Human Reasoning Research Articles

Assessing the performance of ChatGPT in medical ethical decision-making: a comparative study with USMLE-based scenarios

IntroductionThe integration of artificial intelligence (AI) into healthcare introduces innovative possibilities but raises ethical, legal and professional concerns. Assessing the performance of AI in core components of the United States...

The inviolability and value of the human body and the corpse in the Islamic legal tradition

This article examines how the inviolability (hurma) of the human body is conceived of and negotiated within juristic discourses on the paradigmatic hypothetical case of the starving person from the fifth/eleventh through seventh/thirteenth centuries. The Islamic legal tradition documents lively juristic discussions and debates on hypothetical cases of necessity, many of which implicate foundational questions of value and the inviolability of the human body. At a basic level, necessity applies to a situation in which someone must ‘break’ the law or fail to fulfill a legal obligation, in order to escape the threat of serious harm to something of high value (usually, life, limb, or health). The jurists constructed multiple hypothetical cases of necessity that implicate the inviolability of the human body, including the cases of the starving person, excision of gangrenous tissue, and cesarian section, as well as numerous fact patterns of coercion. This article considers the challenging issues of inflicting serious bodily harm on others and on oneself in order to preserve life, including inflicting harm on the living and on the dead, as well as self-harm. It provides an account and analysis of both the substance and mechanics of fiqh argument, or how jurists reasoned through complex questions of value that also have substantial bases in conceptions of the human body, medicine, logic, and reasonableness through the language and method of the fiqh. This article argues that the evidence and methods of argumentation employed by the jurists reveal significant diversity and tensions in juristic method and understandings of value and subjectivity in the fiqh, as well as in the construction of the foundational concepts of inviolability (hurma) and (legal) protection ('isma). Furthermore, it shows that in the starving person paradigm, the jurists explored the boundary between (non-human) animality and humanity, and grappled with how to locate the (dead) human body within legal thought, and how to operationalize the higher degree of respect accorded to the human body than to non-human bodies. Finally, it examines the ways in which these discourses provide material for considering the functions of subjectivity, emotions, and moral aversion in the law.

LLM-Based Multi-Agent Systems for Software Engineering: Literature Review, Vision and the Road Ahead

Integrating Large Language Models (LLMs) into autonomous agents marks a significant shift in the research landscape by offering cognitive abilities that are competitive with human planning and reasoning. This paper explores the transformative potential of integrating Large Language Models into Multi-Agent (LMA) systems for addressing complex challenges in software engineering (SE). By leveraging the collaborative and specialized abilities of multiple agents, LMA systems enable autonomous problem-solving, improve robustness, and provide scalable solutions for managing the complexity of real-world software projects. In this paper, we conduct a systematic review of recent primary studies to map the current landscape of LMA applications across various stages of the software development lifecycle (SDLC). To illustrate current capabilities and limitations, we perform two case studies to demonstrate the effectiveness of state-of-the-art LMA frameworks. Additionally, we identify critical research gaps and propose a comprehensive research agenda focused on enhancing individual agent capabilities and optimizing agent synergy. Our work outlines a forward-looking vision for developing fully autonomous, scalable, and trustworthy LMA systems, laying the foundation for the evolution of Software Engineering 2.0.

Interpreting Metaphorical Language: A Challenge to Artificial Intelligence

In recent years, numerous studies have pointed to the ability of artificial intelligence (AI) to generate and analyze expressions of natural language. However, the question of whether AI is capable of actually interpreting human language, rather than imitating its understanding, remains open. Metaphors, being an integral part of human language, as both a common figure of speech and the predominant cognitive mechanism of human reasoning, pose a considerable challenge to AI systems. Based on an overview of the existing studies findings in computational linguistics and related fields, the paper identifies a number of problems associated with the interpretation of non-literal expressions of language by large language models (LLM). It reveals that there is still no clear understanding of the methods for training language models to automatically recognize and interpret metaphors that would bring it closer to the level of human “interpretive competencies”. The purpose of the study is to identify possible reasons that hinder the understanding of figurative language by artificial systems and to outline possible directions for solving this problem. The study suggests that the main barriers to AI’s human-like interpretation of figurative natural language are the absence of a physical body, the inability to reason by analogy and make inferences based on common sense, the latter being both the result and the cognitive process in extracting and processing information. The author concludes that further improvement of the AI systems creative skills should be at the top of the research agenda in the coming years.

