PROBE FOR FORMULATING A PREDICTIVE MODEL OF AIARTWORLD
This scientific article proposes methodological approaches for predicting the development of non-human artificial intelligence art and its conditions, presenting a definitional framework of the AIARTWORLD model. This model reflects the socio-aesthetic dynamics between human and non-human actors. Two derived models for development prediction are presented: the Ideal Model, which removes human influence on art production, and the Possible Model, reflecting the likely near-term implementation potential. The interplay between these models outlines the speculative predictive interval. A methodological approach focusing on Metaculus predictive market operations is proposed to verify speculative models, blending Crowd Wisdom and statistical modeling. The model’s criteria require exact data aggregation and extraction, drawing from scientific fields such as artificial intelligence research, consciousness research, and cognitive sciences, along with completed art research and current AI-related art practices.
Updated project proposal: June 2023/unreviewed
PROBE FOR FORMULATING A PREDICTIVE MODEL OF AIARTWORLD
Tomáš Marušiak MFA
Masaryk University Brno, Faculty of Arts, Department of Musicology (Czech Republic)
Cultcode-Institute of Visual Art (Slovak Republic )
ANNOTATION
The article aims to present several starting points for the formulation of methodological approaches applicable to the prediction of the development of the non-human art of artificial intelligence and its conditions. To fulfill such a goal, a possible definitional framework of the AIARTWORLD model was outlined, which, like human art and aesthetics, reflects the social and aesthetic situation of the relationship between human and non-human actors. Two derived models have been proposed for the prediction of the development. The first Ideal Model eliminates human influence on art production. The Possible Model is a reference point that reflects the highly probable implementation potential in the short-term horizon. The relationship between the Ideal Model and the Possible Model outlines the speculative predictive interval in which the articulation of speculative models is located. To verify speculative models, a methodological approach focused on Metaculus predictive market operations is proposed, which, as a scientific platform, uses a combination of Crowd Wisdom and statistical modeling. The criteria of speculative models are conditioned by the exact aggregation and extraction of data after using the Prisma method and verification procedures from the Delphi method. Data aggregation is proposed from scientific research areas such as artificial intelligence research, consciousness research, and cognitive sciences. In addition to the aggregation and extraction from scientific knowledge, the next step is to offer similar data processing from completed art research and current art practice in the field of AI.
ABSTRACT
This article extends the research field of the research project Aesthetics of Artificial Intelligence Art. The research project examines the relationships between the art of artificial intelligence and computational neuroaesthetics. The subject of the study and approaches to it are articulated from three basic positions. First, the theoretical position is situated in the discourse of artificial intelligence art and aesthetics. The analytical position is based on the study of current artistic practice and identifies artistic strategies in the subject area, especially the objectives of the project. The experimental position represents a practical-speculative approach to carrying out one’s research into the possibilities of art and aesthetics, which can also be understood as artistic research. The goal of the research is to provide an answer to the question of whether it will be possible to create non-human artificial intelligence art and under what conditions. To achieve the research objectives, but also as a significant contribution, it is possible to consider the acceleration of application possibilities of computational neuroaesthetics, the benefits of which have the potential to influence related methodological approaches even beyond the framework of this research project.
The ambition of the research project is to create a research space at the intersections of neurosciences, aesthetics, artificial intelligence, computer science, and artistic research. By subsuming the aforementioned theoretical starting points about the current state of research, I propose constructing a research project, the definition model of Aesthetics of Artificial Intelligence Art (AAIA), which is an area of computational aesthetics, whose aim is to understand and create artificial systems capable of analyzing and producing aesthetic experiences and enabling their transfer between human and machine. This multidisciplinary field combines computer science, artistic practice, art science, neurosciences, psychology, and philosophy. AI technical means are used in AAIA for the analysis, understanding, and simulation of human thinking, perception simulation of consciousness processes, and creating models of the Artificial Intelligence Art World. AAIA research supports the development of intelligent systems that can enhance creative activity but also allow a deeper understanding of natural and artificial consciousness processes. As already mentioned, one of the key ambitions of the research project is to provide an answer to the question of whether it will be possible to create non-human artificial intelligence art and under what conditions. About this, it was important to offer possibilities on how to approach the formulation of the AIARTWORLD predictive model, which directly continues and expands the research field of the research project.
For the design of predictive criteria, the framework of the Ideal AIARTWORLD Model was used, derived as a paraphrase of Danto’s definition of ArtWorld. Based ultimately on the ideas of general artificial intelligence or superintelligence, it was proposed that the Ideal Model include the elimination of human influence on the production of art, which has its consciousness and which will meet the aesthetic needs of an almost will-less recipient based solely on non-invasive neural analysis. As an estimate of the formulation of future hypotheses about the conditions for fulfilling this model, the outlines of the frameworks and necessary assumptions were formulated.
Since it is clear that the achievement and realization of any Ideal Model are not possible, a research problem arises in the form of the question of how to determine the distance between the current state and the idea of such a model when it does not have established criteria that define it. The Possible Model was proposed as a reference point. The Possible Model derives from the general AIARTWORLD model and has a dynamic character. It includes potential realizability formulated based on current scientific knowledge.For the identification of criteria, it was proposed to formulate a methodology of the so-called state of prediction – approximation thus primarily utilizing the predictive market Metaculus (Metaculus is a scientific platform that maximizes epistemic value. It allows users to predict a wide range of topics)
AIARTWORLD
The attempt to formulate the predictive model AIARTWORLD aims to create a fundamental starting point that will allow for methodological and other research operations with predictive models of future aesthetics about the use of artificial intelligence. The speculative model of future aesthetics is currently an open question, as I believe that closer parameters of modeling will be brought to us by the predictive interaction of AIARTWORLD models as I propose them. I believe that if we are to open discourse or have mental adaptability to changes in social relations in the world with the rising dominance of AI, we must admit the emergence of probably new paradigmatic phenomena in the field of art. By this, I want to indicate that generating a work through AI is still, in some form in 2023, generating a work in the old world of art, whose roots can be found in the post-media age, or in the times of shaping modern institutionalization of art, even in the digital age. Assuming that the social habitus[1] of human actors in ARTWORLD is significantly unchanging, the non-human habitus can be more „tuned“ rather than upgraded, analogously to the human one. In this part, we will deal with conceptual starting points, determinants but also proposals that will serve for the construction of the AIARTWORLD model.
