---

Objects

1. Some specific cases

Constant automaton :  no E, juste an output function dependinf on inputs. That is the same that a fix E and its content integrated in the ouptuf function.

No output automation. ;This cas seems uninteresting, and without importance for the evolution of the automaton, as long as the output have no effet on the inputs.

No input automation : no iputs, or at least, neither E nor O depends on them. In this case, the systes is necessarily cyclicla. But the cycle may be very long (problably exponentially long in respec to E).

Reversible automation :
- E may be deduced from E_-1 , E, I_-1 and O ;
- any anterior state may be reconstituted.

Narcissistic automaton : I_i+1 starts again from O_i

Sollipsistic automaton : E = f(E) uniquement

Passive automaton :
- constant  E = E, O = f(E), no I (not necessarily constant, but necessarily cyclical)
- transparent : E = E, O = I
(memoryless filter, or higher order)

Emergence begins when we can make the differenc between I and D (what we perceive about E). In order to indify O, the system must have a mirror. Then we have a global O, image of the system (self) for the external world. And, inside this global O, the explicit O (or controlled O : gesture, speech, attitudes).

2. Emerging automaton

Let us go beyond the finite automaton model. And dive into a continuous wold.
In time : actions and phenomena not depending on cycles.
In structure : actions and events happening out the formal I/O windows. For instance, physical capture of S by another entity. Global elevation of the environment and the S itself.
In environment : not formal.

Note that the telecommunications metrics matters only if the physical world has a role. Unless we got a digital description for the physical world itself, or at least that the analog part can be masked by digital patterns/info.

The is some fuzziness too in the relations betwen parts of the entity.

Emergence of a given bit : degree of fuzz.

Take the emergencde of a bit in the material/physical sense, but first postulate that the bit is logical.
or else : start from any point and work on the dematerialization of its relations.

One could say : there is simultaneous emergence of physical and logical, when emerges the representation, the sign.

Emergence demands that someting external or the environment "wants" the transfer. Up to the moment of a possible revolution, there is an indefinite nesting. And man is not the ultimate root.

Emergence is over all emergence of the self.

b) porosity A,E (software hardware undifferenciation).

There is no.. between E and A. We can adjoint to A the constants bits of E. At the limit, if all E bits are constants, thn for all E, A(E) = E, merges in A.

Conversely, a part of A may be explicited as a a bit string in E. But where is the limit ? Even if A becoms an execution function for a program containd in E, we nevertheless need some minimal functions in A to execute these programs.

Or, then, A ceases to be an automaton and becoms a simple object (program, memory) for another automaton.

Quid of a minimal automaton ? The clock ?

3. Memory

Petrach encouters " Qu'allons nous transmettre à nos enfants " (summer 93, in Houlgate probably on France Culture). With Bernadette Bricard, Paul Weil, Antoine Prot, Emmanuel Ascarat. "Il faut n'avoir l'air de rien : on ne fait pas passer ce qui était l'objet de l'effort volontaire" (children do not memorize this king of speech, but rather, indirectly, the attitudesce n'est pas cela que les enfants retiennent, mais plutôt, indirectement, les attitudes)

Le théorème de Pythagore contient 24 mots.
Le principe d'Archimède, 67.
Les 10 commandements, 179.
La déclaration américaine d'indépendance, 300.
Et la récente législation en Europe limitant les occasions de fumer, 24 942.

Pub Philip Morris  SA dans Le mondedu 22 juin 1995.)Le texte ci-dessus occupe presque toute la page. En bas de la page :

" Une réglementation excessive de la vie des gens peut amener à de dangereuses atteintes à la liberté individuelle. ET en matière de libertés individuelels, il est toujours plus facile de restreindre que d'élargir. Philip Morris, 17 000 employés au service des 97 millions de fumeurs européens".

"Les changements dans les flux qui traversent chaque partie du système caractérisent sa fonction". Mélèse 1972. Autrement dit, la fonction est le traitement, ou la différence entre O et I. La manière extérieure, boite noire de décrire le système, sans aller jusqu'à décrire E avec ses programmes...

"C'est une impression d'étouffante tristesse qui se dégage d'un monde axé sur la fonction... ce monde est vide, il sonnne creux." Gabriel Marcel 1955 1/179

"Une fonction opérationnelle ... est une classe faite d'un ou plusieurs types d'actions exercées apr la même unité fonctionnelle ou par plusieurs, et dont le rôle est de réguler les flux entrant et sortant en provenance ou à destination d'une série de niveaux considérés globalement. " Blumenthal 1969

"Un pont, des conditions normales d'utilisation, se comporte simplement comme une surface relativement lisse sur laquelle des véhicules peuvent se déplacer. Ce n'est que lorsqu'il est en surcharge que nous apprenons les propriétés physiques des matériaux dont il est fait." HA Simon 1969

"Fermer par une détermination supplémentaire le système de causalité réciproque des structures déjà existanes équivant à ajouter une structure nouvelle spcialisée dans l'accomplissmeent d'une fonction déterminée. ... La différenciation va dans le même sens que la condensation de fonctions multiples sur la même structure." Simondon 1/30, 38

4. Tentative automaton which makes the models better

Let us set a memory of infinite capacity
an a bit flow which a system S tries to predict as well as possible
1st bit : S bets that the second one will be the same
2d bit : if OK, confirmation
else : stop prediction and wait
3d bit : if the same as the 1st, predict alternancde.
else, if same as the 2d, predict that the 4h will be the same ..

