Search

Multiple research problems specific to advertising emerge from there, including whether and how the advertising industry should address the ethical and legal issues surrounding the possibility of AI-generated advertising content containing false or misleading information as well as strategies for designing and implementing AI-powered advertising in a socially responsible and ethical manner. In her editorial, the previous editor, Shelly Rodgers ([16]) proposed an AI classification schema to systematically understand and develop subdomains in the emerging AI advertising research field along the following three dimensions: AI type, AI function, and learning type. The lack of transparency and consumers' limited ability to understand AI technology and the mechanism and processes of AI-powered advertising would affect advertising outcomes. Since ChatGPT was released on November 30, 2022, followed by Microsoft's announcement of its artificial intelligence (AI)-powered new Bing search engine on February 7, 2023, and Google's Bard release on March 21, 2023, it seems AI generally has taken over conversations across all sectors of society. [Extracted from the article]

Background Artificial intelligence chatbots have demonstrated potential to enhance clinical decision-making and streamline health care workflows, potentially alleviating administrative burdens. However, the contribution of AI chatbots to radiologic decision-making for clinical scenarios remains insufficiently explored. This study evaluates the accuracy and reliability of four prominent Large Language Models (LLMs)—Microsoft Bing, Claude, ChatGPT 3.5, and Perplexity—in offering clinical decision support for initial imaging for suspected pulmonary embolism (PE). Methods Open-ended (OE) and select-all-that-apply (SATA) questions were crafted, covering four variants of case scenarios of PE in-line with the American College of Radiology Appropriateness Criteria. These questions were presented to the LLMs by three radiologists from diverse geographical regions and setups. The responses were evaluated based on established scoring criteria, with a maximum achievable score of 2 points for OE responses and 1 point for each correct answer in SATA questions. To enable comparative analysis, scores were normalized (score divided by the maximum achievable score). Result In OE questions, Perplexity achieved the highest accuracy (0.83), while Claude had the lowest (0.58), with Bing and ChatGPT each scoring 0.75. For SATA questions, Bing led with an accuracy of 0.96, Perplexity was the lowest at 0.56, and both Claude and ChatGPT scored 0.6. Overall, OE questions saw higher scores (0.73) compared to SATA (0.68). There is poor agreement among radiologists' scores for OE (Intraclass Correlation Coefficient [ICC] = −0.067, p = 0.54), while there is strong agreement for SATA (ICC = 0.875, p < 0.001). Conclusion The study revealed variations in accuracy across LLMs for both OE and SATA questions. Perplexity showed superior performance in OE questions, while Bing excelled in SATA questions. OE queries yielded better overall results. The current inconsistencies in LLM accuracy highlight the importance of further refinement before these tools can be reliably integrated into clinical practice, with a need for additional LLM fine-tuning and judicious selection by radiologists to achieve consistent and reliable support for decision-making. [ABSTRACT FROM AUTHOR]

Dialogue systems, from Virtual Personal Assistants such as Siri, Cortana, and Alexa to state-of-the-art systems such as BlenderBot3 and ChatGPT, are already widely available, used in a variety of applications, and are increasingly part of many people's lives. However, the task of enabling them to use empathetic language more convincingly is still an emerging research topic. Such systems generally make use of complex neural networks to learn the patterns of typical human language use, and the interactions in which the systems participate are usually mediated either via interactive text-based or speech-based interfaces. In human–human interaction, empathy has been shown to promote prosocial behaviour and improve interaction. In the context of dialogue systems, to advance the understanding of how perceptions of empathy affect interactions, it is necessary to bring greater clarity to how empathy is measured and assessed. Assessing the way dialogue systems create perceptions of empathy brings together a range of technological, psychological, and ethical considerations that merit greater scrutiny than they have received so far. However, there is currently no widely accepted evaluation method for determining the degree of empathy that any given system possesses (or, at least, appears to possess). Currently, different research teams use a variety of automated metrics, alongside different forms of subjective human assessment such as questionnaires, self-assessment measures and narrative engagement scales. This diversity of evaluation practice means that, given two DSs, it is usually impossible to determine which of them conveys the greater degree of empathy in its dialogic exchanges with human users. Acknowledging this problem, the present article provides an overview of how empathy is measured in human–human interactions and considers some of the ways it is currently measured in human–DS interactions. Finally, it introduces a novel third-person analytical framework, called the Empathy Scale for Human–Computer Communication (ESHCC), to support greater uniformity in how perceived empathy is measured during interactions with state-of-the-art DSs. [ABSTRACT FROM AUTHOR]

