Development Of Artificial Intelligence Research Articles

Bias in medical AI: Implications for clinical decision-making.

Biases in medical artificial intelligence (AI) arise and compound throughout the AI lifecycle. These biases can have significant clinical consequences, especially in applications that involve clinical decision-making. Left unaddressed, biased medical AI can lead to substandard clinical decisions and the perpetuation and exacerbation of longstanding healthcare disparities. We discuss potential biases that can arise at different stages in the AI development pipeline and how they can affect AI algorithms and clinical decision-making. Bias can occur in data features and labels, model development and evaluation, deployment, and publication. Insufficient sample sizes for certain patient groups can result in suboptimal performance, algorithm underestimation, and clinically unmeaningful predictions. Missing patient findings can also produce biased model behavior, including capturable but nonrandomly missing data, such as diagnosis codes, and data that is not usually or not easily captured, such as social determinants of health. Expertly annotated labels used to train supervised learning models may reflect implicit cognitive biases or substandard care practices. Overreliance on performance metrics during model development may obscure bias and diminish a model's clinical utility. When applied to data outside the training cohort, model performance can deteriorate from previous validation and can do so differentially across subgroups. How end users interact with deployed solutions can introduce bias. Finally, where models are developed and published, and by whom, impacts the trajectories and priorities of future medical AI development. Solutions to mitigate bias must be implemented with care, which include the collection of large and diverse data sets, statistical debiasing methods, thorough model evaluation, emphasis on model interpretability, and standardized bias reporting and transparency requirements. Prior to real-world implementation in clinical settings, rigorous validation through clinical trials is critical to demonstrate unbiased application. Addressing biases across model development stages is crucial for ensuring all patients benefit equitably from the future of medical AI.

Examining teachers’ behavioural intention of using generative artificial intelligence tools for teaching and learning based on the extended technology acceptance model

The rapid development of generative artificial intelligence (GenAI) tools has given rise to a growing discussion of the potential challenges and benefits that the use of these technologies may present in the field of education. This study examines the acceptance of the use of GenAI tools for teaching and learning among primary and secondary school teachers in Hong Kong. It uses an extension of the technology acceptance model (TAM) with a modified framework that incorporates two key factors: self-efficacy and subjective norm. Data were collected from a sample of 367 primary and secondary school teachers in Hong Kong using questionnaires containing items for six constructs: self-efficacy, perceived usefulness, perceived ease of use, attitude towards using, subjective norm, and behavioural intention. The results show that fostering teachers’ self-efficacy, perceived usefulness, and attitude is essential for successfully increasing their behavioural intention to use GenAI tools. Subjective norm was also found to influence teachers’ behavioural intention. To enhance teachers’ effective use of GenAI for teaching, teacher development programmes should focus on equipping teachers with comprehensive conceptual knowledge and skills and an understanding of the application of these tools to teaching and learning. Policy support to create a conducive environment for the use of GenAI in teaching and learning would also be beneficial. The study has theoretical implications in its extension of the TAM model as well as implications for enhancing teachers’ AI literacy and developing pedagogies for the meaningful use of GenAI tools for teaching and learning in K–12 settings.

Enhancing high-resolution reconstruction of flow fields using physics-informed diffusion model with probability flow sampling

The application of artificial intelligence (AI) technology in fluid dynamics is becoming increasingly prevalent, particularly in accelerating the solution of partial differential equations and predicting complex flow fields. Researchers have extensively explored deep learning algorithms for flow field super-resolution reconstruction. However, purely data-driven deep learning models in this domain face numerous challenges. These include susceptibility to variations in data distribution during model training and a lack of physical and mathematical interpretability in the predictions. These issues significantly impact the effectiveness of the models in practical applications, especially when input data exhibit irregular distributions and noise. In recent years, the rapid development of generative artificial intelligence and physics-informed deep learning algorithms has created significant opportunities for complex physical simulations. This paper proposes a novel approach that combines diffusion models with physical constraint information. By integrating physical equation constraints into the training process of diffusion models, this method achieves high-fidelity flow field reconstruction from low-resolution inputs. Thus, it not only leverages the advantages of diffusion models but also enhances the interpretability of the models. Experimental results demonstrate that, compared to traditional methods, our approach excels in generating high-resolution flow fields with enhanced detail and physical consistency. This advancement provides new insights into developing more accurate and generalized flow field reconstruction models.