Formulation "A tout moment, le rendement des objets s'accroît, ou du moins de certaines catégories d'objets, avec du fonctionnel au numérateur et de l'organique au dénominateur.

5. Some definitions 

Un système abstrait, fait de concepts, est totalement défini par les hypothèses, axiomes et postulats de ses créateurs. Le Moigne 1974, d'après Ackoff .

Les systèmes abstraits ont un nombre limité et défini de psychèmes. Lussato 1972

Phase abstraite : Ensemble des objets n'ayant pas de signification intrinsèque, mais représentant (ou étant couplés à) des objets concrets, des classes d'objets ou des objets psychologiques. Ces objets (symboles et signes) sont véhiculés par un agent concret, en phase physique, le support. Lussato 1972

On ne peut contester qu'il n'y ait là une puissance réelle. L'absence de doute et de scrupules, la parfaite unité automatique, garde cet être à part des tergiversations humaines. J. Michelet, Histoire de France, Ed. Rouff. tome VI.

Il suffit qu'un processus soit explicité pour qu'il soit réalisable par un automate. Lussato (et divers)

L'automatisation est la substitution d'une machine à un homme pour effectuer un travail déterminé. L'automation cherche à créer des automates et met en oeuvre des moyens nouveaux qui modifient complètement les procédés de fabrication. Chenique 1971

6. Analog absorption capability

Thesis. When it grows a digital object may absorb/integrate more analog parts/aspects, in particular in its inteface with humans.

///icono

the look of the curve, in the three zones I, II and III, must differ according the nature of the problmes
cf. p? 156 (in Delahaye ?)
I. non compressible string, length pg. print + long string
J. time of printing x string length
L. time print . lprint. (l string)2

//icono

2
- length = print + decode + l string/2
- time = (print + decode) * l.string/8

total cost
time power (set, number of cycles). memory size

when an automaton "plays always the same game", it is a limit in generative power.
it could be augmented by interaction with other automata. eg. light

Symmetry with analysis. Factors here are used as multiplicators.
Numeration with base b as a multiplying example (number is written as a sequence of digits, and its value is the sum of the digits times b^rank)

Would it be possible to design an algebra of multiplications. Probably possible only for very narrow object spaces.
Then, the other fundamental example is embryology, which shows clearly the complexity ot the thing. The multiplication is not at all a sort of simple sequences applied globally !

Ops opf the factors Vs. Ops of the product.
Programs and data have no proper ops, only a length.
We find again ghe ops in the followin way :
- a program of a given length (KC if optimized), in bits
- data of a given volume (same optimisation),

then bits², to be multiplied by the processing time (perhaps a division...)
Beware here, recursion. A short program may generate an infinite looop (10GOTO210)

Then decompression and possibly amplificators
- alias
- interpolation
- noise, texture, timber generation... rendering.

A basic aspect of DU is that the output may be considerably amplified and broadcasted to billions of auditors.

7. Haecceity

By contrast, what makes haecceity interesting is the contradiction/loss ratio versus product of abstract models. That could be measured :

The scheme must be perfected.

rejoins somewhere the capacity of concrete to generate complexity ad infinitum
then, in some way F =0 for a real objecdt. But, webare, the numetor of F must be countable
If it has a finite number of bits ; is it, indeed... concrete ?
Rejoints the idea that the higher level is richer thatn the lower one. Hence eternity, etc.

It remains to show how a finite number of bits can, in interesting way, be presented as the concretization of an indefinite number of complex abstarct models
and, reciprocally, how they can be generated
(show that  R³ orthonormed is optimal ?)

On the generation /
- works parly by projection, then reconstitution work
- reduction on this subspace
but in fact multiplier, since afterwards it will work by product

n' bit smaller (because subspace) then reduction by transition to metafile, to model
Goees to : multiplication/reduction of concretisations

n = N³ goes into n' = N²
n" > or << n' by transition to meta (but not forcibly)à
N = n"'
quid of the n"'.n"?

and the fuzzy sets heres ?