Impaired renal transporter gene expression and uremic toxin excretion as aging hallmarks in cats with naturally occurring chronic kidney disease.

Aging leads to nephron senescence and chronic kidney disease (CKD). In cats, indoxyl sulfate (IxS) has been previously quantified and associated with CKD, and little is known about tubular transporters. Two cohorts of cats aged 6 to 21 years were enrolled. Cohort 1 included 41 colony cats with 28 control and 13 CKD cats. Cohort 2 had 30 privately-owned cats with 10 control and 20 CKD cats. In cohort 1, serum concentrations of IxS, trimethylamine N-oxide (TMAO), p-cresol sulfate (PCS), and phenyl sulfate were higher in CKD vs. control cats (all P<0.05). This observation was independently validated in cohort 2. Renal cortical and medullar tissues were collected from a third cohort of cats euthanized for humane reasons unrelated to the study. We provided the evidence that renal tubular transporter genes, OAT1, OAT4, OATP4C1, and ABCC2, but not OAT3, were expressed in the kidneys of cats, and their expressions were downregulated in CKD (all FDR<0.1). Cats and humans share 90.9%, 77.8%, and 82.5% identities in OAT1, OATP4C1, and ABCC2 proteins, respectively. In healthy cats, circulating TMAO and IxS are significantly correlated with age. Our study reveals impaired uremic toxin secretion and tubular transporter expression in cats with CKD.

Wacana Program Bansos untuk Korban Judi Online dalam Tinjauan Al-Maṣlaḥaḥ Al-Mursalah Perspektif Wahbah Az-Zuḥaili

The Minister of Human Development and Cultural Affairs, Muhadjir Effendy, opened the opportunity for victims of online gambling to be included in Social Welfare Integrated Data (DTKS) so that they could receive social assistance. The victims are family members or close relatives who have been financially harmed by online gambling perpetrators. This programme sparked support and opposition from various circles. The emergence of these differing views is influenced by varying perspectives on assessing the maṣlaḥaḥ (public interest) of the program. Indeed, in the context of Islamic law, maṣlaḥaḥ can be used as a basis for legitimizing something, a method known as al-maṣlaḥaḥ al-mursalah. However, according to Wahbah az-Zuḥaili in Uṣūl al-Fiqh al-Islāmi, the measure of maṣlaḥah is not human logic but the legitimacy of syariat regarding a particular form of maṣlaḥah. According to Wahbah az-Zuḥaili, human reasoning will always differ in assessing maṣlaḥah, as seen in the differing opinions of some scholars regarding the maṣlaḥah of social assistance programs. The aim of this study is to analyze the discourse surrounding the social assistance program for victims of online gambling from the perspective of al-maṣlaḥaḥ al-mursalah, according to Wahbah az-Zuḥaili. To achieve this goal, the research method used was qualitative, with a library research approach. The data collection technique involves reviewing various literature, including classical and contemporary scholars' books, scientific articles, and other relevant sources. The results of this study indicate that this program is in line with the standards of the concept of al-maṣlaḥaḥ al-mursalah, as outlined by Wahbah az-Zuḥaili.