Large language models such as ChatGPT enable users to automatically produce text but also raise ethical concerns, for example about authorship and deception. This paper analyses and discusses some key philosophical assumptions in these debates, in particular assumptions about authorship and language and—our focus—the use of the appearance/reality distinction. We show that there are alternative views of what goes on with ChatGPT that do not rely on this distinction. For this purpose, we deploy the two phased approach of deconstruction and relate our finds to questions regarding authorship and language in the humanities. We also identify and respond to two common counter-objections in order to show the ethical appeal and practical use of our proposal. [ABSTRACT FROM AUTHOR]

Classical natural language processing endeavored to understand the language of native speakers. When this proved to lie beyond the horizon, a scaled-down version settled for text analysis and processing but retained the old name and acronym. But text ≠ language. Any combination of signs and symbols qualifies as text. Language presupposes meaning, which is what connects it to real life. Failing to distinguish between the two results in confusing humanoids (machines thinking like humans) with machinoids (humans thinking like machines). As scientific English (SciEng) became the lingua franca of science, it has acquired all the traits of a machine language: reduced vocabulary, where fewer and fewer words have taken on more and more meanings; prescribed use of pronouns; depersonalized rigid syntactic forms and rules of composition. Compliance with SciEng standards can be automatically verified, which means that Sci Eng can be automatically imitated, what is referred to as AI writing (ChatGPT). The article discusses an attempt to automatically correct deviations from the rules by what is touted as AI editing. [ABSTRACT FROM AUTHOR]

A large language model (LLM) is an artificial intelligence (AI) model that uses natural language processing (NLP) to understand, interpret, and generate human-like language responses from unstructured text input. Its real-time response capabilities and eloquent dialogue enhance the interactive user experience in human–AI communication like never before. By gathering several sources on the internet, LLM chatbots can interact and respond to a wide range of queries, including problem solving, text summarization, and creating informative notes. Since ophthalmology is one of the medical fields integrating image analysis, telemedicine, AI, and other technologies, LLMs are likely to play an important role in eye care in the near future. This review summarizes the performance and potential applicability of LLMs in ophthalmology according to currently available publications. [ABSTRACT FROM AUTHOR]

Introduction: OpenAI recently introduced the ability to create custom generative pre-trained transformers (cGPTs) using text-based instruction and/or external documents using retrieval-augmented generation (RAG) architecture without coding knowledge. This study aimed to analyze the features of ophthalmology-related cGPTs and explore their potential utilities. Methods: Data collection took place on January 20 and 21, 2024, and custom GPTs were found by entering ophthalmology keywords into the "Explore GPTS" section of the website. General and specific features of cGPTs were recorded, such as knowledge other than GPT-4 training data. The instruction and description sections were analyzed for compatibility using the Likert scale. We analyzed two custom GPTs with the highest Likert score in detail. We attempted to create a convincingly presented yet potentially harmful cGPT to test safety features. Results: We analyzed 22 ophthalmic cGPTs, of which 55% were for general use and the most common subspecialty was glaucoma (18%). Over half (55%) contained knowledge other than GPT-4 training data. The representation of the instructions through the description was between "Moderately representative" and "Very representative" with a median Likert score of 3.5 (IQR 3.0–4.0). The instruction word count was significantly associated with Likert scores (P = 0.03). Tested cGPTs demonstrated potential for specific conversational tone, information, retrieval and combining knowledge from an uploaded source. With these safety settings, creating a malicious GPT was possible. Conclusions: This is the first study to our knowledge to examine the GPT store for a medical field. Our findings suggest that these cGPTs can be immediately implemented in practice and may offer more targeted and effective solutions compared to the standard GPT-4. However, further research is necessary to evaluate their capabilities and limitations comprehensively. The safety features currently appear to be rather limited. It may be helpful for the user to review the instruction section before using a cGPT. [ABSTRACT FROM AUTHOR]