Determinants of AI performance and their effect on national GDP: an empirical study

Abstract. Artificial intelligence is one of the most widely discussed topics in recent years, remarkably following the pandemic. During the pandemic, numerous industries and organizations have focused on the development of artificial intelligence due to the shift to remote learning and working. This shift also caused expanded investments and research on AI driven by the growing demand of for AI-powered tools across various sectors such as education, healthcare, and business. This paper analyses the determinants of AI performance and examines the relationship between a countrys AI performance and its economic growth, measured by Gross Domestic Product (GDP). This paper employs Pearson Pairwise Correlation Multiple Linear Regression, and Simple Linear Regression models with a sample of 62 different countries to uncover the most influential factors impacting a country's AI performance and its subsequent influence on GDP. During the process of constructing the multiple linear regression model, this paper also evaluates the collinearity diagnostics such as VIF and tolerance level to prevent multicollinearity. This study finds that a countrys innovation skill, featuring its research and development capabilities, is the primary driver influencing its success in AI development. Subsequently, this paper concludes that advanced AI performance can significantly contribute to a countrys economic growth, further emphasizing the importance of progressive AI development.

Artificial Intelligence: Intensifying or Mitigating Unemployment?

The rapid development of Artificial Intelligence (AI) is simultaneously fostering a proliferation of novel job opportunities while rendering some traditional roles obsolete and specific skills outdated. Previous research has failed to consider the short-, medium-, and long-term variations in AI’s impact on unemployment, which may lead to an incomplete understanding of the AI-employment relationship. This paper examines daily data from January 4, 2013, to August 12, 2024, utilising advanced wavelet-based Quantile on Quantile Regression (QQR) methodology to assess AI’s impact on the Unemployment Index (UI) across quantiles and time scales, with a sample size of 2820 drawn from a larger dataset totalling 4241 observations. The conclusions reveal that AI generally positively impacts UI in the short term, especially with AI at 0.6-0.7 quantiles, as automation replaces workers faster than new job roles emerge and skills transform. However, in the medium term, positive and negative effects balance as new jobs and skills emerge through continuous industrial restructuring. In the long run, AI predominantly mitigates UI by further enhancing economic development, fostering skill upgrading, and facilitating market adjustments, but this result does not hold during AI at 0.7 quantiles and UI at the highest quantiles, such as Coronavirus Disease 2019 (COVID-19). Under new technological revolution and industrial transformation, we formulate China-specific suggestions to avert potential AI-induced unemployment crisis from short-term, medium-term, long-term, and sector-specific perspectives.

Making pathologists ready for the new AI era: changes in required competencies.

In recent years, there has been an increasing interest in developing and using artificial intelligence (AI) models in pathology. While pathologists generally have a positive attitude towards AI, they report a lack of knowledge and skills regarding how to use it in practice. Furthermore, it remains unclear what skills pathologists would require to use AI adequately and responsibly. Yet adequate training of (future) pathologists is essential for successful AI use in pathology. In this paper, we assess which entrustable professional activities (EPAs) and associated competencies pathologists should acquire in order to use AI in their daily practice. We make use of available academic literature, including literature in radiology, another image-based discipline, which is currently more advanced in terms of AI development and implementation. Although microscopy evaluation and reporting could be transferrable to AI in the future, most of the current pathologist EPAs and competencies will likely remain relevant when using AI techniques and interpreting and communicating results for individual patient cases. In addition, new competencies related to technology evaluation and implementation will likely be necessary, and knowing one's own strengths and limitations in human-AI interaction. Because current EPAs do not sufficiently address the need to train pathologists in developing expertise related to technology evaluation and implementation, we propose a new EPA to enable pathology training programs to make pathologists fit for the new AI era "using AI in diagnostic pathology practice" and outline its associated competencies.

