Field Of AI Research Articles

What Should we Reasonably Expect from Artificial Intelligence?

Objective: the objective of this article is to address the misalignment between the expectations of Artificial Intelligence (or just AI) systems and what they can currently deliver. Despite being a pervasive and cutting-edge technology present in various sectors, such as agriculture, industry, commerce, education, professional services, smart cities, and cyber defense, there exists a discrepancy between the results some people anticipate from AI and its current capabilities. This misalignment leads to two undesirable outcomes: Firstly, some individuals expect AI to achieve results beyond its current developmental stage, resulting in unrealistic demands. Secondly, there is dissatisfaction with AI's existing capabilities, even though they may be sufficient in many contexts.Methods: the article employs an analytical approach to tackle the misalignment issue, analyzing various market applications of AI and unveils their diversity, demonstrating that AI is not a homogeneous, singular concept. Instead, it encompasses a wide range of sector-specific applications, each serving distinct purposes, possessing inherent risks, and aiming for specific accuracy levels.Results: the primary finding presented in this article is that the misalignment between expectations and actual AI capabilities arises from the mistaken premise that AI systems should consistently achieve accuracy rates far surpassing human standards, regardless of the context. By delving into different market applications, the author advocates for evaluating AI's potential and accepted levels of accuracy and transparency in a context-dependent manner. The results highlight that each AI application should have different accuracy and transparency targets, tailored on a case-by-case basis. Consequently, AI systems can still be valuable and welcomed in various contexts, even if they offer accuracy or transparency rates lower or much lower than human standards.Scientific novelty: the scientific novelty of this article lies in challenging the widely held misconception that AI should always operate with superhuman accuracy and transparency in all scenarios. By unraveling the diversity of AI applications and their purposes, the author introduces a fresh perspective, emphasizing that expectations and evaluations should be contextualized and adapted to the specific use case of AI.Practical significance: the practical significance of this article lies in providing valuable guidance to stakeholders within the AI field, including regulators, developers, and customers. The article's realignment of expectations based on context fosters informed decision-making and promotes responsible AI development and implementation. It seeks to enhance the overall utilization and acceptance of AI technologies by promoting a realistic understanding of AI's capabilities and limitations in different contexts. By offering more comprehensive guidance, the article aims to support the establishment of robust regulatory frameworks and promote the responsible deployment of AI systems, contributing to the improvement of AI applications in diverse sectors. The author's call for fine-tuned expectations aims to prevent dissatisfaction arising from unrealistic demands and provide solid guidance for AI development and regulation.

Evaluating the quality of visual explanations on chest X-ray images for thorax diseases classification

AbstractDeep learning models are extensively used but often lack transparency due to their complex internal mechanics. To bridge this gap, the field of explainable AI (XAI) strives to make these models more interpretable. However, a significant obstacle in XAI is the absence of quantifiable metrics for evaluating explanation quality. Existing techniques, reliant on manual assessment or inadequate metrics, face limitations in scalability, reproducibility, and trustworthiness. Recognizing these issues, the current study specifically addresses the quality assessment of visual explanations in medical imaging, where interpretability profoundly influences diagnostic accuracy and trust in AI-assisted decisions. Introducing novel criteria such as informativeness, localization, coverage, multi-target capturing, and proportionality, this work presents a comprehensive method for the objective assessment of various explainability algorithms. These newly introduced criteria aid in identifying optimal evaluation metrics. The study expands the domain’s analytical toolkit by examining existing metrics, which have been prevalent in recent works for similar applications, and proposing new ones. Rigorous analysis led to selecting Jensen–Shannon divergence (JS_DIV) as the most effective metric for visual explanation quality. Applied to the multi-label, multi-class diagnosis of thoracic diseases using a trained classifier on the CheXpert dataset, local interpretable model-agnostic explanations (LIME) with diverse segmentation strategies interpret the classifier’s decisions. A qualitative analysis on an unseen subset of the VinDr-CXR dataset evaluates these metrics, confirming JS_DIV’s superiority. The subsequent quantitative analysis optimizes LIME’s hyper-parameters and benchmarks its performance across various segmentation algorithms, underscoring the utility of an objective assessment metric in practical applications.