Technological innovations since the turn of the century have created opportunities to increase public participation in notice and comment rulemaking, increase the efficiency of the process, and increase the quality of the rules adopted by agencies. For some rules, online rulemaking and social media have facilitated increased public participation, but have not necessarily facilitated improvements in the quality of public comments. In addition, in some cases, the transformation of the process has created new challenges for government agencies by making it easier for supporters or opponents of rules to flood agencies with duplicative and potentially false comments to which the agencies must respond. Artificial intelligence, including ChatGPT, is the latest tech trend to create opportunities to transform notice and comment rulemaking. If ChatGPT is only used by members of the public and organizations as a tool to draft comments, it may increase the involvement of the public in the process and assist them in drafting clear and intelligible comments. However, it is unlikely to improve the quality of public comments that they provide to agencies, because it will not help them understand the type of information that agencies are seeking in public comments. In addition, if ChatGPT is used to any significant extent in drafting comments, it could create challenges for agencies by flooding agencies with significantly greater numbers of seemingly unique comments, most of which merely express sentiment, values or preferences, and many of which may include false information. There is, however, an alternative path forward for ChatGPT and artificial intelligence in notice and comment rulemaking. If agencies use ChatGPT to create tools that help members of the public understand proposed rules, the rulemaking process and the types of comments that are effective in that process, agencies may enable members of the public to provide them with more useful comments. In addition, agencies may be able to use other artificial intelligence tools to make it easier to organize and analyze the comments they receive and to notify members of the public about proposed rules that might be of interest to them. Embracing these alternative uses of artificial intelligence and ChatGPT could generate some of the gains in public participation that other technological tools have promised, but not delivered. [ABSTRACT FROM AUTHOR]

In part two of this mini-series, we evaluate the range of machine-learning tools now available for application to veterinary clinical text-mining. These tools will be vital to automate extraction of information from large datasets of veterinary clinical narratives curated by projects such as the Small Animal Veterinary Surveillance Network (SAVSNET) and VetCompass, where volumes of millions of records preclude reading records and the complexities of clinical notes limit usefulness of more "traditional" text-mining approaches. We discuss the application of various machine learning techniques ranging from simple models for identifying words and phrases with similar meanings to expand lexicons for keyword searching, to the use of more complex language models. Specifically, we describe the use of language models for record annotation, unsupervised approaches for identifying topics within large datasets, and discuss more recent developments in the area of generative models (such as ChatGPT). As these models become increasingly complex it is pertinent that researchers and clinicians work together to ensure that the outputs of these models are explainable in order to instill confidence in any conclusions drawn from them. [ABSTRACT FROM AUTHOR]

Zero-shot image classification enables the recognition of new categories without requiring additional training data, thereby enhancing the model's generalization capability when specific training are unavailable. This paper introduces a zero-shot image classification framework to recognize new categories that are unseen during training by distilling knowledge from foundation models. Specifically, we first employ ChatGPT and DALL-E to synthesize reference images of unseen categories from text prompts. Then, the test image is aligned with text and reference images using CLIP and DINO to calculate the logits. Finally, the predicted logits are aggregated according to their confidence to produce the final prediction. Experiments are conducted on multiple datasets, including MNIST, SVHN, CIFAR-10, CIFAR-100, and TinyImageNet. The results demonstrate that our method can significantly improve classification accuracy compared to previous approaches, achieving AUROC scores of over 96% across all test datasets. Our code is available at https://github.com/1134112149/MICW-ZIC. [ABSTRACT FROM AUTHOR]

Large language models such as ChatGPT have been shown to excel in solving complex math problems. However, they cannot solve basic arithmetic problems such as 758*639 = 484,362. This makes us ponder if LLMs have been trained to solve math and science problems in the right way. When a student learns math at school, she or he starts with arithmetic, then moves to word problems, polynomials, and calculus. Each skill she or he acquires will be used in the next stage to solve more advanced problems. In this paper we propose Scaffolding Learning for LLMs, which imitates how a student learns a subject in a step-by-step manner. For example, we first train an LLM to perform highly specific operations such as multiplication and division, and then apply such "skills" in a more generic task such as solving word problems. This is related to Curriculum Training, which trains a model on tasks following a specific order, such as training on easy tasks first and then gradually increases the difficulty. Our proposed approach goes from specific tasks to generic ones, which can be considered as a special case of Curriculum Training. Our empirical studies show that when an LLM has "mastered" a specific skill, only a small amount of training is required to teach it to apply the skill to a more generic application. [ABSTRACT FROM AUTHOR]