The essay The Artworld[2] by Arthur Danto was published in 1964 and became a key text for understanding the nature and definition of art. Danto introduces the concept of the „world of art“ in which the understanding of art is decomposed into extensive contextual connections in which it is presented and interpreted. Danto offers a „World of art“ where art cannot be fully understood or appreciated without knowledge of the social and cultural context. For completeness, however, it is necessary to add that in opposition to the world of art stands the model of artistic fields enforced by Bourdieu[3], who rejects the concept of context because it obscures the processes of conflict. The most accepted concept appears to be the symbolic interactionism of „social production“ by Howard S. Becker. Becker[4] defines such a world as an art world: „consisting of all interested people and organizations whose activities are necessary for the production of a certain type of events and objects that the art world significantly produces.“ The problem of the world of art may have different methodological starting points, but the core of the problem, which can undoubtedly be sought in subjective human interactions, remains probably without major changes. Even if human art production is replaced by non-human, the human recipient or at least one of the trio of actors, author-producer-recipient, remains in play.
I consider it important to mention the very sketchy description of the world of art that such an idea currently only serves as a dialectical aid, which is more scaffolding than a framework. Certainly, some positions arising mainly from the historical frame of such discourse, such as Dickie’s[5] standpoint that there can be no scientific information relevant to aesthetics, I consider to be a heuristic judgment. Since the neuro-aesthetic perspective is fundamental for the entire research project, I find it complementary or informative to state the claims of Maxwell Bennett and Peter Hacker, who thoroughly argued in their book Philosophical Foundations of Neuroscience[6], that searching for correlates of cognitive activities leads neuroscientists to the mistaken notion that it is not the brain, but the human being, that is embedded in a certain social-cultural environment, because it is the one who sees, feels, and thinks.
„Life 3.0[7]“ or the idea of being human and not just a brain, or a dataistic object in the age of artificial intelligence, demonstrates the defining framework of human living space and advanced AI technologies. In his eponymous book, Tegmark explores the potential impact of artificial intelligence (AI) on society, humanity, and human understanding of life. He introduces the key starting point of his considerations as a classification of life into three stages or levels. Life 1.0 represents biological life, driven by biological evolution and limited by the abilities of biological organisms. Life 2.0 represents cultural life, where humans have gained the ability to shape their environment, create technologies, and modify their bodies and minds. Life 3.0 futuristically defines the era in which we develop general artificial intelligence (AGI) or superintelligence. Tegmark examines and analyzes the socio-technological-ethical contexts and movements associated with the ultimate AGI, which surpasses human abilities while formulating and also urging to formulate questions about potential benefits and risks. He also offers a great deal of speculation about Intelligence Explosion as a hypothetical scenario where AGI surpasses human intelligence and accelerates its development, leading to rapid and potentially uncontrollable growth in intelligence. I consider the opening of the discourse on superintelligence to be the most important contribution of this book. Superintelligence[8] represents a hypothetical form of artificial intelligence (AI) that surpasses human intelligence in practically all aspects. It refers to AI systems that have cognitive abilities significantly exceeding human abilities. In connection with the analysis of the „Terminator“ phenomenon, Tegmark states that the Terminator story does not lie in Skynet taking over the world, but in distracting attention from the real risks and opportunities offered by AI. Here he metaphorically offers a guide to taking over the world in three steps: Step 1 – Create AGI at a human level, Step 2 – Use this AGI to create superintelligence, Step 3 – Use or release this superintelligence to take over the world. Tegmark’s constructions may not be too far from AIARTWORLD, which already operates in a geochronological period that can be named the idea of AI-Anthropo-and-non-Anthropo-cene.
Ideal Model and Model of the Possible
The basic premise for creating a prediction of AIARTWORLD is formulated as a proposal for a methodology for comparing two models of AIARTWORLD. The first Ideal Model is formulated as a model with maximum dominance of superintelligence. The second, Model of the Possible, is defined by technological or scientific conclusions that have real outlines and there is no insurmountable obstacle to realizing them within a certain time horizon.
The general model of AIARTWORLD is derived from the sociological discourse about the world of art. It is formulated as a paraphrase of Danto’s essay[9]. Such a world of art reflects a shared living space, an art field used by processes in the mutual interaction between AI and humans. Subsequently, I propose a possible definition of AIARTWORLD:
AIARTWORLD „World of Artificial Intelligence Art“ refers to a network of human and non-human: actors, institutions, practices, and discourses that participate in the creation, presentation, interpretation, and reception of art, which contains dominant elements of AIART art strategies, or the use of approaches that omit or limit the impact of man in the process of art production and perception.
Attempt to define the boundaries of the worked field
The sociology of art offers a unique lens through which to examine the complex relationship between art and society. The scientific principles of art sociology can formulate the social and cultural forces that shape artistic creation in exact or approximate parameters. It may seem erroneous not to use a sociological reference to the post-media era or digital art. I did not consider it necessary to include the analysis of Digital Sociology[10] or the view of digital art in the post-media era from the position of sociology[11] as far as the formulation of basic models is concerned, since they do not radically differ from those I mention.
The defining framework „Field of AI Art“ can be opened from the perspective of Pierre Bourdieu, who sees it in interaction with its associated economic rules. However, a conceptual problem arises in how to establish these economic rules. Despite this, I propose using the term AI aesthetic capital, as a kind of symbolic capital.