Active learning inspired method in generative models

In the decade, researchers have proposed many remarkable algorithms in structural design, training modes, etc., in the field of Generative AI. However, with the explosive growth of demand for data and annotation in generative models and the emphasis on data-centric algorithms, improving model performance by improving data-efficiency has become another way for the future development of generative models. In this paper, we propose a new strategy to improve the performance of the generative models inspired by active learning. Through the cognitive feature extractor of generative models (similar oracle in active learning) to complete the annotation of data informativeness. By guiding the generative models to focus on samples in low-feature density region, i.e., with high informativeness, the models can gradually achieve “full-cognition” of the data. We conduct extensive experiments to verify the effectiveness of our strategy and demonstrate its widespread application in generative models and downstream data-driven tasks.

Information retrieval from textual data: Harnessing large language models, retrieval augmented generation and prompt engineering

This paper describes how recent advancements in the field of Generative AI (GenAI), and more specifically large language models (LLMs), are incorporated into a practical application that solves the widespread and relevant business problem of information retrieval from textual data in PDF format: searching through legal texts, financial reports, research articles and so on. Marketing research, for example, often requires reading through hundreds of pages of financial reports to extract key information for research on competitors, partners, markets and prospective clients. It is a manual, error-prone and time-consuming task for marketers, where until recently there was little scope for automation, optimisation and scaling. The application we have developed combines LLMs with a retrieval augmented generation (RAG) architecture and prompt engineering to make this process more efficient. We have developed a chatbot that allows the user to upload multiple PDF documents and obtain a summary of predefined key areas as well as to ask specific questions and get answers from the combined documents’ content. The application’s architecture begins with the creation of an index for each of the PDF files. This index includes embedding the textual content and constructing a vector store. A query engine, employing a small-to-big retrieval method, is then used to accurately respond to a set of predefined questions for each PDF to create the summary. The prompt has been designed in a manner that minimises the risk of hallucination which is common in this type of model. The user interacts with the model via a chatbot feature. It utilises similar small-to-big retrieval techniques over the indices for straightforward queries, and a more complex sub-questions engine for in-depth analysis, providing a comprehensive and interactive tool for document analysis. We have estimated that the implementation of this tool would reduce the time spent on manual research tasks by around 60 per cent, based on the discussions we have had with potential users.

The Application of Convolutional Neural Networks in Organ-on-a-Chip Technology: A Review

Recent developments in microfluidics and biomaterials have enabled the creation of organs-on-chips (OoCs), which provide a controlled and lifelike in vitro microenvironment resembling an actual organ. Organ-on-a-chip technology offers a tremendous opportunity for efficient, cost-effective, and ethical drug testing and research. However, the high throughput and large quantities of data these complex microenvironments produce make it difficult for researchers to effectively analyze the data and draw valuable conclusions. Convolutional Neural Networks (CNNs) are most aptly positioned to address many of OoC’s challenges because of their ability to interpret microscopic images and facilitate the analytical process. Despite the growing field of AI, there have been a limited number of studies summarizing the various applications of CNNs in OoCs. This review aims to provide 1) an overview of the technology involved with CNNs and OoCs 2) an insight into the state-of-the-art applications of CNNs in OoCs including device parameters, predicting and tracking cell trajectories, super-resolution image segmentation, and image classification, and 3) an overview of existing challenges and opportunities ahead for clinical translation of this technology. Various applications of CNNs have been classified by the type of task. Different CNN models such as Faster R-CNNs, fully convolutional networks (FCNs), and Mask R-CNNs are explained and highlighted. This review article can be used as a resource for a better understanding of the potential of CNNs in biomedical research and clinical applications, particularly OoCs.

Empirical legal analysis simplified: reducing complexity through automatic identification and evaluation of legally relevant factors.