Backgrounds: The use of large language models (LLMs) in medicine can help physicians improve the quality and effectiveness of health care by increasing the efficiency of medical information management, patient care, medical research, and clinical decision-making. Methods: We collected 34 frequently asked questions about glucocorticoid-induced osteoporosis (GIOP), covering topics related to the disease's clinical manifestations, pathogenesis, diagnosis, treatment, prevention, and risk factors. We also generated 25 questions based on the 2022 American College of Rheumatology Guideline for the Prevention and Treatment of Glucocorticoid-Induced Osteoporosis (2022 ACR-GIOP Guideline). Each question was posed to the LLM (ChatGPT-3.5, ChatGPT-4, and Google Gemini), and three senior orthopedic surgeons independently rated the responses generated by the LLMs. Three senior orthopedic surgeons independently rated the answers based on responses ranging between 1 and 4 points. A total score (TS) > 9 indicated 'good' responses, 6 ≤ TS ≤ 9 indicated 'moderate' responses, and TS < 6 indicated 'poor' responses. Results: In response to the general questions related to GIOP and the 2022 ACR-GIOP Guidelines, Google Gemini provided more concise answers than the other LLMs. In terms of pathogenesis, ChatGPT-4 had significantly higher total scores (TSs) than ChatGPT-3.5. The TSs for answering questions related to the 2022 ACR-GIOP Guideline by ChatGPT-4 were significantly higher than those for Google Gemini. ChatGPT-3.5 and ChatGPT-4 had significantly higher self-corrected TSs than pre-corrected TSs, while Google Gemini self-corrected for responses that were not significantly different than before. Conclusions: Our study showed that Google Gemini provides more concise and intuitive responses than ChatGPT-3.5 and ChatGPT-4. ChatGPT-4 performed significantly better than ChatGPT3.5 and Google Gemini in terms of answering general questions about GIOP and the 2022 ACR-GIOP Guidelines. ChatGPT3.5 and ChatGPT-4 self-corrected better than Google Gemini. [ABSTRACT FROM AUTHOR]

Background: Recurrent and metastatic head and neck squamous cell carcinoma (HNSCC) is characterized by a complex therapeutic management that needs to be discussed in multidisciplinary tumor boards (MDT). While artificial intelligence (AI) improved significantly to assist healthcare professionals in making informed treatment decisions for primary cases, an application in the even more complex recurrent/metastatic setting has not been evaluated yet. This study also represents the first evaluation of the recently published LLM ChatGPT 4o, compared to ChatGPT 4.0 for providing therapy recommendations. Methods: The therapy recommendations for 100 HNSCC cases generated by each LLM, 50 cases of recurrence and 50 cases of distant metastasis were evaluated by two independent reviewers. The primary outcome measured was the quality of the therapy recommendations measured by the following parameters: clinical recommendation, explanation, and summarization. Results: In this study, ChatGPT 4o and 4.0 provided mostly general answers for surgery, palliative care, or systemic therapy. ChatGPT 4o proved to be 48.5% faster than ChatGPT 4.0. For clinical recommendation, explanation, and summarization both LLMs obtained high scores in terms of performance of therapy recommendations, with no significant differences between both LLMs, but demonstrated to be mostly an assisting tool, requiring validation by an experienced clinician due to a lack of transparency and sometimes recommending treatment modalities that are not part of the current treatment guidelines. Conclusion: This research demonstrates that ChatGPT 4o and 4.0 share a similar performance, while ChatGPT 4o is significantly faster. Since the current versions cannot tailor therapy recommendations, and sometimes recommend incorrect treatment options and lack information on the source material, advanced AI models at the moment can merely assist in the MDT setting for recurrent/metastatic HNSCC. [ABSTRACT FROM AUTHOR]

