The current state of affairs in the field of using polymer hydrogels for the creation of innovative systems for signal and image processing, of which computing is a special case, is analyzed. Both of these specific examples of systems capable of forming an alternative to the existing semiconductor-based computing technology, but assuming preservation of the used algorithmic basis, and non-trivial signal converters, the nature of which requires transition to fundamentally different algorithms of data processing, are considered. It is shown that the variability of currently developed information processing systems based on the use of polymers, including polymer hydrogels, leads to the need to search for complementary algorithms. Moreover, the well-known thesis that modern polymer science allows for the realization of functional materials with predetermined properties, at the present stage, receives a new sounding: it is acceptable to raise the question of creating systems built on a quasi-biological basis and realizing predetermined algorithms of information or image processing. Specific examples that meet this thesis are considered, in particular, promising information protection systems for UAV groups, as well as systems based on the coupling of neural networks with holograms that solve various applied problems. These and other case studies demonstrate the importance of interdisciplinary cooperation for solving problems arising from the need for further modernization of signal processing systems.

It is shown that there are prerequisites for returning to the original interpretation of logic as the science of the laws of thought. Using the methods of the neural network theory of the noosphere, it is proved that the nature of thinking is significantly influenced by the socio-cultural code of the ethnic/social system. A connection is revealed between philosophical categories as basic undefined concepts and archetypes, understood according to Jung. It is proved that the formalization of the laws of thinking, built on multi-valued logic, should operate not only and not so much with the concepts of natural language, but with “reflection quanta” lying on the archetypal level of consciousness. It is shown that the category of truth underlying classical logic can also be interpreted as a “reflection quantum”, but which has undergone a formalization procedure. It is demonstrated how exactly the apparatus of multi-valued logic can be applied to identifying links between the "quanta of reflection", which can be considered as a generalization of the rules of reasoning established in classical logic, dating back to Aristotle. The category “Qi”, which underlies ancient Chinese philosophy and is inseparable from the Chinese sociocultural code, was used as an example. Meaning of this category in the version that is assimilated by European and Eurasian sociocultural codes is revealed.

The analysis of the existing literature demonstrates that in order to address the fundamental challenges associated with the origin of life, it is essential to consider this problem from a comprehensive perspective, specifically from the vantage point of the general theory of evolution of complex systems. From these positions, life should be regarded as a distinctive instance of an information storage and processing system that emerges naturally. Evolutionary processes should be examined from the vantage point of the coevolution of material and informational components, which has not been sufficiently emphasized hitherto. It is shown that a specific example in this respect is analogues of neural networks spontaneously formed in solutions of some hydrophilic polymers. Such systems lead to the formation of non-trivial information objects. A wide range of other examples is considered, proving that the processes occurring with the participation of hydrophilic polymers should be interpreted, among other things, from the point of view of formation of information objects, which, under certain conditions, influence the processes occurring at the molecular and supramolecular level. It is shown that it is reasonable to use the tools of classical dialectics to solve such fundamental problems as that of the origin of life.

The nature of qualitative transformations of warfare at the present stage is analyzed. It is shown that these transformations will inevitably lead to the question of partial transfer of individual's intelligence to non-biological information carrier. In particular, in the foreseeable future it is predicted that AI will emerge based on the intelligence of officers who possess advisory combat experience. It is shown that tasks of this kind can be solved by existing software tools. It is also shown that the solution of the problem of partial transfer of an individual's consciousness to a non-biological information carrier goes beyond the concept of transhumanism in its existing version. AI built on such a basis, regardless of its intended purpose, becomes a kind of an intermediary between a person and the suprapersonal level of information processing. It is shown that the emergence of such intermediaries can significantly affect social relations, in particular, significantly increase the influence of those philosophical concepts that insist on the importance of collectivism as a factor that is closely related to the humanistic picture of the world. It is shown that, paradoxically, the development of military AI will inevitably serve as one of the factors giving a new impetus to the ideals of humanism.

The increasing volume of unstructured textual data in social networks requires automated tools for efficient classification and monitoring. This study presents the design and evaluation of an AI-based system for multi-class text classification using a dataset collected from VKontakte. The dataset was annotated into several semantic categories of the word "hero," serving as a domain-specific case study for testing classification models under real-world constraints such as class imbalance, overlapping categories, and limited training samples. Three approaches were implemented and compared: a Long Short-Term Memory (LSTM) network, the transformer-based DeBERTa model, and an AutoML solution (LightAutoML). Experimental results show that DeBERTa achieves the best-balanced performance with a macro-F1 score of 0.32, while AutoML provides the highest raw accuracy (~65%) with lower resource requirements. LSTM demonstrated limited effectiveness due to the dataset size and complexity. Additional experiments with class balancing and refined labeling improved performance across underrepresented classes. The findings highlight the trade-off between model complexity, computational cost, and classification performance, and confirm the applicability of transformer-based architectures for text analysis in noisy and imbalanced environments. The proposed system can serve as a foundation for automated monitoring tools in social media and other real-world NLP applications.