The study Constructing an Artworld Influencer Network by Mining Social Media[12] (2022) is aimed at mapping the course of history, in which ideas about who creates the art world and how its members interact can change, but what remains constant across periods is the phenomenon of the cultural-symbolic value of a work of art, which is socially conditioned and constructed by those who participate in the art world. Conflict or power struggle takes place in various published art-related charts. This study triggers a discussion on how to mine data from social media for the analysis of the art world. It should be noted that social media are key not-entirely-human actors. Also key is the publication Cultural Analytics,[13] which defines data cultural analytics as a new discipline along with the outlining of new terms, tools, and methodologies. Data cultural analytics can be defined as the study of the data image of cultural artifacts, experiences, and dynamics in global space. It can be established that one of the significant conditions, such as the power struggle, takes place in the spirit of advanced datatism.
„Is there already a power struggle in the field of AIART?“, „Certainly yes, but it is necessary to name it.“ In a certain sense, Manovich’s approach mentioned is groundbreaking. Data analysis as Manovich uses it, is comparable to global analyses of world markets. Most important in this context is the search for a wide range of possible correlates of digitized and digital culture, assuming that today the entire visual culture is digital. Based on this, it offers a reflection on trying to identify the conflict-power struggle of non-human actors. I come up with the idea of reconnaissance of such situations, which can help with identifying interactions in the AI art field.
For the definition of a non-human author, the definition is offered: A non-human author is a designation of an entity or AI system that creates creative works without direct human involvement or intervention. However, it seems that the perception of a non-human author and authorship, as well as all actors of the struggle, can be as complex as the perception of the human ones. The concept of the author can and should be understood as a social product for these purposes, especially in Euro-American culture. The roots of this concept[14] as „alter Deus“ go back to antiquity and the Middle Ages. Riegl and Wöllflin are advocates of history without names. Likewise, structuralist attitudes join the aforementioned direction. Bourdieu does not perceive the author as an isolated subject but as a comprehensive understanding of the art field. The artist as a creator fades and yet remains, physically exists in the physical world of art. The belief in the myth of the creator and his exceptionality has remained and continues to remain. Faith as a complex cognitive and emotional phenomenon has attracted human interest for centuries. I suggest that there is a certain correlation between the perception of traditional mythology (materially unprovable) and AI mythology (materially provable, by code, etc.).
Ideal Model of AIARTWORLD
The Ideal Model of AIARTWORLD is based on the idea of minimizing or eliminating human intervention in the production of art, with the maximum possibility of deducting the aesthetic experience and expectations of the recipient, moving them to the position of an inactive perceiver as the owner, or manager of aesthetic capital, produced by AI. Such a model is based on the feasibility of superintelligence and the maximum possible legibility of the recipient’s neurobiological processes, a community of recipients mutually influencing each other in perceiving aesthetic experiences and preferences, as common ownership or availability of aesthetic capital produced by AI. The outline of the definitional framework results from the condition of meeting the mentioned criteria, which have the dimension of science fiction film content. However, I consider hyperbolization to be an inevitable and completely ordinary tool for formulating models. Sets of conditions and criteria are the subject of investigation, but for now, I can assume that they will include: the world in the concept of Life 3.0 and the rise of superintelligence maximized by human fading.[15] The production of art will only come from a package of aesthetic needs or the need for ownership of aesthetic capital.[16] Such a model could offer an artificial experience of the power struggle in the world of art with the intention of gamification[17] without human actors. The result and course of the experience monitoring would be conditioned by aesthetic needs, etc. Society, however, is differentiated, and there is no certainty that, despite the possibility of fulfilling the mentioned criteria, it will be possible to internally accept such a world. I consider it necessary to determine the basic conditions for the feasibility of the Ideal Model:
- Cognitive openness to the acceptance of such a model by existing creators of visual art and recipients.
- Definition of the term inactive recipient, which may also mean involuntary acceptance of the work solely based on the neurobiological calculation and evaluation of the aesthetic need, which may also mean a natural limitation of conscious and subconscious processes. The basic conditions of feasibility, as I have defined them, are only estimates. Their formulation stems from the presumption of the existence of a certain mass closure or user disinterest, as in the case of Google Glasses or 3D televisions. The second factor is the element of fear of the unknown and a high degree of societal neuroticism.[18] The verification and risk framework of such criteria will be included in the detailed definition of the Ideal Model.
Model of the Possible
The Model of the Possible is derived from the general AIARTWORLD model and has a dynamic character. It encompasses potential feasibility formulated based on current scientific knowledge. The aggregation of such data and conclusions from verifiable scientific sources will be performed based on the degree of approximation to the Ideal Model. As an example, I mention the condition of non-invasive thought reading „semantic recoding of the recipient’s cortical data“. Such a method had not been experimentally verified by May 1, 2023, and can only be achieved through functional Magnetic Resonance Imaging (fMRI). For the Ideal Model, it will be important for such a method to be realized without the need for fMRI, etc. The Model of the Possible, already today (June 2023), allows this method to be applied using near-infrared spectroscopy, which means that the fMRI tunnel will be replaced by a sensing cap, and the object under study can move freely. Based on the above, I propose an auxiliary definition of the Model of the Possible:
The Model of the Possible World of AIARTWORLD „The World of Art of Artificial Intelligence“ refers to a network of human and non-human: actors, institutions, practices, and discourses, that contribute to the creation, presentation, interpretation, and reception of art, which contains dominant elements of AIART artistic strategies, or uses approaches that omit or limit the impact of humans in the process of art production and perception. This model is dynamic, and its formulation depends on current scientific knowledge, production of AI art, artistic research, etc.
METHODOLOGICAL BASIS FOR FORMING PREDICTIVE MODELS
The predictability of technological trends in social behavior, or trends in shaping AIARTWORLD, requires a very complex approach, even though the goal is to offer a series of speculative models depending on established or emerging variables. It’s important to note that predicting future AI development is inherently uncertain, and the accuracy of predictions may vary, even if they’re only approximations of the Ideal Model.
