Why Is a Robot-Deity Both Necessary and Possible for the Scientific Canon?

Why is a Robot-Deity both necessary and possible for the scientific canon?

László Pitlik (My-X team) 2015.I.

Keywords: definition of Good, evaluation, objectivity, general problem solving, sustainability, similarity analysis, automation, big data, consistency-oriented thinking, argumentation techniques, robotized inference machine, academic writing skills

Abstract: The mankind, so probably each of us is always searching (in a conscious or an unconscious way) for answers of several basic questions: e.g.: Why we are existing at all? Or: What is the Good-Better-Best? The “Forbidden Fruit” is “a metaphor, the phrase typically refers to any indulgence or pleasure that is considered illegal or immoral.”[1] -. Therefore, to be aware what is Good-Better-Best (c.f. the Tree of Good and Evil)[2], should not be given for the human race according to the Genesis[3].

A Robot-Deity is not a simulation or substitution of one or more Gods. The Robot-Deity is the knowledge (‘closed’ in source code through human intuition and technology and) being able: to evaluate (to rank) constellations, to detect anomalies (bubbles, force-fields), and to define sustainability (strategic and/or operative goals) without any particular interests of persons and/or groups (it means: in the most objective way).

The capability - to decide about Good or Evil – may not be a simple ‘black-and-white’ approximation (c.f. already Good ‘enough’<Better<Best). The evaluation (always based on limited ‘information/knowledge’) is mostly ‘grayish’ (or fuzzy) even if an ideal constellation could be defined.

These evaluation processes should be able to be automated, and also to handle with antidiscriminative objectives (like the hypothesis, where each evaluated object may be seen as the same).

Parallel to the interpretation of Good or Evil, each word of the human languages can be evaluated in the same methodical process, whether the use of a word in a sentence is ‘good’ enough comparing to each previous words and sentences[4].

The similarity analysis tries to approximate the above mentioned rules and expectations (c.f. general problem solving) through creation of a multileveled framework for optimization from entirely context free level (explorative modeling) via antidiscrimination models to classic modeling of production functions so that a type of norm level always can be derived…

The article tries also to demonstrate how relative simple messages should be proved in frame of argumentation techniques, where the self-checking needs more volume as the message itself.

Part I – The Message

Introduction

The mankind, so probably each of us is always searching (in a conscious or an unconscious way) for answers of several basic questions: e.g.: Why we are existing at all? Or:

What is the Good-Better-Best? The “Forbidden Fruit” is “a metaphor, the phrase typically refers to any indulgence or pleasure that is considered illegal or immoral.”[5] . Therefore, to be aware what is Good-Better-Best (c.f. the Tree of Good and Evil)[6], should not be given for the human race according to the Genesis[7].

The phrase ‘Robot Gods’ or ‘Robot Deities’ can only be found in context with sci-fi – without any complex/serious quality. Other virtual robot roles (like teacher, citizen, etc.) got already interpreted in MIAU leading articles – with the conclusion: the development of robots is possible based on big data and artificial intelligence.[8] Further interdisciplinary backgrounds can be found in the new book about the hermeneutics of similarity analyses.[9]

Problem-identification

The most important question of the science should be the question about evaluation capability, where each phenomenon (like persons, groups, institutions, regions, devices, models, etc.)can become evaluated object.

A seemingly easy question is: Which model is the best among a lot of models handling the same problem? The well-known answer is by William Ockham[10]: the simplest (it means: which is based on the fewest assumptions) is the best. Yet, a proved answer needs a framework, where models can be compared based on arbitrary attributes in a parallel way.[11] Simplicity can have a lot of interpretations: the number of assumptions can only be counted, if each assumption has the same complexity. The number of steps of the solution process can also be counted. The time needed to solve a problem is countable, too. The volume of the handled data (information) is also a quantitative attribute of the simplicity, etc.