AI in the Learning Environment: Examination of Pedagogical, Psychological, and Ethical Implications

This paper provides an overview of how AI tools and capabilities are being integrated in teaching and learning. It also addresses ethical aspects of integrating AI and other advanced technologies in these processes as well as the resulting implications. These issues include algorithmic fairness, privacy and surveillance, intellectual property rights, misinformation and violence, health concerns, and social implications. In addition, how access to such tools will impact the cognitive functions of their users, particularly learners, has become a relevant question. Much more information is now becoming easily accessible to a wider population, making it important to understand and mitigate AI ramifications to human learning, reasoning, and perception, especially for young learners whose brains are still in the early developmental stages. To that end, this paper also reviews the current literature regarding the impact of AI integration in education, focusing on the potential effects this can have on the cognitive abilities and functions of learners. Thus, with AI making its way into the educational realm, from pre-college to advanced studies levels, understanding the resulting pedagogical, psychological, and ethical ramifications can guide thoughtful development and appropriate integration of AI tools in support of the learning environment.

AI-assisted human clinical reasoning in the ICU: beyond "to err is human".

Diagnostic errors pose a significant public health challenge, affecting nearly 800,000 Americans annually, with even higher rates globally. In the ICU, these errors are particularly prevalent, leading to substantial morbidity and mortality. The clinical reasoning process aims to reduce diagnostic uncertainty and establish a plausible differential diagnosis but is often hindered by cognitive load, patient complexity, and clinician burnout. These factors contribute to cognitive biases that compromise diagnostic accuracy. Emerging technologies like large language models (LLMs) offer potential solutions to enhance clinical reasoning and improve diagnostic precision. In this perspective article, we explore the roles of LLMs, such as GPT-4, in addressing diagnostic challenges in critical care settings through a case study of a critically ill patient managed with LLM assistance.

Theory Is All You Need: AI, Human Cognition, and Causal Reasoning

Scholars argue that artificial intelligence (AI) can generate genuine novelty and new knowledge and, in turn, that AI and computational models of cognition will replace human decision making under uncertainty. We disagree. We argue that AI’s data-based prediction is different from human theory-based causal logic and reasoning. We highlight problems with the decades-old analogy between computers and minds as input–output devices, using large language models as an example. Human cognition is better conceptualized as a form of theory-based causal reasoning rather than AI’s emphasis on information processing and data-based prediction. AI uses a probability-based approach to knowledge and is largely backward looking and imitative, whereas human cognition is forward-looking and capable of generating genuine novelty. We introduce the idea of data–belief asymmetries to highlight the difference between AI and human cognition, using the example of heavier-than-air flight to illustrate our arguments. Theory-based causal reasoning provides a cognitive mechanism for humans to intervene in the world and to engage in directed experimentation to generate new data. Throughout the article, we discuss the implications of our argument for understanding the origins of novelty, new knowledge, and decision making under uncertainty.

Signs of consciousness in AI: Can GPT-3 tell how smart it really is?

The emergence of artificial intelligence (AI) is transforming how humans live and interact, raising both excitement and concerns—particularly about the potential for AI consciousness. For example, Google engineer Blake Lemoine suggested that the AI chatbot LaMDA might become sentient. At that time, GPT-3 was one of the most powerful publicly available language models, capable of simulating human reasoning to a certain extent. The notion of GPT-3 having some degree of consciousness could be linked to its ability to produce human-like responses, hinting at a basic level of understanding. To explore this further, we administered both objective and self-assessment tests of cognitive (CI) and emotional intelligence (EI) to GPT-3. Results showed that GPT-3 outperformed average humans on CI tests requiring the use and demonstration of acquired knowledge. However, its logical reasoning and EI capacities matched those of an average human. GPT-3’s self-assessments of CI and EI didn’t always align with its objective performance, with variations comparable to different human subsamples (e.g., high performers, males). A further discussion considered whether these results signal emerging subjectivity and self-awareness in AI. Future research should examine various language models to identify emergent properties of AI. The goal is not to discover machine consciousness itself, but to identify signs of its development, occurring independently of training and fine-tuning processes. If AI is to be further developed and widely deployed in human interactions, creating empathic AI that mimics human behavior is essential. The rapid advancement toward superintelligence requires continuous monitoring of AI’s human-like capabilities, particularly in general-purpose models, to ensure safety and alignment with human values.