Adaption of Artificial Intelligence (AI) to enhance business and collaboration between countries, focusing on Saudi Arabia

Artificial Intelligence (AI) is one of the most ingenious inventions drastically transforming the business world. The question is whether the current growth of AI is all hype or if it has the potential to revolutionize the world is raised by the abundance of intelligent goods and services that have emerged in recent times. Technological developments in AI pave the way for creating technologies that resemble humans and can operate independently and imitate our intellectual processes. Numerous industries have shown considerable interest and enthusiasm in adopting AI due to technological advancements. As a result, numerous businesses are making significant investments to leverage this opportunity through innovative business models based on AI. The study aims to establish a model based on an evidence-synthesis approach that demonstrates how the business sector in Saudi Arabia can incorporate AI to achieve better outcomes and strengthen its relationships with other countries. However, the study also highlights some advantages and disadvantages of adopting AI models in the business industry so a balanced approach can be used to incorporate AI in different aspects of the business. The study recommends incorporating AI in different health, education, and business sectors in Saudi Arabia to foster international relationships in Saudi Arabia.

AI in Soil Health Monitoring: A Data-Driven Approach

Abstract: The sustainability of ecosystems and agricultural productivity depend on healthy soil. However, conventional techniques for evaluating soil health are sometimes time-consuming, labor-intensive, and restricted in their geographic reach. New opportunities for scalable, real-time soil health monitoring have been made possible by recent developments in artificial intelligence (AI) and data-driven methodologies. This study examines a thorough AI-powered system for tracking soil health, emphasizing methods for data collection, processing, and predictive modeling. AI models can provide precise forecasts of soil characteristics, health indicators, and possible crop yields by combining data from multiple sources, such as remote sensing, soil sensors, and historical data. This study offers a comprehensive analysis of recent AI applications in soil health monitoring and suggests a reliable, scalable approach intended to incorporate diverse data sources, guaranteeing precise and effective soil health assessment.

A multimodal educational robots driven via dynamic attention.

With the development of artificial intelligence and robotics technology, the application of educational robots in teaching is becoming increasingly popular. However, effectively evaluating and optimizing multimodal educational robots remains a challenge. This study introduces Res-ALBEF, a multimodal educational robot framework driven by dynamic attention. Res-ALBEF enhances the ALBEF (Align Before Fuse) method by incorporating residual connections to align visual and textual data more effectively before fusion. In addition, the model integrates a VGG19-based convolutional network for image feature extraction and utilizes a dynamic attention mechanism to dynamically focus on relevant parts of multimodal inputs. Our model was trained using a diverse dataset consisting of 50,000 multimodal educational instances, covering a variety of subjects and instructional content. The evaluation on an independent validation set of 10,000 samples demonstrated significant performance improvements: the model achieved an overall accuracy of 97.38% in educational content recognition. These results highlight the model's ability to improve alignment and fusion of multimodal information, making it a robust solution for multimodal educational robots.

Artificial Intelligence: Communication, Technology, and Society (a Systematic Literature Review)

The development of artificial intelligence (AI) technology has transformed the way various social sectors operate, including healthcare, communications, and decision-making. The integration of AI into these domains not only opens up new opportunities for efficiency and innovation, but also presents significant challenges. Among these challenges, the emergence of algorithmic bias, inadequate data protection, technical limitations, and complex ethical dilemmas are of major concern. These issues affect the reliability and fairness of the use of AI technologies, potentially harming individuals and society as a whole. This study aims to investigate the barriers associated with the implementation of AI in society and provide theoretical and methodological solutions to address these challenges. This study uses a systematic literature review methodology to review 40 journals, with the aim of identifying key issues and suggesting solutions for the ethical and successful integration of AI. The findings show that the potential of AI is often hampered by biased algorithms, inadequate privacy laws, and a lack of transparency in decision-making processes. The study recommends the use of qualitative and quantitative methodologies, including experimental research and case studies, to investigate practical applications of AI. In addition, a multidisciplinary approach is suggested to address the technical and social dimensions of AI applications.