I try to get the concret out of the abstract like a rabbit out of hat. My models lets itt go parlty.
Basic solution : sqare pattern with origin
Better solution : from an abstract (powerful) object, define a reference.
The first solution is more organic, the second more functional.
See also the

of Bourbaki.  (what there is ther, formally).

the load/distance rule
18.6. Full capacity

See meaning aspects

also the 7L theory. Place of parsing (change of level...)

all the factors separated in the analysis phase could be used in the reconstruction.
but there is a choice of certain, plus other multiplicators

The four causes. Printing press example
Matter : the paper (but also the hardware of the machine and mall matter around)
Form : the image/form to be printed. (but also the form, structure of the machine)
Efficient : moves of the operator, typograph
Final : the reader, customer

by itself, this process is static, without changes other than wear and tear

Laws result directly from inside DY, or the L geodesic, or from physical laws (Moore). The Kantian a priori/a posteriori

Internal/external convergence
- knowledge then control and possiblty engineering in bioogical genetic more generally : biology, neurosciences
- genetic programming, neural nets, more generally : more complex machines, partly mimicking biological processes

The core, operations at core
here the n+1th bit ? and the main legitimation of my ops definition ?
Decision
limits of the mix, see 18.6. on "self"

8. Self-reproduction

in bio, the parsing sends to material, mixed
in DU ? chomosome are not probably a message (but enzymes, yes)
the adult body/soul is the message

9. Telepathy

The real issue here is not the spatial distance (teleportatio like, etc)
but the direct coupling of the cores, without following the normal synthesis/analysis process to communicate
one could say : it is a problem of digital distance
but with distance even shorter than S and S' diameters

To be studied : It is normal that the communication process reach autonomy on the same way as the genetic/processing process.

As long as processing and communicating (and storing ?) progress at the same rate, we observe the same raios of bits per cycle. in some way, the communication process does not progress, in apparence. (the old problem : if I had more hours per day, accelerating ther ate of thogh... something like that in Eggheads

Aim : look for the simplest function vanishing in the points (optimum = length of equation over precision). Thus we get another kind of KC, "with loss", on a strictly logic standpoint, but perhaps with gain if we consider that the original bitmap was a forced digitization of continuous sub-jacents beings (e.g cicles). A sort of Platon cave. So do we a true work of vision.

Dialectical opposition of binary of the meta (noumenon) and the binary or phenomena.
You go from meta to phenomena through : alias, paper grain, noise.
And from phenomenon to meta by analysis, etc. You climb back up looking tor the intention (see something like that in Delahaye)

A given S has a yield more or less near the maximal (use time/total time, good or bad use of the resources relatively to the aimed at functions)rhaps it would be better here to say "heigth" of processing rather than depth

More bizarre cases, wirh non progressive distribution of levels
Architecture examles
one only bit in input, and two messages (as lon as one will) in output. Then we have a choixe processor at input, a storage of the two mesages and a copy generator in output, with the following description in mode GG :

Examples
I et O are messages of same length
a/ the transformation is a random permutation ; the only one way is o store in memory the full truh table
b/ if they are not permutations, there is several tims the same string in output, then there are classes in input (all those which have the same outpu) and then we can have a class selection processor in input (recognition)
c/if they are permutations, perhaps can they be decomposed (the first ten bits in input give the first ten bits in output) ; then the truth table is decomposed into sub-tables, of whith the sum is sensibly smaller
d/ the most general case, where we have a transfer low ; see how this law is formulated
a general case is an answer to a request ; then we can consider that we have in input the reuest + the whole database (or this later is in E, but is not very different)
This depth analysis will habe to be passed in concrete (physical world)
journa/press, hypertexte, server
granularity, zap, cuting, emssion (Nulle part ailleurs)
According to the average temperature of E and of the distribuion of these temperatures

If S, I and O are of high level, processing depth is not incompatible wiht inputs respect. If there are many bits, standardization is not incompatible with personnality.

Saturation for the observing man

Could be rather near to filtering.

A maximal depth could be said reached when the analysis leads to a perfect decomposition in orthogonal factors. At the limit, a set of fully orthogonal bits.

At the bottom, minimal volumes, high pressures, a totally crystalline and orthogonal world, and coherent.
But never totally. Faults (like in gems) and contradictions, and constraintes opposed to orthogonalitéy.

At the heart, the genetic code, the most specific traits of Roxame, but also the essential about the external world. Decision, creativity.
From a judgment standpoint : pure value, absolute a priori.

When there are inputs, ideally, they are radically factorized, up to the moment when they can add some orthogonal bits to the center.

Quand il y a des inputs, idéalement, ils sont radicalement décomposés, jusqu'à ajouter quelques bits orthogonaux au centre.

Reduction/generation : two categories of choices. In the first, select what is interesting in the ouptut, what "revient au même", does not matter.Then, from he core, what should be added. And, in the center, the basic genetic combinations.