Exploring the interplay between AI and human logic in mathematical problem-solving

This paper investigates the dynamic interplay between Artificial Intelligence (AI) and human logic in the domain of mathematical problem-solving. By critically examining a series of case studies, we compare the efficacy of AI-generated solutions, particularly those offered by ChatGPT, against traditional human problem-solving methods. The study employs various mathematical challenges, ranging from abstract logical puzzles to applied numerical problems, to evaluate AI's problem-solving approach and alignment with human cognitive processes. Our analysis highlights instances where AI's computational strategies complement or diverge from human reasoning, shedding light on AI's potential and limitations in deciphering mathematical problems. Furthermore, we explore the implications of integrating AI tools in educational contexts, specifically their role in enhancing students' mathematical problem-solving skills. The paper aims to contribute to the ongoing discourse on the optimal utilization of AI in education, proposing a balanced approach that leverages AI's computational power while fostering the depth and creativity of human logic. Through this comparative study, we advocate for a collaborative model where AI and human reasoning merge to enrich the educational landscape, particularly in the teaching and learning of mathematics.

On the Interplay Between Interpretation and Reasoning in Compelling Fallacies.

We investigate the articulation between domain-general reasoning and interpretive processes in failures of deductive reasoning. We focus on illusory inferences from disjunction-like elements, a broad class of deductive fallacies studied in some detail over the past 15 years. These fallacies have received accounts grounded in reasoning processes, holding that human reasoning diverges from normative standards. A subset of these fallacies, however, can be analyzed differently: human reasoning is not to blame, instead the premises were interpreted in a nonobvious, yet perfectly predictable and reasonable way. Once we consider these interpretations, the apparent fallacious conclusion is no mistake at all. We give a two-factor account of these fallacies that incorporates both reasoning-based elements and interpretive elements, showing that they are not in real competition. We present novel experimental evidence in favor of our theory. Cognitive load such as induced by a dual-task design is known to hinder the interpretive mechanisms at the core of interpretation-based accounts of the fallacies of interest. In the first experiment of its kind using this paradigm with an inferential task instead of a simpler truth-value-judgment task, we found that the manipulation affected more strongly those illusions where our theory predicts that interpretive processes are at play. We conclude that the best way forward for the field to investigate the elusive line between reasoning and interpretation requires combining theories and methodologies from linguistic semantics and the psychology ofreasoning.

Service quality management under risk prioritization and imprecise information: a hybrid fuzzy approach

PurposeConventional risk prioritization methods rely on crisp inputs but struggle with imprecise data and hesitancy, resulting in inaccurate assessments that affect service and information quality and performance monitoring. This study proposes a fuzzy data-driven risk prioritization model for service quality under imprecise information.Design/methodology/approachEnterprise risk management is crucial for service quality management, ensuring effective identification, assessment and mitigation of risks impacting service delivery and customer satisfaction. This paper proposes a fuzzy data-driven multi-criteria model for risk prioritization involving multiple decision-makers. It introduces a hybrid method combining intuitionistic and hesitant fuzzy group decision-making to assess better and prioritize risks based on decision-maker preferences.FindingsThe proposed hybrid fuzzy model improves service quality in business operations by efficiently representing uncertain information in traditional frameworks. It helps identify potential risks in advance and enhances control over business operations, enabling organizations to benchmark service quality and identify best practices. Accordingly, organizations acquire information and background knowledge to benchmark their service quality. This, in turn, improves service quality under performance management.Research limitations/implicationsDespite the advantages of fuzzy models in risk prioritization, such as mimicking human reasoning more accurately, their complexity can hinder adoption. The intricate computational steps may deter shop-floor managers who prefer the more straightforward conventional crisp RPN approach, which is easier to understand and implement. However, while developing a hybrid fuzzy risk prioritization model may require more effort, its benefits become apparent over time. Once developed, the model can be integrated into software applications, allowing decision-makers to use it easily. This integration simplifies fuzzy computations and enhances risk prioritization, leading to more informed decision-making and improved risk management in the long term.Practical implicationsThe proposed robust fuzzy framework improves risk management by integrating uncertain information and multiple decision-makers expertise, leading to more reliable outputs that enhance strategic decisions and operational efficiency.Originality/valueWe validate the proposed approach at an integrated steel plant’s risk management process, covering broad areas of the service quality domain. To the best of our knowledge, no study exists in existing literature attempting to explore the efficacy of the proposed hybrid fuzzy approach in risk management practices at prime sectors like steel. The study’s novelty is backed by this validation experiment, which indicates that the effectiveness of the results obtained from the proposed multi-attribute hybrid fuzzy methodology is more practical. The model’s outcome substantially adds value to the current risk assessment and prioritization literature that significantly affects service quality.