I. Definition of the research question and the predicted problem
The main goal of the research is to provide an answer to the question of whether it will be possible to create non-human artificial intelligence art and under what conditions. The basic platform for formulating the central research problem is primarily to detect significant factors that, depending on their mutual interaction, can create a model of artificial consciousness capable of creating art as a unique identity. The research problem can be established at several levels. At the first level, it is the description of the problem: Is the concept, including artificial forms of AI actors in the social field of art such as author-producer-recipient, achievable? The second relational-causal level presents the relationship between humans and AIARTWORLD: What is the role of humans in the AIARTWORLD concept, how will it be influenced by humans, and will it be at all? The third level is already formulated as a problem of forming speculative models of future AIART aesthetics. It’s also necessary to point out the risks associated with the relatively strong and rapid onset of progress in the field of AI. The rapid trends mentioned will likely change during the entire research implementation, as well as the perspective from positions that are directly related to specific technologies and social moods.
II. Selection of the time horizon
The time horizon has a fluid character, it does not offer a prediction for a certain period, but only the relationship of distance from the Model of the Possible to the Ideal Model of AIARTWORLD. However, the Ideal Model is unachievable because it is formulated within ultimate criteria. However, this does not mean that the criteria will change, thereby setting time horizons. The relationship mentioned currently demarcates the speculative predictive interval in which the articulation of speculative models is located.
III. Collection and analysis of published and unpublished data and studies.
a) Selection of all appropriate studies according to the established conditions of mutual relationships, extraction of data, assessment of data homogeneity, effort to increase it, and integration of the model of mutual relationships, by the Prisma method. This method is based on the principle of Preferred Reporting Items for Systematic Reviews and Meta-Analyses, or established PRISMA. It is a minimal set of items based on reporting in the context of systematic reviews and meta-analyses, as a set of guidelines used to report on systematic reviews and meta-analyses in the field of research and evidence synthesis. This method was developed to improve transparency, completeness, and accuracy in reporting these types of studies. Systematic reviews and meta-analyses are research methods used to summarize and synthesize findings from multiple studies on a specific research question or topic. Used studies include a comprehensive and systematic review of existing literature, selection, and evaluation of relevant studies, and statistical analysis of data from these studies. The methodology and criteria for reporting will depend on many factors, including technological progress, discoveries in research, political decisions, societal influences, and unforeseen events. The methodology and criteria for reporting will depend on a multitude of factors, including technological progress, discoveries in research, political decisions, societal influences, and unforeseen events. Among the presented methods for predicting the development of AIARTWORLD, I include areas such as AI, art, neurosciences, sociology, and their relevant outputs. These include the aggregation of knowledge and conclusions from researchers and industry experts, as well as AI influencers who can provide valuable perspectives on the future of AI; reports on ongoing research on new technologies and trends in various AI technologies such as machine learning, natural language processing, computer vision, and robotics; patent analysis; academic publications, etc. Through an open or predefined analysis for the research project, a data system will be created that will be usable to ensure that the delivery of subsumptive conclusions is systematic, consistent, and transparent about meta-analyses.
IV. ANALYSIS AND HYPOTHESIS PROCESSING
As previously stated, the sources for decision-making in the selection of criteria will be derived from a) Assessment of data extraction, data homogeneity assessment, efforts to increase it, and integration of the model of mutual relationships in AIARTWORLD models and the results in the experimental part of the research project. b) Creation of a predictive question as a sub-model of AIARTWORLD. For example: „Can art experiment XXX bring XXX about artificial consciousness XXX?“.
V. PLACING THE PREDICTIVE QUESTION AS A SPECULATIVE MODEL ON THE PREDICTIVE MARKET OF METACULUS
The placement of the predictive question will be on the online platform Metaculus [20] as a predictive market, which focuses on collective forecasting or formulating predictions. However, labeling Metaculus[21] as a predictive market may not be entirely accurate. The principle of the market is to determine the price, and the actors always strive to maximize profits. The purpose of science is to develop knowledge of the world and the actors, typically scientists and researchers, but also users, who are motivated to find the truth. Predictive markets combine these two missions and motivations. Metaculus is a scientific platform specially designed to maximize epistemic value, not monetary value. It allows users to predict a wide range of topics, such as technology, science, politics, and more. Metaculus uses a combination of Crowd Wisdom and statistical modeling for aggregating predictions and generating predictive models for various events and outcomes. In addition to using prediction aggregation, Metaculus allows users to:
- create their predictions by assigning probabilities to various outcomes
- derive evidence to support their predictions
- engage in discussions to mutually question their predictions and update their predictions based on new information
- participate in tournaments aimed at specific events or discourses, thus demonstrating their prognostic abilities.
The popularity of the Metaculus platform is mainly seen among those who are interested in the intersection of technology, science, and future trends. Naturally, the resulting aim of Metaculus can be characterized as introducing procedures that analyze user prediction performance over time and assign scores based on accuracy. The platform architecture is inspired by Bayesian statistics, ensemble modeling, time series analysis, machine learning, and related techniques. The creators of Metaculus cooperate with researchers to create hybrid human ML models that match or surpass the state-of-the-art predictive systems. Long-term collaboration with researchers from the Institute of Biocomplexity at the University of Virginia, as well as with Tom McAndrew’s biostatistics laboratory at Lehigh University, has led to a multitude of promising experiments using human judgment predictions for the parametrization of computational models.
VI. CONSTRUCTION OF FUTURE DEVELOPMENT SCENARIOS.
The construction of future scenarios will be derived from:
a) Recommendations of the predictive market,
b) Modified formulation of the predictive question on the predictive market and its repositioning,
c) Free interpretation based on the collection of recommendation data.
VII. VERIFICATION OF FORECASTS ON THE PREDICTIVE MARKET OF METACULUS AND USE OF THE DELPHI METHOD
The verification of the forecast in relation to the idea AIARTWORLD method and future aesthetics will be realized as follows: a) Extraction of data about similar or related forecasts b) Using the Delphi Method[22] (Delft method), which is structured as an iterative forecasting or decision-making technique and involves the collection and synthesis of inputs from a group of experts to make predictions, reach consensus, or make informed decisions. The method is designed to minimize biases and facilitate the convergence of expert opinions. The Delphi method allows experts to anonymously share their views, which helps to mitigate biases caused by the dominance of certain individuals or group dynamics. It also encourages open discussion and reassessment of initial positions, leading to the convergence of expert opinions and the identification of areas of consensus or disagreement. The Delphi method is commonly used in various fields including technology forecasting, policy making, risk assessment, strategic planning, and decision-making in healthcare. It is particularly useful when dealing with complex and uncertain problems where expert opinions and insights are valuable for decision-making or prediction purposes.