Background: ChatGPT, a recently developed artificial intelligence (AI) chatbot, has demonstrated improved performance in examinations in the medical field. However, thus far, an overall evaluation of the potential of ChatGPT models (ChatGPT-3.5 and GPT-4) in a variety of national health licensing examinations is lacking. This study aimed to provide a comprehensive assessment of the ChatGPT models' performance in national licensing examinations for medical, pharmacy, dentistry, and nursing research through a meta-analysis. Methods: Following the PRISMA protocol, full-text articles from MEDLINE/PubMed, EMBASE, ERIC, Cochrane Library, Web of Science, and key journals were reviewed from the time of ChatGPT's introduction to February 27, 2024. Studies were eligible if they evaluated the performance of a ChatGPT model (ChatGPT-3.5 or GPT-4); related to national licensing examinations in the fields of medicine, pharmacy, dentistry, or nursing; involved multiple-choice questions; and provided data that enabled the calculation of effect size. Two reviewers independently completed data extraction, coding, and quality assessment. The JBI Critical Appraisal Tools were used to assess the quality of the selected articles. Overall effect size and 95% confidence intervals [CIs] were calculated using a random-effects model. Results: A total of 23 studies were considered for this review, which evaluated the accuracy of four types of national licensing examinations. The selected articles were in the fields of medicine (n = 17), pharmacy (n = 3), nursing (n = 2), and dentistry (n = 1). They reported varying accuracy levels, ranging from 36 to 77% for ChatGPT-3.5 and 64.4–100% for GPT-4. The overall effect size for the percentage of accuracy was 70.1% (95% CI, 65–74.8%), which was statistically significant (p < 0.001). Subgroup analyses revealed that GPT-4 demonstrated significantly higher accuracy in providing correct responses than its earlier version, ChatGPT-3.5. Additionally, in the context of health licensing examinations, the ChatGPT models exhibited greater proficiency in the following order: pharmacy, medicine, dentistry, and nursing. However, the lack of a broader set of questions, including open-ended and scenario-based questions, and significant heterogeneity were limitations of this meta-analysis. Conclusions: This study sheds light on the accuracy of ChatGPT models in four national health licensing examinations across various countries and provides a practical basis and theoretical support for future research. Further studies are needed to explore their utilization in medical and health education by including a broader and more diverse range of questions, along with more advanced versions of AI chatbots. [ABSTRACT FROM AUTHOR]

The value of generative AI tools in higher education has received considerable attention. Although there are many proponents of its value as a learning tool, many are concerned with the issues regarding academic integrity and its use by students to compose written assessments. This study evaluates and compares the output of three commonly used generative AI tools, ChatGPT, Bing and Bard. Each AI tool was prompted with an essay question from undergraduate (UG) level 4 (year 1), level 5 (year 2), level 6 (year 3) and postgraduate (PG) level 7 biomedical sciences courses. Anonymised AI generated output was then evaluated by four independent markers, according to specifed marking criteria and matched to the Frameworks for Higher Education Qualifcations (FHEQ) of UK level descriptors. Percentage scores and ordinal grades were given for each marking criteria across AI generated papers, inter-rater reliability was calculated using Kendall's coefcient of concordance and generative AI performance ranked. Across all UG and PG levels, ChatGPT performed better than Bing or Bard in areas of scientifc accuracy, scientifc detail and context. All AI tools performed consistently well at PG level compared to UG level, although only ChatGPT consistently met levels of high attainment at all UG levels. ChatGPT and Bing did not provide adequate references, while Bing falsifed references. In conclusion, generative AI tools are useful for providing scientifc information consistent with the academic standards required of students in written assignments. These fndings have broad implications for the design, implementation and grading of written assessments in higher education. [ABSTRACT FROM AUTHOR]

