  Thus, in my opinion, this must be made an explicit element
of any measure of probable success.

Your "exponential life parameter" is, in fact, a perfect example here.
Why do you suppose that so many people have the intuition that a
"moderate" lifespan is to be biased so heavily?  There are costs
associated with mere existence, and these make themselves known over
time.  If it takes too long to figure out how to get those people out
of their frozen state, the nasty end of this function is going to
reveal itself.  (Of course, I'm only using my intuition here, but
given what you've stated previously, I'm sure you'd agree.)

If time is, as I believe it to be, the major obstacle barring any of
us from participating successfully in the use of suspended animation,
then I argue that it is to time that we must focus our attention.  If
the time it takes for reversible cryonics to be perfected is the most
significant factor that a person in suspended animation will have to
wait for, then it is the reversing of cryonic storage that must be
achieved with all possible speed.  That is where our resources should
be applied.  Regardless of what one thinks of the eventual probability
of success of cryonics, this is (IMO) the best way to affect that
chance in a positive manner.


John

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=9546