Unveiling the Dynamic Mechanisms of Generative AI in English Language Learning: A Hybrid Study Based on fsQCA and System Dynamics

The burgeoning development of generative artificial intelligence (GenAI) has unleashed transformative potential in reshaping English language education. However, the complex interplay of learner, technology, pedagogy, and contextual factors that shape the effectiveness of GenAI-assisted language learning remains underexplored. This study employed a novel mixed-methods approach, integrating qualitative comparative analysis (QCA) and system dynamics (SD) modeling, to unravel the multi-dimensional, dynamic mechanisms underlying the impact of GenAI on English learning outcomes in higher education. Leveraging a sample of 33 English classes at the Harbin Institute of Technology, the QCA results revealed four distinct configurational paths to high and low learning effectiveness, highlighting the necessary and sufficient conditions for optimal GenAI integration. The SD simulation further captured the emergent, nonlinear feedback processes among learner attributes, human–computer interaction, pedagogical practices, and ethical considerations, shedding light on the temporal evolution of the GenAI-empowered language-learning ecosystem. The findings contribute to the theoretical advancement of intelligent language education by constructing an integrative framework encompassing learner, technology, pedagogy, and context dimensions. Practical implications are generated to guide the responsible design, implementation, and optimization of GenAI in English language education, paving the way for learner-centric, adaptive learning experiences in the intelligence era.

Trends in the industry of clinical trials

Background. The clinical trial (CT) industry is subject to changes, some of which are rapidly developing and their directions can be predicted, while others develop gradually, forming stable tendencies, which requires analysis to confirm and predict them. According to the generally accepted time gradation, there are an operational (up to 1 month), short-term (up to 1 year), medium-term (up to 5 years), long-term (up to 20 years), and long-term (over 20 years) prognoses. A short-term forecast is common in CT industry.Objective: to identify trends in the development of CT industry until the end of 2024.Material and methods. We searched publications in Russian and English segments of the Internet in open access sources in ClinicalTrials.gov, PubMed/MEDLINE, Google Scholar, Academia, ResearchGate, CyberLeninka, eLibrary databases by key query “trends in clinical trials” for the period from January to March 2023. The query “history of clinical trials” was also used in Russian part of the Internet through Google search. The analysis included publications on the history of CT development over 1936–2023. The total number of sources analyzed was 59 (9 Russian and 50 English).Results. The tendencies in CT development can be divided into two groups. The first one is related to CT organization and conducting, while another is associated with development of innovative drugs. In the first group, the trends have persisted since 2022 and are expressed in the ongoing digitalization of operational activities, a shift from centralized research to decentralization, while the protocol design has changed towards patient-centricity. In the second group, the number of expected drugs has decreased and a shift towards biological drugs, gene and cell therapy has become more pronounced.Conclusion. Trends in CI are characterized by a number of innovations, primarily related to digitalization, the development of telemedicine technologies, mathematical modeling, artificial intelligence, virtual CI, decentralization and patient-centricity.

Exploring the Landscape of Explainable Artificial Intelligence (XAI): A Systematic Review of Techniques and Applications

Artificial intelligence (AI) encompasses the development of systems that perform tasks typically requiring human intelligence, such as reasoning and learning. Despite its widespread use, AI often raises trust issues due to the opacity of its decision-making processes. This challenge has led to the development of explainable artificial intelligence (XAI), which aims to enhance user understanding and trust by providing clear explanations of AI decisions and processes. This paper reviews existing XAI research, focusing on its application in the healthcare sector, particularly in medical and medicinal contexts. Our analysis is organized around key properties of XAI—understandability, comprehensibility, transparency, interpretability, and explainability—providing a comprehensive overview of XAI techniques and their practical implications.