Exploring Facilitators and Barriers to Managers’ Adoption of AI-Based Systems in Decision Making: A Systematic Review

Introduction—Decision making (DM) is a fundamental responsibility for managers, with significant implications for organizational performance and strategic direction. The increasing complexity of modern business environments, along with the recognition of human reasoning limitations related to cognitive and emotional biases, has led to a heightened interest in harnessing emerging technologies like Artificial Intelligence (AI) to enhance DM processes. However, a notable disparity exists between the potential of AI and its actual adoption within organizations, revealing skepticism and practical challenges associated with integrating AI into complex managerial DM scenarios. This systematic literature review aims to address this gap by examining the factors that influence managers’ adoption of AI in DM. Methods—This study adhered to the PRISMA guidelines. Articles from 2010 to 2024 were selected from the Scopus database using specific keywords. Eligible studies were included after rigorous screening and quality assessment using checklist tools. Results—From 202 articles screened, a data synthesis of 16 eligible studies revealed seven major interconnected factors acting as key facilitators or barriers to AI integration within organizations. These factors—Managers’ Perceptions of AI, Ethical Factors, Psychological and Individual Factors, Social and Psychosocial Factors, Organizational Factors, External Factors, and Technical and Design Characteristics of AI—were then organized into a complex analytical framework informed by existing theoretical constructs. Discussion—This contribution provides valuable insights into how managers perceive and interact with AI systems, as well as the conditions necessary for successful integration into organizational DM processes.

Going Round in Circles: A Cognitive Bias in Geometric Reasoning.

Deductive reasoning is essential to most of our scientific and technological achievements and is a crucial component to scientific education. In Western culture, deductive reasoning first emerged as a dedicated mode of thinking in the field of geometry, but the cognitive mechanisms behind this major intellectual achievement remain largely understudied. Here, we report an unexpected cognitive bias in geometric reasoning that challenges existing theories of human deductive reasoning. Over two experiments involving almost 250 participants, we show that educated adults systematically mistook as valid a set of elementary invalid inferences with points and circles in the Euclidean plane. Our results suggest that people got "locked" on unwarranted conclusions because they tended to represent geometric premisses in specific ways and they mainly relied on translating, but not scaling, the circles when searching for possible conclusions. We conducted two further experiments to test these hypotheses and found confirmation for them. Although mathematical reasoning is considered as the hallmark of rational thinking, our findings indicate that it is not exempt from cognitive biases and is subject to fundamental counter-intuitions. Our empirical investigations into the source of this bias provide some insights into the cognitive mechanisms underlying geometric deduction, and thus shed light on the cognitive roots of intuitive mathematical reasoning.

A text analytics study on the alignment of an Indonesian Muslim community’s municipal corporation policy with Sustainable Development Goals

PurposeThis study analyses the supervision policy of Municipal Corporations (MCs) in the Indonesian Muslim community regarding its alignment with the United Nations Sustainable Development Goals (SDGs).Design/methodology/approachAs a single case study, purposive sampling was adopted to select a municipal corporation policy issued by Ciamis Regency, West Java, Indonesia. A novel mixed methods approach, combining computer analytics and human intelligence, was introduced to perform text analytics.FindingsText mining identified the most frequently occurring words — e.g., ‘perumda’ (municipal corporation), ‘daerah’ (regional), ‘pengawas’ (supervisor), ‘peraturan’ (regulation) — from the sample but found no single word indicating business alignment with SDGs. Further qualitative inductive analysis was conducted, revealing the critical role of MCs’ supervisory boards in business planning, execution, and reporting to align the MC businesses with local SDG initiatives.Originality/valueAligning MC business activities with local SDG actions is entirely within the authority of supervisory boards which demands transformational leadership. This research pioneered an innovative blend of computer-assisted techniques and human reasoning to investigate the supervision policies of MCs concerning local SDG actions, with evidence from a Muslim community in Indonesia.