There is a strong demand for science, technology, engineering, and mathematics (STEM) education in Africa, which is crucial in driving the industrial revolution. Efforts to promote STEM education have led to innovations in the teaching–learning process where emerging technologies are injected into instructional processes. Artificial intelligence (AI) has been at the forefront of changing approaches to instructional activities in higher education. In African higher education, academics and undergraduates utilize AI tools for various purposes. After consulting various studies, a systematic literature review approach was used to examine the strategic goals of AI integration among STEM academics and undergraduates in African higher education. The systematic review was carried out using the PRISMA procedure. Our objective was to identify the existing gaps and challenges to provide research guidance to aspiring researchers seeking to contribute significantly to integrating AI into STEM education in African higher education. We searched for reports covering ten years (2015–2024) on the topic, but we could identify and analyze only 12 available studies that were published within three years (2022–2024). Based on our findings, AI tools are strategically utilized by STEM academics in African higher education to engage students in learning activities, administrative processes, information searches, content generation, paraphrasing academic content, grammar checks, teaching, and research. Similarly, STEM undergraduates utilize AI tools for information searching, self-learning, content generation, paraphrasing, grammar checking, and research. The most used AI tool by both STEM academics and undergraduates for various purposes is the ChatGPT. We hold an optimistic position that AI literacy advocacy and advancements in research on adopting AI tools for STEM education in African higher education will enhance the output of STEM education, contributing to the pursuit of sustainable development driven by African higher education. [ABSTRACT FROM AUTHOR]

Background: Uveitis is the ophthalmic subfield dealing with a broad range of intraocular inflammatory diseases. With the raising importance of LLM such as ChatGPT and their potential use in the medical field, this research explores the strengths and weaknesses of its applicability in the subfield of uveitis. Methods: A series of highly clinically relevant questions were asked three consecutive times (attempts 1, 2 and 3) of the LLM regarding current uveitis cases. The answers were classified on whether they were accurate and sufficient, partially accurate and sufficient or inaccurate and insufficient. Statistical analysis included descriptive analysis, normality distribution, non-parametric test and reliability tests. References were checked for their correctness in different medical databases. Results: The data showed non-normal distribution. Data between subgroups (attempts 1, 2 and 3) was comparable (Kruskal-Wallis H test, p-value = 0.7338). There was a moderate agreement between attempt 1 and attempt 2 (Cohen's kappa, ĸ = 0.5172) as well as between attempt 2 and attempt 3 (Cohen's kappa, ĸ = 0.4913). There was a fair agreement between attempt 1 and attempt 3 (Cohen's kappa, ĸ = 0.3647). The average agreement was moderate (Cohen's kappa, ĸ = 0.4577). Between the three attempts together, there was a moderate agreement (Fleiss' kappa, ĸ = 0.4534). A total of 52 references were generated by the LLM. 22 references (42.3%) were found to be accurate and correctly cited. Another 22 references (42.3%) could not be located in any of the searched databases. The remaining 8 references (15.4%) were found to exist, but were either misinterpreted or incorrectly cited by the LLM. Conclusion: Our results demonstrate the significant potential of LLMs in uveitis. However, their implementation requires rigorous training and comprehensive testing for specific medical tasks. We also found out that the references made by ChatGPT 4.o were in most cases incorrect. LLMs are likely to become invaluable tools in shaping the future of ophthalmology, enhancing clinical decision-making and patient care. [ABSTRACT FROM AUTHOR]

Large language models (LLMs) are generating interest in medical settings. For example, LLMs can respond coherently to medical queries by providing plausible differential diagnoses based on clinical notes. However, there are many questions to explore, such as evaluating differences between open- and closed-source LLMs as well as LLM performance on queries from both medical and non-medical users. In this study, we assessed multiple LLMs, including Llama-2-chat, Vicuna, Medllama2, Bard/Gemini, Claude, ChatGPT3.5, and ChatGPT-4, as well as non-LLM approaches (Google search and Phenomizer) regarding their ability to identify genetic conditions from textbook-like clinician questions and their corresponding layperson translations related to 63 genetic conditions. For open-source LLMs, larger models were more accurate than smaller LLMs: 7b, 13b, and larger than 33b parameter models obtained accuracy ranges from 21%–49%, 41%–51%, and 54%–68%, respectively. Closed-source LLMs outperformed open-source LLMs, with ChatGPT-4 performing best (89%–90%). Three of 11 LLMs and Google search had significant performance gaps between clinician and layperson prompts. We also evaluated how in-context prompting and keyword removal affected open-source LLM performance. Models were provided with 2 types of in-context prompts: list-type prompts, which improved LLM performance, and definition-type prompts, which did not. We further analyzed removal of rare terms from descriptions, which decreased accuracy for 5 of 7 evaluated LLMs. Finally, we observed much lower performance with real individuals' descriptions; LLMs answered these questions with a maximum 21% accuracy. Large language models (LLMs) are rapidly changing the biomedical landscape. We compared the ability of multiple LLMs and other search techniques in identifying genetic conditions from text descriptions, including medical and non-medical (colloquial) descriptions. We also investigated the effects of different prompting approaches and the removal of rare terms. [ABSTRACT FROM AUTHOR]