This paper investigates the potential for reducing the complexity of AI and Law and empirical legal studies projects through a novel annotation methodology that relies on GPT Family Models to assist human annotators. Improving the speed, cost and quality of annotation could greatly benefit such projects. In modelling types of legal claims, researchers in the fields of empirical legal studies and AI and Law have long relied on manually annotating factors in case texts. To demonstrate our methodology, we employ cases and factors regarding whether a police officer has constitutional authority to detain a motorist on the basis of the officer's suspicion that the motorist is trafficking drugs. Our results demonstrate how recent advances in text analytics can reduce the burden of identifying factors in large numbers of cases and improve machine learning models' predictions of case outcomes. This article is part of the theme issue 'A complexity science approach to law and governance'.

Intelligence and Consciousness

One can trace a possible leap from intelligence to consciousness in the debates in and around the field of AI. We argue that this leap is a result of an unscrutinized assumption. We attempt to analyze the relationship between intelligence and consciousness by proposing a stripped-down working definition of intelligence. Based on this definition, we argue that there is already at least one system that is more intelligent than at least one of the conscious systems, yet that more intelligent system is not conscious. Therefore, there is not enough evidence to sustainably support the leap from intelligence to consciousness.

The impact of artificial intelligence on managerial attention allocation for discontinuous change: a conceptual framework

AbstractThe early detection of discontinuous change plays a decisive role in the performance and competitiveness of organizations. Whether and how top managers perceive change is mostly determined by their attention. However, research has shown that many top managers are often unable to allocate their attention properly and may therefore react with inertia or not at all to relevant changes. This raises the question of how managerial attention can be influenced to make top managers more receptive to discontinuous changes. A promising approach to this problem may come from the growing research field on the impact of AI on strategic decision-making. In this paper, I provide a conceptual framework of how the use of AI might help top managers better focus their attention on discontinuous change. Based on a systematic literature review and an attentional model, I highlight factors that influence top managers' attention allocation and likely enhance or inhibit it through the use of AI. This allows me to derive propositions for the application of AI in discontinuous change detection that can serve as a starting point for future empirical research. My paper contributes to broadening the research field of AI in the area of managerial attention.

Artificial Intelligence and Metaverse

The word meta is derived from latin and translates to mean “beyond.” Thus, the concept of the Metaverse aims to travel beyond mundanity, to go beyond the boundaries of the present and to impossibly reach beyond the constraints of reality itself. The fundamental concept of the Metaverse was developed and proposed by Neil Stephenson in the year 1992 and is as such a very new possibility. The computational resources for the metaverse now exist in the form of the internet, the cloud and AI. This paper aims to discuss the potential contained in the metaverse and more specifically for AI in the metaverse. As society progresses, so does its Artificial Intelligence and this can have far-reaching impact on the metaverse as a whole. The paper also discusses new applications for the metaverse as well in the field of AI and the world at large.

From Clustering to Cluster Explanations via Neural Networks.

A recent trend in machine learning has been to enrich learned models with the ability to explain their own predictions. The emerging field of explainable AI (XAI) has so far mainly focused on supervised learning, in particular, deep neural network classifiers. In many practical problems, however, the label information is not given and the goal is instead to discover the underlying structure of the data, for example, its clusters. While powerful methods exist for extracting the cluster structure in data, they typically do not answer the question why a certain data point has been assigned to a given cluster. We propose a new framework that can, for the first time, explain cluster assignments in terms of input features in an efficient and reliable manner. It is based on the novel insight that clustering models can be rewritten as neural networks-or "neuralized." Cluster predictions of the obtained networks can then be quickly and accurately attributed to the input features. Several showcases demonstrate the ability of our method to assess the quality of learned clusters and to extract novel insights from the analyzed data and representations.

Description of Parameter Variation Learning with Artificial Intelligence and GeoGebra in Students of a Differential Equations Course

In this era of social and technological change, the advent of AI has had great impact on the lives of many people and especially in the educational field, as mentioned in [1], wich prove that the perceptions that students and instructors have about artificial intelligence are very positive to implement them in learning. In this [2] assert that it is possible to have a collaborations between the fields of mathematics and artificial intelligence AI.