Human Reasoning for Visual Analytics in the Moment of Emergent Artificial Intelligence

Human Reasoning for Visual Analytics in the Moment of Emergent Artificial Intelligence

Sequential merging and construction of rankings as cognitive logic

We introduce and evaluate a cognitively inspired formal reasoning approach that sequentially applies a combination of a belief merging operator and a ranking construction operator. The approach is inspired by human propositional reasoning, which is understood here as a sequential process in which the agent constructs a new epistemic state in each task step according to newly acquired information. Formally, we model epistemic states by Spohn's ranking functions. The posterior representation of the epistemic state is obtained by merging the prior ranking function and a ranking function constructed from the new piece of information. We denote this setup as the sequential merging approach. The approach abstracts from the concrete merging operation and abstracts from the concrete way of constructing a ranking function according to new information. We formally show that sequential merging is capable of predicting with theoretical maximum achievable accuracy. Various instantiations of our approach are benchmarked on data from a psychological experiment, demonstrating that sequential merging provides formal reasoning methods that are cognitively more adequate than classical logic.

Comparative analysis of moral decision-making and trust dynamics: human reasoning vs. ChatGPT-3 narratives

Comparative analysis of moral decision-making and trust dynamics: human reasoning vs. ChatGPT-3 narratives

Assessing the need for coronary angiography in high-risk acute coronary syndrome patients using artificial intelligence and computed tomography

Abstract Background/Introduction The use of coronary computed tomography angiography (CCTA) in managing high-risk acute coronary syndrome (ACS) patients is increasingly common. Yet, its role remains unvalidated in this context, posing challenges to traditional clinical decision-making algorithms. Physicians, accustomed to a hierarchical and structured approach involving symptom assessment, electrocardiograms, and biomarkers for deciding on the necessity of invasive coronary angiography (ICA), now face dilemmas when CCTA results challenge established diagnostic pathways and contradict clinical decisions based on usual criteria. Meanwhile, the potential of artificial intelligence (AI) which processes information in a way fundamentally different from human clinical reasoning, has been shown to aid decision-making in cardiovascular medicine. Yet, it has not been tested in this acute setting. Purpose This study aimed to evaluate the efficacy of the Chat Generative Pre-trained Transformer (ChatGPT) in guiding the need for ICA in high-risk ACS patients based on both standard clinical data and CCTA findings. Methods This investigation is a sub-study of a larger research project where high-risk ACS patients (all having a theoretical guideline-based indication for ICA) underwent CCTA prior to ICA. ChatGPT analyzed clinical vignettes containing patient data, electrocardiograms, troponin levels, and CCTA results to determine the necessity of ICA. The AI's recommendations were then compared to actual ICA findings to assess its decision-making accuracy. Results Among 86 patients with a clear indication for ICA, ChatGPT recommended against it for 19 patients, 16 of whom indeed had no significant findings at ICA (Figure). For 67 patients, ChatGPT advised proceeding with ICA, and a significant lesion was confirmed in 58 of them. Consequently, ChatGPT's overall accuracy stood at 86%, with a sensitivity of 95% and a specificity of 64%. The model's negative predictive value was 84%, and its positive predictive value was 87%. Conclusion These preliminary data suggest that ChatGPT can effectively assist in guiding the need for ICA among high-risk ACS patients, potentially reducing unnecessary procedures but further clinical studies to refine AI's role in clinical decision-making are required.ChatGPT's decision as per ICA's results