To rank models, it seems to be universal enough – but the list of benchmarking phenomena is long (not only in case of solutions can be spoken about ideality)[12]:

Which is the best rock-and-roll motive among alternative solutions? (c.f. robot-musician)
Which is the best variation of hymn among potential initiatives? (c.f. robot-citizen)
Which is the best candidate for the presidency? (c.f. robot-citizen)
Which is the best client for credits? (c.f. robot-economist)
Which application is the best in case of a public procurement process? (c.f. robot-economist)
Which is the best word to complete a sentence in certain situations? (c.f. robot-speaker)
Which penalty is the best for a concrete person? (c.f. robot-judge)
Which method is the best for a child being educated? (robot-teacher, robot-parent)
Which patient should wait no longer for medicine/operation? (c.f. selection of sacrificial lambs – c.f. robot-doctor)
Which behavior is the most suspected? (c.f. Robocop)
Which word is here and now the best for a sentence? (c.f. Robot-Speaker)
etc. (c.f. Robot-xxx)

The searching for the best object (it means: for the most ideal attribute-constellation) lets assume, that the evaluators are capable to rank each object involved into the process. They should be able to prove, whether each object can be evaluated as the same?

Summa summarum: alternative constellations (objects) are existing in each context. Decision support systems (or committees) always have to rank them. The human knowledge about ideality (benchmarking) can be transformed more and more into source codes…

If somebody should think: the human race has not the right and/or the knowledge to rank objects in any way, the institutions like Fitch, Moody's and Standard & Poor's[13] may not do this,too[14]. Nobody may also think of prizes like The Prima Primissima Award in Hungary (c.f. object of the year, Time Person of the Year), where a lot of factors should have influence of the selection process…

The core problem, what should be solved since time immemorial, is: what is the algorithm (being as far as possible independent from its creator’s interest) to derive the most ideal object in case of selection procedures?

The capability of derivation of the Best assumes that the evaluator can also define both the second best and the rank position of each other object. Parallel, the ranking of objects can be interpreted as the capability of a norm, from which level it can be spoken about Good at all. Therefore the word ‘Good’ should exist both in an absolute and a relative approximation (c.f. fuzzy/grayish evaluations). So, the capability - to decide about Good or Evil – may not be a simple ‘black-and-white’ approximation (c.f. already Good ‘enough’<Better<Best). The evaluation (always based on limited ‘information/knowledge’) is mostly ‘grayish’ (or fuzzy) even if an ideal constellation could be defined.

These evaluation processes should be able to be automated, and also to handle with antidiscriminative objectives (like the hypothesis, where each evaluated object may be seen as the same). The automation makes possible to ensure the approximation of the highest level of objectivity…

Parallel to the direct interpretation of Good or Evil, each word of the human languages can be evaluated in the same methodical process, whether the use of a word in a sentence is ‘good’ enough comparing to each previous words and sentences[15].

Solution

The solution may be identified in the phenomenon of similarity![16] Similarity analysis[17] is a process family, where (based on more or less restrictions) a kind of optimal parameter set of an adequate staircase function can be derived in an automated way.

Experiments about capabilities of self-controlled chains based on similarity analyses run since 10 years.[18] The results show: the similarity could be used to approximate the highest level of objectivity for the decision support in an automated way.

Similarity analysis can derive both a sort of norm level and a scale for approximation levels of a probably ideal constellation.

Similarity analyses can be seen as a thinking methodology being able to collect general human capabilities which can approximate objectivity itself. Therefore the transformation of phenotypically and heuristically (c.f. intuition) existing human capabilities in source code can ensure a type of Robot-Deity (aiming objectivity).

So, the similarity analysis tries to approximate the above mentioned rules and expectations (c.f. general problem solving) through creation of a multileveled framework for optimization from entirely context free level (explorative modeling) via antidiscrimination models to classic modeling of production functions so that a type of norm level always can be derived…

Conclusion

Assumed, if the Robot Deity (it means now: the algorithm to ensure the highest level of objectivity for the widest area of phenomena) can be created and used in fact, then each evaluation process and each opinion about them should be changed in the entire society. The frequency of useful human intuitions (in the segments of phenomena being described in the world of big data more and more complex) will be more and more a kind of rarity (c.f. frequency of win strategies by chess master against chess-computers). The phenotypic knowledge of mankind will be more and more transferred into source codes. The ‘human’ and ‘real’ science will be smelted again. The so called human-and-‘machine’ symbiotic interactions speak about a ‘machine’, which is a sort of conserved human intuitions…

On the same complexity level as the problem about the Robot-Deity, the problem of an Inference Machine can be set. An Inference Machine should be able to deliver argumentation checks (e.g. through robotized using of rules in academic writing skills) in an automated way (c.f. Annex / Part II)…