Mathematics & Artificial Intelligence: Intersections and Educational Implications

Educational jurisdictions worldwide are integrating AI education in their curricula, across grades K-12, and across subject areas, with a focus on AI applications, societal implications, and AI ethics. Jurisdictions also focusing on how AI works and how AI is developed are realizing that AI relies heavily on mathematical algorithms. The jurisdictions that are advancing K-12 AI mathematics curricula to prepare students to understand and apply the mathematics concepts used by AI systems are focused on grades 11-12 courses. This paper investigates how AI mathematics curricula may be designed for younger grades. First, we take a close look at the nature of a neural network and identify the mathematics typically used. Second, we review K-12 AI curricula in Canada and internationally and note that they lack a focus on AI mathematics. Third, we offer examples of how we may engage students across grades with mathematics used in the neural networks. Last, we look at future directions of AI mathematics education and research. Neural networks are not the only approach to AI, and there is more to AI than neural networks. However, neural networks have led to impressive progress in the field of AI, such as the development of large language models like ChatGPT. For our paper, focusing on neural networks gives us a sufficient starting point for addressing the questions we raise. This paper contributes to conversations about the intersection of AI education and mathematics education, and the development and research of AI mathematics curricula and teaching and learning resources across K-12.

Exploring Variational Auto-encoder Architectures, Configurations, and Datasets for Generative Music Explainable AI

Generative AI models for music and the arts in general are increasingly complex and hard to understand. The field of explainable AI (XAI) seeks to make complex and opaque AI models such as neural networks more understandable to people. One approach to making generative AI models more understandable is to impose a small number of semantically meaningful attributes on generative AI models. This paper contributes a systematic examination of the impact that different combinations of variational auto-encoder models (measureVAE and adversarialVAE), configurations of latent space in the AI model (from 4 to 256 latent dimensions), and training datasets (Irish folk, Turkish folk, classical, and pop) have on music generation performance when 2 or 4 meaningful musical attributes are imposed on the generative model. To date, there have been no systematic comparisons of such models at this level of combinatorial detail. Our findings show that measureVAE has better reconstruction performance than adversarialVAE which has better musical attribute independence. Results demonstrate that measureVAE was able to generate music across music genres with interpretable musical dimensions of control, and performs best with low complexity music such as pop and rock. We recommend that a 32 or 64 latent dimensional space is optimal for 4 regularised dimensions when using measureVAE to generate music across genres. Our results are the first detailed comparisons of configurations of state-of-the-art generative AI models for music and can be used to help select and configure AI models, musical features, and datasets for more understandable generation of music.

Incorporating Machine Learning into Sociological Model-Building

Quantitative sociologists frequently use simple linear functional forms to estimate associations among variables. However, there is little guidance on whether such simple functional forms correctly reflect the underlying data-generating process. Incorrect model specification can lead to misspecification bias, and a lack of scrutiny of functional forms fosters interference of researcher degrees of freedom in sociological work. In this article, I propose a framework that uses flexible machine learning (ML) methods to provide an indication of the fit potential in a dataset containing the exact same covariates as a researcher’s hypothesized model. When this ML-based fit potential strongly outperforms the researcher’s self-hypothesized functional form, it implies a lack of complexity in the latter. Advances in the field of explainable AI, like the increasingly popular Shapley values, can be used to generate understanding into the ML model such that the researcher’s original functional form can be improved accordingly. The proposed framework aims to use ML beyond solely predictive questions, helping sociologists exploit the potential of ML to identify intricate patterns in data to specify better-fitting, interpretable models. I illustrate the proposed framework using a simulation and real-world examples.