Mastering Data Pipelines for Al: A Beginner's Guide to Building Efficient Workflows

This comprehensive article explores the critical role of data pipelines in Artificial Intelligence (AI) development, emphasizing their importance in ensuring high-quality, relevant data for machine learning models. It covers key stages of the data pipeline process, including data ingestion, cleaning and preprocessing, transformation and feature engineering, and storage and loading. The article discusses various tools, techniques, and best practices for each stage, addressing challenges in handling diverse data sources, scalability issues, and the need for efficient data management. It highlights the significance of robust data pipelines in enhancing AI model performance and reliability, while also considering the dynamic nature of the AI field and the necessity for continuous learning and adaptation.

FORMATION OF STRATEGIC DIRECTIONS FOR THE USE OF ARTIFICIAL INTELLIGENCE IN THE ENTERPRISE TO ACHIEVE THE GOALS OF SUSTAINABLE DEVELOPMENT

Rapid artificial intelligence (AI) development opens new opportunities for enterprises to achieve sustainable development goals. This article summarizes the research results of applying AI to optimize production processes, reduce environmental impact, increase productivity, and strengthen social responsibility. The article aims to highlight the potential of AI for achieving enterprises' sustainable development goals and identify effective strategies for its application to increase economic efficiency, social responsibility, and environmental sustainability. A qualitative analysis of statistical data, official regulatory documents, books, and articles from accredited journals was used to achieve the goal. The obtained results showed that AI could be used to analyze production processes and forecast the demand for raw materials, optimize supply chains and reduce energy consumption, identify potential sustainability risks and develop strategies to prevent them, identify defects in production and control product quality in real-time, create personalized services and products for customers. The study highlights the importance of AI's ethical and safe use to ensure its positive impact on society and the environment. The practical value of the article lies in the systematic review of studies on the effects of AI on the sustainable development of enterprises. The obtained results can be used to improve production processes, increase enterprises' competitiveness, and promote sustainable development.

THE INNOVATIVE IMPACT OF GENERATIVE ARTIFICIAL INTELLIGENCE ON DIGITAL BUSINESS TRANSFORMATION

In the context of intensifying competition and evolving market dynamics, the deployment of cutting-edge technologies has become not merely a discretionary choice, but an indispensable imperative for any enterprise aspiring to achieve successful growth. Generative artificial intelligence, with its substantial potential for automation, personalisation and optimisation of business processes, is emerging as a highly promising avenue of digital transformation. This study is dedicated to investigating approaches and delineating strategies for aligning generative artificial intelligence with the requirements of digital business transformation. The research examines the development of artificial intelligence, with a focus on symbolic artificial intelligence, machine learning, deep learning and generative artificial intelligence. In addition, it considers the impact of these developments on business processes. The article identifies the potential benefits and challenges associated with the adaptation of generative artificial intelligence to the needs of modern business, in the areas of marketing, sales and data analysis. The utilisation of diverse methodologies and techniques, including prompts, fine-tuning, and the incorporation of interactive guidance systems, can enhance the efficacy and precision of generative AI in a business setting, thereby facilitating optimal outcomes in a multitude of tasks. The authors put forth the proposition of employing generative artificial intelligence technology in conjunction with Retrieval-Augmented Generation, with the objective of enhancing the quality and relevance of responses to user queries. Additionally, they advocate for the utilisation of agents or orchestration tools to provide guidance to models. The successful implementation of generative artificial intelligence hinges on three key factors: the clear definition of objectives, the selection of suitable tools and technologies, and the assurance of managerial and staff support. The implementation of generative artificial intelligence will contribute to increased efficiency through the automation of routine tasks, enhanced competitiveness through personalisation and innovation, optimised cost structures that increase profitability, and expanded opportunities for research and development.

