ФИЛОСОФИЯ / 6.Философия науки

Elkhova Oksana Igorevna, Doctor of Philosophy, professor, Department of Philosophy and Politology, Bashkir State University, Russia;

Kudryashev Alexandr Fedorovich, Doctor of Philosophy, professor, Department of Philosophy and Politology, Bashkir State University, Russia.

METHODOLOGY OF THE CREATIVE PROCESS IN SYSTEMS WITH ARTIFICIAL INTELLIGENCE

It is necessary to state the following about artificial intelligence and the modeling of the creative process in systems with artificial intelligence. Systems with artificial intelligence cannot yet compete with man in creativity. The creation of technological systems that are capable of creativity, by copying the logic of the individual’s step-by step inference, is a difficult task and ends in a very modest result.

The development of systems with artificial intelligence is very different from conventional programming. If an ordinary computer program can be represented in a paradigm: a program is an algorithm plus data, then for systems with artificial intelligence the paradigm is different: here the program is a knowledge base plus a knowledge-management strategy. The main distinguishing feature of systems with artificial intelligence is that they work with a knowledge based.

In conventional programs, data representation for the algorithm is not difficult. For systems with artificial intelligence, the representation of knowledge becomes a problem. This problem includes a lot of questions: what is knowledge, what knowledge is stored in the system as a knowledge base, in what form and how much, how to use it or replenish it, etc.

It is worth mentioning the differences between data and knowledge. Unlike data, knowledge has the following properties: internal interpretability, structure, connectivity and activity. Internal interpretability assumes that the data have unique names and attributes that make it possible to operate with them as information units.

Structure means the decomposition of complex objects into simpler objects while establishing the following relationships between them: “part-whole”, “class-subclass”, “genus-species”, and so on.

Connectedness reflects the patterns of facts, processes, phenomena and causeand-effect relationships between the elements of knowledge. Human cognitive activity has specific characteristics. In other words, human knowledge is active, while in ordinary programs, data are passively stored in the computer’s memory. This fact fundamentally differentiates knowledge from data. For example, the discovery of contradictions in knowledge becomes the motivation for overcoming them and engenders the emergence of new knowledge. The same stimulus of activity is the incompleteness of knowledge, expressed in the necessity of their replenishment. Of course, the idea of providing knowledge activity in systems with artificial intelligence generates methodological difficulties in terms of its implementation.

At present, systems with artificial intelligence cannot compete with a person in creativity, in the absence of a database at their disposal that is comparable to the human potential of the knowledge of common sense.

A few words about the knowledge processing strategy are necessary. The strategy of processing knowledge is closely related to the skills that people have in solving creative tasks based on heterogeneous knowledge that cannot be formalized. It is difficult to implement this function within the software and hardware systems. The knowledge on which the person relies for solving creative tasks is heterogeneous and cannot be formalized. It includes a set of concepts and their interrelations, knowledge about the structure and interaction of parts of different objects and quantitative and qualitative characteristics of objects, phenomena and their elements.

An ordinary computer program carries out a process for the logical operation of data which are given in a single formalized form. The strategy of processing knowledge in systems with artificial intelligence is based on the hardware and information-software complex. The action of this complex is analogous to the action of the mechanisms of the thinking and decision-making of a person.

In conclusion, speaking about the methodology of the creative process in systems with artificial intelligence, we once again emphasize that the creation of technological systems capable of creativity must be carried out according to the principle of the knowledge base plus the strategy of processing knowledge.

The widely known classical Ada Lovelace’s Objection boils down to the assertion that the computer is not capable of independent creativity since creativity means the production of a new result. Computers cannot invent anything new; their fate is the strict implementation of the requirements specified by the person who writes the programs for them. The implicit precondition of the Ada Lovelace objection is the classification of concepts new to all kinds of results, without considering who produces these results. If we distinguish between the results by the criterion: who receives these results, we get two types of novelty. On the one hand, the new results that a person receives, on the other hand, new results that artificial intelligence receives. It can be said that at the present time creative competition between man and artificial intelligence is born. Returning to the goals set at the beginning of this article, we can note the absence of insurmountable obstacles to the growing field of artificial intelligence which, in principle will be able to compete with man in creativity in the future.

The composite stage in systems with artificial intelligence creation is realized as programmed search option of a new product in the sphere of possible combinations. The creation of technological systems capable of creativity must be carried out according to the following principle: the knowledge base plus the strategy of processing knowledge. The strategy of processing knowledge is closely related to the skills that people have in solving creative tasks based on heterogeneous knowledge that cannot be formalized.

References

Кудряшев А.Ф., Елхова О.И. Творческий процесс в системах с искусственным интеллектом // ITIDS+RRS'2014 (Information Technologies for Intelligent Decision Making Support)», Volume 2, May 18-21, Ufa, Russia, 2014. С. 191-195.