Making Sense of Machine Learning: A Review of Interpretation Techniques and Their Applications

Transparency in AI models is essential for promoting human–AI collaboration and ensuring regulatory compliance. However, interpreting these models is a complex process influenced by various methods and datasets. This study presents a comprehensive overview of foundational interpretation techniques, meticulously referencing the original authors and emphasizing their pivotal contributions. Recognizing the seminal work of these pioneers is imperative for contextualizing the evolutionary trajectory of interpretation in the field of AI. Furthermore, this research offers a retrospective analysis of interpretation techniques, critically evaluating their inherent strengths and limitations. We categorize these techniques into model-based, representation-based, post hoc, and hybrid methods, delving into their diverse applications. Furthermore, we analyze publication trends over time to see how the adoption of advanced computational methods within various categories of interpretation techniques has shaped the development of AI interpretability over time. This analysis highlights a notable preference shift towards data-driven approaches in the field. Moreover, we consider crucial factors such as the suitability of these techniques for generating local or global insights and their compatibility with different data types, including images, text, and tabular data. This structured categorization serves as a guide for practitioners navigating the landscape of interpretation techniques in AI. In summary, this review not only synthesizes various interpretation techniques but also acknowledges the contributions of their original authors. By emphasizing the origins of these techniques, we aim to enhance AI model explainability and underscore the importance of recognizing biases, uncertainties, and limitations inherent in the methods and datasets. This approach promotes the ethical and practical use of interpretation insights, empowering AI practitioners, researchers, and professionals to make informed decisions when selecting techniques for responsible AI implementation in real-world scenarios.

The Ethical Dilemma with Open AI ChatGPT: Is it Right or Wrong to prohibit it?

Digitalisation and innovation in learning and research are rapidly becoming crucial drivers of society's sustainable and progressive growth. AI's technological advancements and landscapes have significant strengths, and their diversity and quality have grown in recent years. This has facilitated the impressive development of AI apps and software, such as ChatGPT, which has become popular around the world. ChatGPT is an OpenAI access to users in education to generate essays, song lyrics and stories. It is an AI language model that can understand and generate human-like responses to text inputs, making it a valuable tool for various economic and cultural applications. This study examines the ethical dilemma of banning ChatGPT. Using a range of argumentative examples, I address the concept of moral obligations to OpenAI access but also its limitations. Some possible ethical issues that may arise in the use of AI-powered chatbots include concerns about data privacy, algorithmic bias, and the potential for chatbots to replace human interaction and support. Can OpenAI's cutting-edge technology and tools truly help corporate operations and institutions, and improve decision-making? Can it also give students and researchers significant resources to help them develop their knowledge, critical thinking skills and understanding in a variety of fields? Allowing ChatGPT to operate freely could lead to unintended consequences, but it could also promote innovation in the field of AI. Ultimately, finding a balance between regulation and innovation is key to maximising the benefits of ChatGPT while minimising its potential harms. AI software has the potential to degrade and debase our ethics, which are fundamentally different from our critical thinking. Chomsky, Roberts & Watumull concern is that AI software lacks the ability to understand and apply ethical principles in the same way that humans do, which could lead to unintended consequences and ethical dilemmas. Keywords: Critical Thinking; Ethical dilemma; Right or Wrong; ChatGPT

The Knowledge Structure and Development Trend in Artificial Intelligence Based on Latent Feature Topic Model

Currently, with the rapid development of science and technology, the field of artificial intelligence presents characteristics such as a wide crossover of disciplines and fast update, and the field of artificial intelligence has become a new focus of international competition. As an interdisciplinary field, the field of artificial intelligence has rich knowledge and strategic management significance. This article conducts an in-depth study on the knowledge structure and evolution trends in the field of AI, and the main work is as follows. First, a new potential feature topic model New-LDA is proposed for the study of topic recognition, which enhances the feature learning ability of the traditional LDA model, and makes up for the deficiency of the traditional LDA model in the ability of recognizing topics in complex environments. Second, the knowledge structure in the field of AI is analyzed from two aspects: topic recognition and coword analysis. The time series model is introduced to establish the topic evolution network, and the high-frequency words in three periods are compared and analyzed to find the evolution regular of knowledge structure in the AI domain. Finally, taking the cross-discipline of AI as an example, the thematic evolution of the field and its cross-discipline is analyzed to determine the future development direction and evolutionary trend of the field of AI.

