Adam Maria Gadomski's Meta-knowledge Engineering Server 

ON COMPUTATIONAL AXIOLOGY and ETHICS:
a TOGA meta-theory long-term perspectives 

Working Page since  23 Mar.2001, last updated 20 Oct. 2004.

On "Philosophy updating"
As computational ontology and epistemology as well as, axiology and ethics have not more the classical philosophical notion because their subject is not more an abstract/ideal  human which represents the humanity.   Also autonomous robots and computer systems  react on their own subjective vision of the world.  They do it according to individual perception and conceptualization capabilities, their own information, preferences and knowledge. In this sense, they have a subjective ontology, epistemology and values. Me may argue that they have  a very simple, not conscious yet  "e-philosophy".
Anyway, the e-philosophy will be a symptom of a high-intelligence of  abstract intelligent agents (personoids).
The possibility of computer implementation of "philosophy components/modules" is a base for  its new forms, such as, an experimental and  computational philosophy (using simulation methods) .


In consequence, we may define top-down new interest domains of  emerging systemic and over-disciplinary research branches and theories. They could be, for example, 

-  Unified Intelligence Axiology and Ethics, UnIA:   It is an hypothetical unified theory of Axiology and Ethics of intelligent socio-cognitive systems, natural and artificial, it can be seen as a complementary part of an anticipatory science which can be called  the Unified theory of Meta-Systems (UTMS). UTMS  is, in my opinion, in an initial state of development. 
As well as, more abstract yet,
-  Unified Theory of Meta-Systems integrates common theoretical properties of real-world and possible abstract systems: large homogeneous systems, system aggregates, global systems, meta-system hierarchies, intelligent systems, knowledge systems, system decomposition, evolution, self-developing. as well as all properties of human knowledge which are independent on the nature of the concepts used. 

For example, TOGA  is a pioniering approach to the top-down construction of an UTMS.
In parallel, , the theories of complex networks/webs/grids/graphs also contribute to the UTMS development. 



An intelligent beings/entity/agent is difficult to imagine in a not intelligent environment. Multi-intelligent systems create them self social relations, and, in consequence, they needs ethics system for common existence/ surviving.
Aggregates of social groups/communities/organizations and nations follow different ethics . Their systemic generalization and a development of a common coexistence and collaboration plataform lead us to the meta-ethics research.

Humans-computers symbiosis in the modern Information and Knowledge Technological Society, has enlarge our interest to the  meta-ethics study of  homogenious ethical bases of the possible personoidssociety/ the link in preparation/. 
- Here personoids we can see as abstract high-intelligent entities with structural intelligence, which basic/essential functional properties are independent on their carrier system (a physical realization). 

- The pragmatic studies in this matter performed in ENEA has conducted us to the formulation of  the EoI (Cordis) MEFIT  proposal.

References with remarks:

In  Jack Worlton's paper : Toward a Philosophy of Computational Science and Engineering , Cray Users Workshop,1995: 
This paper outlines an approach to the development of a philosophy of computational science and engineering, in four topics. 

   o  Computational epistemology. Epistemology (a specific study of knowledge) asks the question: "How do you know?"
       Computational epistemology is a study of the reliability of the results generated through computational science and
       engineering; eight methods for assuring computational reliability will be presented: Fermi solutions, computational elasticity,
      empirical consistency, analytical solutions, conservation equations, alternative numerical methods, parameter studies, and
      peer review.

   o  Computational axiology. Axiology (the study of values) asks the question: "How do/can/should we determine the value of
       computing systems? Eight indicators of value will be presented: performance, costs, performance/ costs, robustness,
       compatibility, applications, transparency, and quality.

   o  Computational metaphysics. Metaphysics(the study of reality (rather:  reality conceptualizations ! [A.M.Gadomski]) ) covers a 
       wide variety of subjects for study, including models ( rather: reality model building theories [A.M.Gadomski]). For purposes of 
       this paper, we will present a model of modeling ( really, the author talks about meta-modeling [A.M.Gadomski]). 

   o  Computational methodology. Methodology (the study of methods) includes, among other things, the questions of precise
       terminology and comprehensive classification. Here we present a set of taxonomies of interest to high-performance
       computing, including 24-way, 48-way, and 96-way taxonomies of parallel computing systems, and a 36-way taxonomy of
       communications in parallel computations.

Interesting: Anthony A. Aaby, Computational Ethics, 2002, web

Kari Coleman, University of British Columbia: Android Arete: Virtue Ethics for Computational Agents :

There exist computer ethics, the ethics of using computers, but this statement is simplistic. Computer ethics could be use of computers to create ethical systems for computers. In this sense computer ethics would describe how computers would act in a human world.
Coleman talked about computers as slaves for serving the ends of humans. "What are the computational ethics of these computers?" She did advocate the three laws of robotics articulated by Asimov, but not the appearance of robots.
She went on to enumerate quite a number of the virtues computers/robots/agents would have in order to be ethical. After she finishes collecting these virtues she will be trying to classify them. Examples include: agentive, social, environmental, and moral.
At the same time, these virtues may be interpreted as the virtues of programmers, the creators of the software/robots in the first place ( but, as well as, their organization axiologies, assumed intuitively, embedded in various local business ethics being developed  under different socio-cognitive-economic-political constrains [A.M.Gadomski]).


- A design of intelligent robots requires  the design of  human-robot interactions, and in consequence,  we need a robots' computational ethics and, in general, a computational axiology.
In order to comprehend the concept of computational axiology we need to have a definition of axiology.

Axiology is a general theory/science of human values, their orgines, interrelations and dynamics. It is considered as a part of philosophy, when "theory of value" is a branch of  economy  and political economy ( see Google below). 

Remark: They start from different implicit axiomatic assumptions.

Formal Axiology is a specific branch of the science of Axiology. The late Dr. Robert S. Hartman  developed this science between 1930 and 1973, He introduced the concept Dimensions of Value, going towards a mathematical axiology.  This theory can be considered as a branch of  early socio-cognitive research (see Web). 

In the above context, computational axiology is yet an anticipatory science which should be an unified, mathematically formal approach to the problem of values from the perspective of  intelligent entities. It should generalize the properties of : meta-ethics and the Theory of Value.  I claim, it should  be a socio-cognitive engineering task, possible to the realization according to the TOGA Meta-theory framework.


Google Search: "human-robot interaction", "computational ethics" (30 Oct.2003: 0 docs) 
                         "Axiology "(19 Apr.2004 - 10 300 docs)  and   "Theory of Value"   (41.000 docs)
                         "Computational Axiology" (20 Oct..2004 - 4 docs), only from this server.
                        "Computational ethics" (17 Sep.2004 -46 docs),  from this server 3,4 pos. 
 

    MKS (Meta-Knowledge Server)    |   Intell. Agents Ethics    |   Socio-cognitive Engineering HID  |  A.M.Gadomski