IDENTIFICATION OF SOURCES FOR THE PURPOSE OF SETTING SEARCH CLASSIFICATION AND DATA AGGREGATION FOR THE METHODS: PRISMA, DELPHI, AND METACULUS.
Current knowledge in fields such as AI-related technologies, artificial consciousness research, and research on the limits of human cognitive abilities is a key data source in shaping potential AIARTWORLD models and the Ideal AIARTWORLD Model. This section focuses primarily on the identification of fundamental knowledge sources. The identification of sources and their subsequent free evaluation aims to lay the foundation for the articulation of conditions, search classifications, data aggregation, and conclusions for the use of the Prisma or Delphi method. The principle of free evaluation will also be important for setting predictive visions for Metaculus.
The Impact of Technologies and AI For the purposes of using the predictive methodology, I cite the study Latest Trends in Artificial Intelligence Technology: A Scoping Review (2023) [23], which examines the scope of current cutting-edge artificial intelligence technologies using the PRISMA methodology. The aim of this study was to evaluate the most advanced technologies used in various areas of artificial intelligence technology research. For the stated purposes, 3 journals exploring the areas of AI and machine learning published during 2023 were used: Journal of Artificial Intelligence Research[24], Journal of Machine Learning Research[25], and Machine Learning[26]. Certain qualifications were set for technological solutions: • The technology must be tested against comparable solutions, • Standard approved or otherwise well-justified data sets must be used in the application, • The results must show improvements over comparable solutions.
Data can be highly unstructured and the technological solution should be able to utilize data with minimal manual human labor. The results of this review suggest that creating labeled datasets is very labor-intensive and solutions using unsupervised learning or semi-supervised learning technologies are increasingly being investigated. Learning algorithms should be able to be efficiently updated and predictions should be interpretable. When using artificial intelligence technologies in real-world applications, safety and explainable predictions need to be considered before mass adoption takes place.
For the purposes of applying in the research project Art Aesthetics of Artificial Intelligence in relation to the study mentioned, the procedure was modified and repeated. For the extraction of data sources, the Prisma method was freely reformulated and reused for 2023 in connection with previous periods. The sources were expanded to include the journals: NeuroImage,[27] Journal of Cognitive Neuroscience,[28] Frontiers in Human Neuroscience,[29] Creativity Research Journal,[30] Neurocomputing.[31] The repeated analysis showed very similar results, which helped to determine the basic sources of information for the proposed implementation of the Prisma method in the research project’s intentions. Among all the studies that contain agreement and examine the central research problem from another perspective, I cite the study on the imitation of animal or human cognition.[32] The study Detect, Understand, Act: A Neuro-symbolic Hierarchical Reinforcement Learning Framework (2023) is also focused on testing Animal-AI (AAI) and referencing the study The Animal-AI Olympics (2019),[33] which examines Animal-level-AI as a springboard to the path of human-level AI.
However, the approach mentioned significantly aided in the aggregation of sources on the research topic of Deep Image Reconstruction, loosely translated as the reconstruction of an image based on records of human cortical activity. This is a progressive multidisciplinary field of research whose goal is to transfer and reconstruct the experienced visual content from records of measurable activities such as blood oxygenation or electric signals created in the human brain. The data comparison of studies: Inter-individual deep image reconstruction via hierarchical neural code conversion (2023), Attention modulates neural representation to render reconstructions according to subjective appearance (2020), Within-participant statistics for cognitive science (2022), with sources studying a similar principle, but with the possibility of replacing fMRI with a mobile approach via fNIRS, was not very successful. The answer to the problem, however, was known by ChatGPT model 4: „Replacing fMRI with fNIRS in the context of deep image reconstruction would certainly pose new challenges and opportunities.“
Artificial Consciousness
Artificial consciousness is important to understand as a theoretical-speculative model, which is the subject of a multidisciplinary field of research. In the last decade, research has made significant progress. The challenge, as well as the defining criteria related to artificial consciousness as a machine with real conscious experiences, present an open discourse. Interdisciplinary collaboration is mainly formed in the field of AI, neuroscience, and philosophy, as well as medical sciences. For the purposes of the research project, I propose bypassing the Prisma aggregation platform, as only a few institutions address the subject matter, and therefore I focus the mapping of scientific knowledge mainly on:
I. Integrated Information Theory[34] (IIT) is a theoretical framework developed by Giulio Tononi that aims to explain the nature of consciousness. According to IIT, consciousness arises as a result of integrated information generated by the brain. Information is considered integrated if it cannot be broken down into separate parts, thus they must or can be deducible when examining the brain. IIT also suggests that consciousness is a property of any system that generates integrated information. This means it is not limited to biological systems but can also exist in artificial systems capable of generating integrated information.
II. Cognitive architectures, such as the Global Workspace Theory[35] and ACT-R[36] (Adaptive Control of Thought-Rational), provide frameworks for understanding higher-level cognitive processes and have been extended to include aspects of consciousness. These architectures aim to capture the mechanisms underlying consciousness by modeling information processing and decision-making. The Global Workspace Theory is a simple cognitive architecture that qualitatively considered a large set of conscious and unconscious processes. The goal of ACT-R is to define the fundamental and irreducible cognitive and perceptual operations as operational models, allowing the functions of the human mind. The representation of such a model lies in the fact that every task that people can perform consists of discrete operations. Thus, the brain itself is organized in a way that allows individual processing modules to produce knowledge.