Summary

In case of each phenomenon, a kind of necessity to be able to define the ideal (and a scale for measurement the differences to it) can be detected since time immemorial. The capability to rank potential constellations cannot be seen as a black-and-white problem. The automated approximation of ‘grayish’ outputs based on big data impulses seems to be more and more realistic e.g. in frame of similarity analyses…

Literature

see the hyperlinks in the footnotes

Annex/Part II. – Self-checking experiment

The article tries also to demonstrate how relative simple messages should be proved in frame of argumentation techniques (c.f. inference machine), where the self-checking needs more volume as the message itself.

A self-checking can be executed based on argumentation techniques. It means: First, each declaration should be identified. In the next step: each partial conclusion should be derived. Finally, the closing conclusion should be set. For example (calculation of average body height for a population):

Basic declaration set (or facts or given values pro cell in a spreadsheet):

measured body height in a given case: 160 cm
measured body height in a given case: 170 cm
measured body height in a given case: 180 cm
measured body height in a given case: 190 cm
+ definition of average

Partial conclusion (results calculated through function in a spreadsheet):

sum of the body heights = 160+170+180+190 = 700 cm
number of measured object = 4 persons (c.f. count()-function)

Final conclusion (closing result calculated through function in a spreadsheet):

average body height (per definition) = sum of body heights / number of observed objects = 700 cm / 4 capita = 175 cm/capita

How could be used the above derived thinking pattern in case of the article about a robot deity?

Basic declarations: e.g.

The knowledge about Good and Evil has a central role in the thinking procedures of mankind about itself since time immemorial. (c.f. Bible)
The words for Good (Better/Best) are existing at all. (Even in English Good and God seem to be especially similar – c.f. central role of similarity in the potential solution).
A Google Search for ‘Robot Deity’ or ‘Robot God’ do not produced any scientific article, where the quality of an article could be interpreted through Google Scholar (it means: no sign of Google Scholar could be identified).
“The development of robots is possible based on big data and artificial intelligence.” (see footnote Nr. 5., which can be interpreted as a type of evidence proving)
Similarity analysis exists as mathematical approximation, and seems to be useful for creating robots. (see footnote Nr. 6., which can be interpreted as a type of evidence proving)
Definition of the Robot Deity as declaration of the author.
Definition of the connection between existence and consciousness[19]

Partial conclusion on the first level:

Necessity of a Robot Deity is given since time immemorial (based on basic declaration Nr. 1. and 2., where necessity means phenomena being permanent existing but not clarified enough. On other hand, the basic declaration Nr. 3. could be seen as a kind of argumentation against the necessity, assumed, if no traces of a solution are given in a qualitative level, then a problem may not be important enough. However: the definition of a Robot Deity in scientific level assumed a sort of accumulation of knowledge elements: c.f. experiments of artificial intelligence.)
Possibility of a Robot Deity is given: Traces of a Robot Deity according to the given definition are not given, but a lot of other robots seem to be imaginable e.g. based on similarity analyses. Robots are robots, it means: the core capability of the robots should be the same. With other words: a Robot Deity may be constructed on the core similarities of other robots. (c.f. basic declaration Nr. 4. and Nr. 5. and Nr. 6.)

Closing conclusion (it means: based on the derived conclusion of the first level):

A problem having necessity and possibility should and may be handled (c.f. partial conclusions of the first level + basic declarations Nr.7.).
The title of the article used the word: WHY? The answer is: if something is possible, then it is also necessary (c.f. basic declaration Nr.7.). Possibilities (c.f. solutions) will be identified (c.f. intuitions, heuristics) through necessities (c.f. basic declaration Nr.7. – c.f. happiness is the recognized necessity!?).

[1] (Remark: in this article will only be used citations from wikipedia, from the ’common knowledge’ or from author’s own articles before

[2]

[3] (Genesis 2:16-17 =

[4]

[5] (Remark: in this article will only be used citations from wikipedia, from the ’common knowledge’ or from author’s own article before

[6]

[7] (Genesis 2:16-17 =

[8]

[9] ill.

[10] ,

[11]

[12]

[13]

[14]

[15]

[16] ill.

[17]

[18]

[19] „their social being that determines their consciousness”