abstraction = compression

you know this software that turns photographs into pixel grids, using a
smart algorithm to decide which pixels in the grid would be black (there)
and which white (not there)? they use it on passports now. it's typical
police software, originating from the search for automated methods to
positively identify people's every characteristics like finger prints, iris
scans and now also photographs by abstracting these to so compact a data
unit as possible, allowing them to be stored on data carriers with
capacities that do not conflict with current reality.

well if you would extend the bit decisions of these algorithms (black or
white) into byte descisions, using ASCII character set SHAPES instead of
pixels (and the space for a blank), provided that the algorith would exist,
you could use a PC-XT to turn any digital photograph or picture into an
automated piece of ascii-art.

which could next be manually adapted to look genuine.

don't know of any algorithms to seperate the two.

but in a universe as large as ours they would undoubtedly exist, or should I
say have evolved?

jer0en wrote:

> positively identify people's every characteristics like finger prints, iris
> scans and now also photographs by abstracting these to so compact a data

Are you equating every characteristic with very limited characteristics
such as finger prints and iris scans?  Also, note that when the police
(spit) match finger prints, they do not match entire finger prints but
just a small number of features--a fact that has led to miscarriages of
justice.  Is there any reason to suppose that no two people have the
same finger print, ditto iris scan?  No.

-- 
He is not here; but far away
  The noise of life begins again
  And ghastly thro' the drizzling rain
On the bald street breaks the blank day.

I'm sorry for reacting this late, but though I'm sceptical about quite a
number of things, I would have to concede to the fact that there seems to be
more than sufficient space in the total of possibilities of human
fingerprints as well as our irises to allow a zero number of matches being
due to coincidence to remain plausible in basicly any form of reality that
would be plausible. That is, without figures as to the contrary.

could you make the paranoia stick numerically?

if it's sufficient to identify any one of 7 billion numbers you would need a
minimum resolution of 10 digits.

this is a minimum. if the specs of fingerprints would be randomly spread
over the fingertip surface you would need more resolution to make an ID than
if the specs would be somehow within a limited region of the total of
location possibilites, that could be interpreted as non-random, like in 7
billion unique numbers starting from 1 to 7 billion instead of, for
instance, freely variegating within the total possibility space.

both the measure of the ID entropy and the proportion of total ID space to
ID total could provide indications to rule out coincidence.

to assess the total ID space the resolution of fingerprint spec images that
the police uses would be a usefull datum.

if it would cover the entire finger tip, it would represent the total ID
space, that is, of course, limited by the number of possible actual
fingeprints within the matrix.

if it would cover just part of the fingertip, it could be indicative of a
lack of entropy, that is of coincidence, and as such of some form of
sequential numbering.

resuming, the three basic characteristics of sequential numbering are

A. withing the total ID space, a lower ID entropy than is plausible by
normal coincidental frequential spreading, typically even zero

B. with regard to storage, a useful, i.e. low total ID space / ID total
ratio

C. the absence of doubles

above characteristics are typical but not necessary. the use of compression
would reduce the need for a low ID space, the use of encryption could
disguise a low ID entropy, which would require you to figure out a key, and
doubles could be introduced to enhance plausibility within a particular
reality.

given the above, it is unsure whether indications of some form of
non-conincidence, of regularity like in sequential numbering, could be found
at all in unique IDs such as our fingerprints, not least since safety levels
could apply that would depend on both the processing ability of the marker
and that of the marked.

otoh, the higher the safety level of a hacked encryption key, the more
certain the fact of the encryption.

the proof would stand like the universe itself.

if processing and storage resources of the marker would be sufficient, the
encryption key could simply be a complete copy of each individual's DNA
which, together with some sort of ontogenetical algorithm, would "render"
the entire individual, fingerprints and irises included.

moreover, if the ontogenetical algorithm could be run in reverse, it would
allow the marker to determine the DNA of individuals to specifications. if
that DNA could be physically reproduced and inserted into egg cells, such
individuals could be physically created.

there are facts even in common reality weakly indicating that this would be
the case. since we do not currently have any such processing and storage
resources ourselves, nor the ontogenetical algorithm, we would already have
the key but not the means to use it.