III. Simulation Neuroscience[37] is a field of science with deep historical roots and, aside from its extremely natural scientific approach, it also inspires by opening various philosophical discourses. Based on this, we can consider simulation neuroscience as a multidisciplinary field at the intersection of neuroscience, computer science, and mathematics. It has proven to be an effective approach to uncovering the complicated functioning of the brain. Extensive brain simulations, such as the Human Brain Project, aim to replicate the entire structure and dynamics of the brain. The Human Brain Project[38] (HBP) is a large-scale scientific research project planned for decades, based on supercomputers, aiming to design and implement an AI-based scientific research infrastructure that will enable researchers across Europe to expand knowledge in the fields of neuroscience, computing, and brain-related medicine. These simulations integrate a vast amount of experimental data and use supercomputers to model neuronal activity, synaptic connections, and network dynamics, providing a holistic view of brain function.
MAP OF POSSIBILITIES FOR SURPASSING HUMAN COGNITIVE ABILITIES.
Current research shows that some human cognitive limits are considered insurmountable according to current scientific views. For example, hippocampal structures in relation to Time cells (perception of time) cannot be modified. Humans cannot perceive time, for instance, to the thousandths of a second. However, factors of representation and imagination can create an impression that there is an expansion of consciousness. The possibilities for overcoming such limits are important to establish in three basic models. The first model represents the field of accepting the possibility of surpassing human cognitive limits. The second is the biological-psychological model of current knowledge about human cognitive limits. The third model represents theoretically feasible ways to overcome cognitive limits through simulation neuroscience. The fourth model is constituted as speculative using neuromorphics.
MODEL OF ACCEPTANCE FOR OVERCOMING HUMAN COGNITIVE LIMITS: Why is it meaningful to analyze „New mysterians“?
The contribution The end of science? On human cognitive limitations and how to overcome them (2020) offers an extensive conceptual set of tools discussing forms and variations of human cognitive limitations associated with the so-called „New mysterians“[39]. New mysterians believe that there are fundamental cognitive aspects of human consciousness and subjective experience that cannot be fully explained or reduced to precisely describable processes. Thomas Nagel[40], one of the founders of this movement, argues that certain subjective experiences cannot be fully understood through purely objective scientific explanations. He suggests that there is a chasm between the physical processes in the brain and the subjective experience itself. For Nagel, this chasm is a mystery that science may never fully solve, hence the name „New mysterians“. They question the idea that all aspects of human consciousness can ultimately be explained by neuroscience or other scientific disciplines. They argue that subjective experience poses a unique problem for scientific reductionism. The mentioned study describes a comparison between the representative approach of science and the imaginative representation of the impossibility of overcoming limits. I also see an important contribution in outlining an analogy between various modalities (hard vs. soft) of cognitive limitation. I believe it is necessary to understand New mysterians as a cultural-social phenomenon of mass anxiety about the endangerment[41],[42] of the human species, which was created with intentions close to religion, offering the ultimate answer „it is not possible, and if so, only a little“. I see the importance of the mentioned contribution in two planes. The argumentation model of the impossible is located in the first plane. Neurosciences deal with conceptual problems of understanding brain functions and delimiting fundamentally probable physiological limitations. The proposal for an upgrade is only of marginal interest. The second plane is important for the research project in that it seems necessary to express speculative considerations about the causal kinship of the phenomenon of mass anxiety about endangerment of New mysterians and the social group in which AI agnostics are located.
The study „Psychological Closure Does Not Entail Cognitive Closure“ (2017) [43], which the mentioned study refers to, discusses the problem of psychological closure. The core of the research problem here stems from some philosophical postulates about „cognitive closure“ in response to certain problems. It reveals McGinn’s fundamental flaw in deriving the representational conclusion of psychological closure, which lies in the error of ambiguity regarding the term „understanding“. McGinn and Chomsky operate with the consideration that humans are „cognitively closed“ to the quantum world. Even such mysterian theses can be refuted, as precise scientific knowledge and descriptions have already been developed for areas to which we are supposedly „closed“, such as the mind in relation to the body, as well as the scientific model of quantum mechanics. Mathematics, for example, uses a tesseract, that is, the concept of a cube from three dimensions to four dimensions. Although it’s hard to quickly envision a tesseract, because we live in a three-dimensional world, we can use analogies and projections to gain some understanding. Mathematicians can imagine what a tesseract looks like, in the same way that any of us can imagine a cube in front of our mind’s eye. A similar analogy can be found in the representation of spacetime curvature and so on. There is often a confusion between cognitive and perceptual closure, which is present in relation to a large part of the scientifically provable part of physical laws. If there really is a boundary to human knowledge (representative or imaginative), it’s because there is a human need to invoke the cult of ultimacy. To explain the scientific and also the new mysterian approach to human representation and imagination, as well as to understand the world, it will probably be necessary to use the measure of the need for cognitive closure [44], which relates to the individual’s desire for definitive answers, certainty, and closure in cognitive processes.
BIOLOGICAL-PSYCHOLOGICAL MODEL: current knowledge about human cognitive limits
The principles of cognitive limits represent a concept that acknowledges the natural limitations of human mental cognition and knowledge or judgment formation. Human cognitive abilities are bounded or limited, resulting in an influence on perception, attention, memory, and decision-making. Natural or pathological limits can be divided into morphological and functional, which is important to import into the model from the position of normal as well as pathological states. Concurrent subjects of study are ethical determinants: I suggest focusing the criteria for aggregating knowledge on: the issue of human cognitive capacity in relation to imagination, opening a discourse on basic problems of representation, and comparison of intension of a representation and task space [45]. The research problem relating to the above can be characterized by a question aimed in two directions. The first direction represents positively verified possibilities of invasive or non-invasive expansion of these possibilities. The second direction consists of the so-called analysis of fear of the unknown.
METHODOLOGICAL APPROACH TO MODELS 3 AND 4
The third model represents theoretically feasible ways of exceeding cognitive limits through simulation neuroscience to establish a methodological framework, which will be approached only after the extraction of knowledge and the establishment of starting points in the part of simulation neuroscience. The fourth model is constituted as speculative using neuromorphic computing. An important factor in building this model is the analysis of the first neuromorphic experiments on humans, planned to be conducted by the company Neuralink [46].
