X-Message-Number: 13317
From: 
Date: Sun, 27 Feb 2000 10:58:48 EST
Subject: probability, religion,research

This stuff is being recycled ad nauseam, but we always have the newcomers to 
consider, as well as those whose memories need refreshing. So once more into 
the breach.

Very roughly, there is a pro-cryonics camp holding that, while use of present 
resources for cryobiological research is important, it is not important to 
the exclusion of other immediate goals.

Then there is a semi-pro-cryonics camp holding that most current resources 
should go into cryobiological research, often saying or implying that 
cryopreservation by past or current methods is nearly hopeless. Some of these 
people, at least some of the time, heap scorn on "blissfully ignorant 
optimists" and even stigmatize optimism about nanotechnology as "religion." 
Today Thomas Donaldson used the "R" word again, even though he is closer to 
being pro than semi-pro.

Today's reminder is about the foundations of probability theory, concerning 
which most people, including most mathematicians, are amazingly ignorant. On 
our web site I have a long discussion of probability and cryonics, which of 
course will be persuasive to few,  but nevertheless is available to anyone 
who is serious about these issues.

The basic premise of probability theory is that what we see is (probably) 
typical of what we don't see. The sample is typical of the population. There 
is no alternative. Larger samples reduce the uncertainty, but the uncertainty 
is always present; yet as long as there is a sample, ignorance is not total.

It is also essential to understand that uncertainty is in the beholder, not 
the world. (I omit a long digression into quantum uncertainties.) The 
probability of an event is the RECORDED ratio of successes to trials in a 
FINITE sequence of reasonably well defined experiments. (I omit a long 
digression into probabilities of probabilities, or second-order 
probabilities.) 

A crucial point is that the experiments in the sequence are never perfectly 
defined, and probabilities are never exact. Yet the theory applies equally to 
a long series of well defined experiments and to a short series of vaguely 
defined experiments. In the latter case the results are less precise and less 
reliable, but no different in kind.

So: Consider two extreme cases.

1. Toss a coin. The known sequence is very long (even though not recorded in 
any one database), and the experiment is rather well defined (even though, if 
someone wanted to, he could train himself to bias the toss). So we get high 
precision and reliability; the probability of getting heads on the next toss 
is  , plus or minus a very small number. 

Important note: The same reasoning would apply to any symmetrical body and 
any random shove, even in the total absence of directly relevant history. It 
would apply, say, to the toss of a 12-sided die, even if such a thing had 
never existed before. We MUST acknowedge the validity of CATEGORIES of 
experiments. 

2. Guess the future of nanotechnology. Here the sequence of experiments is 
poorly defined, but not vacuous. It could be assigned, for example, to the 
category of experiments, "Development of Modern Technologies." Very vague and 
amorphous, but NOT vacuous.

Now one could assemble statistics in many ways. For example, one could ask, 
how often, in the early stages of a technology, have optimists and pessimists 
made predictions, and how often have each been right at later stages? Many 
variations are possible, but history tells us that, in most cases, the 
progress surprises the pessimists. Progress as to goals or ends, that is; it 
is not unusual for particular means to fail. 

Yet again, the big picture is often the only useful one in the longer run. In 
the early days of speculation about rocketry, pessimism was almost totally 
widespread among "experts" and based on their intimate knowledge of the 
difficulties in the details. Optimism among experts was based on 
fundamentals, with details left for later. Optimism among amateurs was based 
on supposedly na ve recognition of the inexorable advance of technology; but 
in fact this alleged naivete was at bottom a thoroughly scientific 
recognition of the sweep of history, or the application of probability theory 
to areas usually ignored.

Now, what about the research priority?

It isn't necessary to choose (say) between research and recruitment. Each 
organization, and each individual, can and will make particular allocations. 
CI does both. But if we want to look at trade-offs, it may be slightly 
helpful to make an analogy with startup tech companies.

What should they go for-early profit or growth? A case can be made either 
way, or for a hybrid, but usually they go for growth, on the theory that 
early growth will be most advantageous for later profit. Many critics scorn 
companies that accept large current losses to gain customers-but in many 
cases there is no other way. One writer recently pointed out that the 
Chrysler minivan line, now hugely profitable, started out for years as a 
loser, with large investments by the company and very small sales. If the 
minivans had been a separate company, the criticism would have been 
deafening, but buried inside Chrysler it was just a small divisional 
experiment, which had resounding success.  

Analogies are unreliable, of course, and the discussion could-and doubtless 
will-go on at great length. I only want readers, especially newcomers, to 
realize that few things are  simple, and in particular that a reasonable 
degree of confidence in nanotech does not deserve the R word.

Robert Ettinger
Cryonics Institute
Immortalist Society
http://www.cryonics.org 

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=13317