Ambiguous AI: Lossy Compression, Model Collapse, and Intercultural Sensitivity

This paper examines the intersections of artificial intelligence (AI), cultural representation, and intercultural sensitivity, critiquing the impact of lossy compression and model collapse on cultural expression in AI systems. The study introduces “intentional patiency” as an alternative framework, conceptualizing AI not merely as an active agent but as a cultural receptor, and in turn, producer. AI systems often rely on compression algorithms that strip away cultural complexities, rendering nuanced traditions as simplified, stereotypical artifacts. Drawing on frameworks from Ted Chiang’s metaphor of lossy compression to Wendy Chun’s critique of noise, the essay explores how AI’s data optimization selectively filters cultural diversity, perpetuating dominant narratives while silencing marginalized voices. Model collapse in generative systems demonstrates while also exacertabting the pitfalls of homogenous data to reduce cultural representations to monolithic “single stories,” echoing concerns raised by Chimamanda Ngozi Adichie. To address this issue, this essay advocates for an interculturally sensitive AI development paradigm, incorporating cultural expansion and ethical safeguards to mitigate cultural erasure. Through the concept of “ambiguous AI,” a paradigm rooted in probabilistic engagement and intentional patiency, the essay seeks to reframe AI’s role as a nuanced participant in cultural discourse, promoting culturally adaptive architectures.

The suitability of skilled talent training in higher vocational colleges based on AI and 5G network connection in the context of educational transformation and upgrading

The development of artificial intelligence (AI) and 5G network technology has changed the production and lifestyle of people. AI also has promoted the transformation of talent training mode under the integration of college industry and education. In the context of the current transformation of education, AI and 5G networks are increasingly used in the education industry. This paper optimizes and upgrades the training mode of skilled talents in higher vocational colleges by using its advanced methods and technologies of information display. This means is helpful to analyze and solve a series of objective problems such as the single training form of the current talent training mode. This paper utilizes the principles and laws of industry university research (IUR) collaboration for reference to construct and optimize the talent training mode based on the analysis of the requirements of talent training and the role of each subject in talent training. Then, the ecological talent training environment can be realized. In the analysis of talent training mode under the cooperation of production and education, the correlation coefficients of network construction, environment construction, scientific research funds, scientific research level, and policy support were 0.618, 0.576, 0.493, 0.785, and 0.451, respectively. This showed that the scientific research level had the greatest impact on talent training in the talent training mode of IUR collaboration, while policy support had less impact on talent training compared with other factors. The combination of AI and 5G network technology with the talent training mode of IUR cooperation can effectively analyze the influencing factors and problems of the talent training mode. The hybrid method is of great significance to the talent training strategy and fitting degree.

A Knowledge Base Driven Task-Oriented Image Semantic Communication Scheme

With the development of artificial intelligence and computer hardware, semantic communication has been attracting great interest. As an emerging communication paradigm, semantic communication can reduce the requirement for channel bandwidth by extracting semantic information. This is an effective method that can be applied to image acquisition of unmanned aerial vehicles, which can transmit high-data-volume images within the constraints of limited available bandwidth. However, the existing semantic communication schemes fail to adequately incorporate the guidance of task requirements into the semantic communication process and are difficult to adapt to the dynamic changes of tasks. A task-oriented image semantic communication scheme driven by knowledge base is proposed, aiming at achieving high compression ratio and high quality image reconstruction, and effectively solving the bandwidth limitation. This scheme segments the input image into several semantic information unit under the guidance of task requirements by Yolo-World and Segment Anything Model. The assigned bandwidth for each unit is according to the task relevance scores, which enables high-quality transmission of task-related information with lower communication overheads. An improved metric weighted learned perceptual image patch similarity (LPIPS) is proposed to evaluate the transmission accuracy of the novel scheme. Experimental results show that our scheme achieves a notable performance improvement on weighted LPIPS while the same compression ratio compared with traditional image compression schemes. Our scheme has a higher target capture ratio than traditional image compression schemes under the task of target detection.

China’s AI Development Model in an Era of Technological Deglobalization

As strategic rivalry between Beijing and Washington centers on new technologies, a trend toward decoupling and deglobalization is challenging the global links upon which China’s artificial intelligence (AI) sector has relied for a long time. This means that Beijing’s AI development strategy must contend with an erosion of global interdependencies. This policy brief examines three key elements of China’s response: an infrastructure megaproject for computing power, a “whole-of-nation” approach to developing AI foundation models, and efforts to forge connections with foreign innovation systems beyond the United States. None of these come without challenges.