CONCLUSION
Preliminarily, it can be determined that the working field, which is placed on the platform of sociology of art, provides a solid space for the articulation of criteria for AIARTWORLD models. A fundamental prerequisite for establishing the Ideal Model seems to be the hypothetical possibility of human acceptance of criteria, which minimize or even eliminate human influence on art production. The perception of art production was different in the Middle Ages or Antiquity than it is today. The question could be whether the questions and models are properly articulated and formed. The aesthetic experience can be, unlike the perception of sexual attraction (I don’t mean attractiveness), predominantly a matter of social convention. The research problem will likely be formulated in intentions that will not submit whether the Ideal Model of AIARTWORLD will be realizable, but how the subjective or social interpretation of this model will be accepted. For testing the above, proposals for experiments, as art research, are offered that could bring a readable reaction of acceptance by artists or creative individuals. The proposals should incorporate latent questions that are necessary for the realization of the Ideal Model of AIARTWORLD: Is it possible for a person to accept a work created by a non-human? Can a person accept the myth of a non-human author? What difference in perception of natural mythology (created by man – materially unprovable) and AI mythology (materially provable, with code, etc.)? Proposals for experimental-art research are approaching the cautious formulation of paths and ideas about how to look at creating AI mythology. AI mythology refers to myths, misconceptions, and exaggerated beliefs about artificial intelligence (AI). Since AI is already creating human space, which has triggered a multitude of myths and misconceptions that create unrealistic expectations or unwarranted fears: AI will replace all human jobs, AI will surpass human intelligence and control the world, Artificial Intelligence is infallible and unbiased. The second support pillar comes from the study of current artistic practice, based on which artistic strategies were identified that have the potential to derive experiments. The selection should be made only by empirical knowledge and should consider the stated criteria that stood in many intersections between the theoretical and experimental position. Successful realization could represent a successful step towards the formulation of predictive criteria, while it is possible to apply conclusions or a description of emerging phenomena that emerged during the experimental-art research. The invisible research potential of art could make a significant contribution to formulating research goals and identifying research problems.
PPREFRENCES:
[1] BOURDIEU, Pierre a Loïc J. D. WACQUANT. An invitation to reflexive sociology. Chicago: University of Chicago Press, 1992. ISBN 9780226067407.
[2] DANTO, Arthur. The Artworld. The Journal of Philosophy [online]. 1964, 61(19) [cit. 2023-06-11]. ISSN 0022362X. Dostupné z: doi:10.2307/2022937
[3] BOURDIEU, P.P. Sociology in Question: Published in Association with Theory, Culture & Society. Sage Publications (CA), 1993. ISBN 9781446236840
[4] BECKER, H.S. Art Worlds. University of California Press, 1982,41. ISBN 9780520052185.
[5] DICKIE, George. Is Psychology Relevant to Aesthetics?. The Philosophical Review [online]. 1962, 71(3), 285–302 [cit. 2023-06-15]. ISSN 00318108. Dostupné z: doi:10.2307/2183429
[6] BENNETT, M.R. a P.M.S. HACKER. Philosophical Foundations of Neuroscience. Wiley, 2021. ISBN 9781119530633.
[7] TEGMARK, Max. Život 3.0: člověk v éře umělé inteligence. Přeložil Markéta IVÁNKOVÁ. Praha: Argo, 2020. Zip (Argo: Dokořán): Dokořán). ISBN 978-80-7363-948-8.
[8] BOSTROM, Nick. Superintelligence: paths, dangers, strategies. 44. Oxford: Oxford University Press, 2014. ISBN 978-0199678112.
[9] Idem: [2]
[10] MARRES, Noortje. Digital Sociology: The Reinvention of Social Research. 1. Warvick: Wiley, 2017. ISBN 978-0-745-68482-6.
[11] GILBERT, Tamsyn. Looking at Digital Art: Towards a Visual Methodology for Digital Sociology. The American Sociologist [online]. 2018, 49(4), 569-579 [cit. 2023-06-16]. ISSN 0003-1232. Dostupné z: doi:10.1007/s12108-018-9384-2
[12] FOKA, Amalia. Constructing an Artworld Influencer Network by Mining Social Media. Leonardo [online]. 2022, 55(1), 24-29. ISSN 0024094X. Dostupné z: doi:10.1162/leon_a_02094
[13] MANOVICH, Lev. Cultural Analytics [online]. The MIT Press, 2020 [cit. 2022-03-09]. ISBN 9780262360647. Dostupné z: https://direct.mit.edu/books/book/4966/Cultural-Analytics
[14] TANNER, Jeremy, ed. Sociology of Art [online]. 1. London: Routledge, 14 August 2003n. l. [cit. 2023-06-16]. ISBN 9781134393305. Dostupné z: doi:10.4324/9780203633649
[15] Human fading ako stav minimalizácie ľudské vplyvu na AIARTWORLD.
[16] Estetický kapitál ako druh symbolického kapitálu
[17] Gamification. Wikipedia: the free encyclopedia [online]. San Francisco (CA): Wikimedia Foundation, 2001- [cit. 2022-03-04]. Dostupné z: https://en.wikipedia.org/wiki/Gamification
[18] Frankl (2006) otvoril diskurz o „masovej neurotickej triáde“ agresie, depresie a závislosti, ku ktorej dochádza, keď jednotlivci zažívajú existenciálne vákuum. Toto vákuum vedie k porušovaniu sociálnych noriem, symptómom stresu a závislosti.
[19] Prisma. Www.prisma-statement.org/ [online]. Dostupné z: www.prisma-statement.org/
[20] Metaculus [online]. [cit. 2023-06-17]. Dostupné z: https://www.metaculus.com
[21] Why I Reject the Comparison of Metaculus to Prediction Markets [online]. In: . 24,2,2023. Dostupné z: https://metaculus.medium.com
[22] WELTY, Gordon. Plato and Delphi. Futures [online]. 1973, 5(3), 281-286 [cit. 2023-06-10]. ISSN 00163287. Dostupné z: doi:10.1016/0016-3287(73)90066-9
[23] NISKANEN, Teemu, Tuomo Sipola SIPOLA a Olli VÄÄNÄNEN. Latest Trends in Artificial Intelligence Technology: A Scoping Review. In: Https://arxiv.org/ [online]. Jamk: Institute of Information Technology, Jamk University of Applied Sciences, 2023 [cit. 2023-06-10]. Dostupné z: https://arxiv.org/abs/2305.04532
[24] Journal of Artificial Intelligence Research. 29. 2022. ISSN 1076 – 9757.