Artificial Intelligence- Oncology and Central Nervous System Tumour Detection

In recent times, in this world of science and technology and recent advancements like machine learning and artificial intelligence, clinicians and medical science are supported with better treatment assistance, increased efficiency and improved methodology in the detection of even the rarest tumour in the human body. In the field of oncology, the help of AI has proved promising results. Diagnosing by imaging and detecting gliomas, its grading can be done easily and accurately. This article focuses on recent advances and technologies in the field of AI and CNS Brain tumour detection. Rare and difficult tumours hard to detect and identify can now be seen and classified with the help of these newer technologies. Pre Intra and post-operative strategies can be planned accurately and most precisely with the help of AI. It is a vast concept that helps enhance various human cognitive abilities in wide ranges.Deep Learning, one of the types of ML, has proved effective in automating many time-consuming steps, including lesion detection and segmentation. AI has several features such as detection and classification, tumour molecular properties, cancer-linked genetics, discoveries of various drugs, prediction of treatment, its outcomes as well as survival, and continued trends in personalized medicine in CNS tumours such as GBM with poor prognosis. Artificial Intelligence is a vast concept that helps enhance various human cognitive abilities in wide ranges. This review focuses on the recent advances in AI and its use in oncology, specifically in CNS, detection, and assessment planning of the underlying cause. Promises and challenges of the same are discussed below.

Challenges and Limitations of Artificial Intelligence Implementation in Modern Power Grid

Ongoing global decarbonization and energy transition have led to significant changes in the traditional power grid, resulting in new challenges for grid operators. These challenges, which are exacerbated by the increased penetration of distributed energy resources (DER), include power system stability, protection and control system malfunction, and supply and demand balance. To better understand the implications of high DER penetration, we proposed a classification of modern grids based on the DER integration level and identified the specific challenges faced by each category. We defined eight key challenges posed by DER on the grid and examined recent academic research on the application of AI techniques to modern power systems. Despite the potential benefits of AI in terms of power system stability and protection, there are currently no commercial applications of this research. Our study analyzed the limitations and barriers to practical AI implementation in power system protection and stability, including the use of synthetic data, insufficient real measurement data, protection selectivity, and the black-box nature of the models. We emphasize the importance of further research and development in the fields of Explainability AI and physics-informed ML to overcome these challenges and enable the successful implementation of AI in modern power grid stability and protection.

SkinLiTE: Lightweight Supervised Contrastive Learning Model for Enhanced Skin Lesion Detection and Disease Typification in Dermoscopic Images.

This study introduces SkinLiTE, a lightweight supervised contrastive learning model tailored to enhance the detection and typification of skin lesions in dermoscopic images. The core of SkinLiTE lies in its unique integration of supervised and contrastive learning approaches, which leverages labeled data to learn generalizable representations. This approach is particularly adept at handling the challenge of complexities and imbalances inherent in skin lesion datasets. The methodology encompasses a two-phase learning process. In the first phase, SkinLiTE utilizes an encoder network and a projection head to transform and project dermoscopic images into a feature space where contrastive loss is applied, focusing on minimizing intra-class variations while maximizing inter-class differences. The second phase freezes the encoder's weights, leveraging the learned representations for classification through a series of dense and dropout layers. The model was evaluated using three datasets from Skin Cancer ISIC 2019-2020, covering a wide range of skin conditions. SkinLiTE demonstrated superior performance across various metrics, including accuracy, AUC, and F1 scores, particularly when compared with traditional supervised learning models. Notably, SkinLiTE achieved an accuracy of 0.9087 using AugMix augmentation for binary classification of skin lesions. It also showed comparable results with the state-of-the-art approaches of ISIC challenge without relying on external data, underscoring its efficacy and efficiency. The results highlight the potential of SkinLiTE as a significant step forward in the field of dermatological AI, offering a robust, efficient, and accurate tool for skin lesion detection and classification. Its lightweight architecture and ability to handle imbalanced datasets make it particularly suited for integration into Internet of Medical Things environments, paving the way for enhanced remote patient monitoring and diagnostic capabilities. This research contributes to the evolving landscape of AI in healthcare, demonstrating the impact of innovative learning methodologies in medical image analysis.