[25] Journal of Machine Learning Research. 20. 2022. ISSN 1533-7928.
[26] Journal of Machine Learning Research. Www.springer.com [online]. [cit. 2023-06-17]. Dostupné z: https://www.springer.com/journal/10994/
[27] NeuroImage. 2022-2023. ISSN 1095-9572.
[28] Journal of Cognitive Neuroscience [online]. [cit. 2023-06-20]. ISSN 1530-8898.
[29] Frontiers: Frontiers in Human Neuroscience [online]. 2023 [cit. 2023-06-20]. Dostupné z: https://www.frontiersin.org/journals/human-neuroscience
[30] Creativity Research Journal [online]. 2023 [cit. 2023-06-20]. ISSN 1532-6934. Dostupné z: https://www.tandfonline.com/journals/hcrj20
[31] Creativity Research Journal [online]. 2023 [cit. 2023-06-20]. ISSN 0925-2312. Dostupné z: https://www.sciencedirect.com/journal/neurocomputing
[32] MITCHENER, Ludovico, David TUCKEY, Matthew CROSBY a Alessandra RUSSO. Detect, Understand, Act: A Neuro-symbolic Hierarchical Reinforcement Learning Framework. Machine Learning [online]. 2022, 111(4), 1523-1549 [cit. 2023-06-10]. ISSN 0885-6125. Dostupné z: doi:10.1007/s10994-022-06142-7
[33] CROSBY, Matthew, Benjamin BEYRET a Marta HALINA. The Animal-AI Olympics. Nature Machine Intelligence [online]. 2019, 1(5), 257-257 [cit. 2023-06-17]. ISSN 2522-5839. Dostupné z: doi:10.1038/s42256-019-0050-3
[34] TONONI, Giulio a Christof KOCH. Consciousness: here, there and everywhere?. Philosophical Transactions of the Royal Society B: Biological Sciences [online]. 2015, 370(1668) [cit. 2022-11-20]. ISSN 0962-8436. Dostupné z: doi:10.1098/rstb.2014.0167
[35] BAARS, Bernard J., Natalie GELD a Robert KOZMA. Global Workspace Theory (GWT) and Prefrontal Cortex: Recent Developments. Frontiers in Psychology [online]. 2021, 12 [cit. 2023-06-14]. ISSN 1664-1078. Dostupné z: doi:10.3389/fpsyg.2021.749868
[36] SINCLAIR, Jacob, Hemmaphan SUWANWIWAT a Ickjai LEE. A hybrid data gathering and agent based cognitive architecture for realistic crowd simulations. Journal of Simulation [online]. 2023, 17(2), 121-148 [cit. 2023-06-17]. ISSN 1747-7778. Dostupné z: doi:10.1080/17477778.2021.1954487
[37] FAN, Xue a Henry MARKRAM. A Brief History of Simulation Neuroscience. Frontiers in Neuroinformatics [online]. 2019, 13 [cit. 2023-06-14]. ISSN 1662-5196. Dostupné z: doi:10.3389/fninf.2019.00032
[38] AICARDI, Christine a Tara MAHFOUD. Formal and Informal Infrastructures of Collaboration in the Human Brain Project. Science, Technology, & Human Values [online]. [cit. 2023-06-17]. ISSN 0162-2439. Dostupné z: doi:10.1177/01622439221123835
[39] New mysterianism. Wikipedia: the free encyclopedia [online]. San Francisco (CA): Wikimedia Foundation, 2001- [cit. 2022-03-11]. Dostupné z: https://en.wikipedia.org/wiki/New_mysterianism
[40] NAGEL, Thomas. Mind and cosmos: why the materialist neo-Darwinian conception of nature is almost certainly false. New York: Oxford University Press, c2012. ISBN 978-0199919758.
[41] WESSELY, Simon. Mass hysteria: two syndromes?. Psychological Medicine [online]. 1987, 17(1), 109-120 [cit. 2023-06-16]. ISSN 0033-2917. Dostupné z: doi:10.1017/S0033291700013027
[42] CHENG, Cecilia, Hsin-yi WANG a Linus CHAN. Multiple forms of mass anxiety in coronavirus disease-2019 pandemic. Journal of Affective Disorders [online]. 2021, 291, 338-343 [cit. 2023-06-16]. ISSN 01650327. Dostupné z: doi:10.1016/j.jad.2021.05.034
[43] VLERICK, Michael a Maarten BOUDRY. Psychological Closure Does Not Entail Cognitive Closure. Dialectica [online]. 2017, 71(1), 101-115 [cit. 2023-06-16]. ISSN 00122017. Dostupné z: doi:10.1111/1746-8361.12176
[44] WEBSTER, Donna M. a Arie W. KRUGLANSKI. Cognitive and Social Consequences of the Need for Cognitive Closure. European Review of Social Psychology [online]. 1997, 8(1), 133-173 [cit. 2023-06-16]. ISSN 1046-3283. Dostupné z: doi:10.1080/14792779643000100
[45] KLEINSORGE, Thomas. Cognitive Capacity, Representation, and Instruction. Frontiers in Psychology [online]. 2021, 12 [cit. 2023-06-16]. ISSN 1664-1078. Dostupné z: doi:10.3389/fpsyg.2021.701687
[46] FITZGERALD, James. Neuralink: Elon Musk’s brain chip firm wins US approval for human study. In: Bbc.com [online]. London: BBC, 2023, 26.5.2023 [cit. 2023-06-16]. Dostupné z: https://www.bbc.com