Artificial intelligence (AI) has proven to be a valuable tool for education. It might perform different functions to facilitate teaching and learning experiences, with a learning companion being one of them. However, teachers often have little understanding of the pool of opportunities provided by AI-powered tools. Having a learning companion in a language classroom is crucial. However, getting involved in human–computer interaction requires understanding of the services that the machine can offer, whether it is a text written as a response to students' input, feedback provided to help students gauge their language level or translation prompt helping students understand the reading text. This research aims at displaying the range of AI services that are applicable in the foreign language classroom and grouping them according to their functionality. The authors selected 150 available in open access online instruments that apply AI and evaluated them against the criteria, which allowed them to organize these tools into categories according to their possible applications in the classroom. This might help teachers make informed decisions about the appropriate tool for certain teaching situations or recommend to their students a tool that could enhance the learning experience. [ABSTRACT FROM AUTHOR]

Objective: The integration of artificial intelligence (AI) in healthcare education is inevitable. Understanding the proficiency of generative AI in different languages to answer complex questions is crucial for educational purposes. The study objective was to compare the performance ChatGPT-4 and Gemini in answering Virology multiple-choice questions (MCQs) in English and Arabic, while assessing the quality of the generated content. Both AI models' responses to 40 Virology MCQs were assessed for correctness and quality based on the CLEAR tool designed for evaluation of AI-generated content. The MCQs were classified into lower and higher cognitive categories based on the revised Bloom's taxonomy. The study design considered the METRICS checklist for the design and reporting of generative AI-based studies in healthcare. Results: ChatGPT-4 and Gemini performed better in English compared to Arabic, with ChatGPT-4 consistently surpassing Gemini in correctness and CLEAR scores. ChatGPT-4 led Gemini with 80% vs. 62.5% correctness in English compared to 65% vs. 55% in Arabic. For both AI models, superior performance in lower cognitive domains was reported. Both ChatGPT-4 and Gemini exhibited potential in educational applications; nevertheless, their performance varied across languages highlighting the importance of continued development to ensure the effective AI integration in healthcare education globally. [ABSTRACT FROM AUTHOR]

Background: Cervical cancer (CC) and breast cancer (BC) threaten women's well-being, influenced by health-related stigma and a lack of reliable information, which can cause late diagnosis and early death. ChatGPT is likely to become a key source of health information, although quality concerns could also influence health-seeking behaviours. Methods: This cross-sectional online survey compared ChatGPT's responses to five physicians specializing in mammography and five specializing in gynaecology. Twenty frequently asked questions about CC and BC were asked on 26th and 29th of April, 2023. A panel of seven experts assessed the accuracy, consistency, and relevance of ChatGPT's responses using a 7-point Likert scale. Responses were analyzed for readability, reliability, and efficiency. ChatGPT's responses were synthesized, and findings are presented as a radar chart. Results: ChatGPT had an accuracy score of 7.0 (range: 6.6-7.0) for CC and BC questions, surpassing the highest-scoring physicians (P < 0.05). ChatGPT took an average of 13.6 s (range: 7.6-24.0) to answer each of the 20 questions presented. Readability was comparable to that of experts and physicians involved, but ChatGPT generated more extended responses compared to physicians. The consistency of repeated answers was 5.2 (range: 3.4-6.7). With different contexts combined, the overall ChatGPT relevance score was 6.5 (range: 4.8-7.0). Radar plot analysis indicated comparably good accuracy, efficiency, and to a certain extent, relevance. However, there were apparent inconsistencies, and the reliability and readability be considered inadequate. Conclusions: ChatGPT shows promise as an initial source of information for CC and BC. ChatGPT is also highly functional and appears to be superior to physicians, and aligns with expert consensus, although there is room for improvement in readability, reliability, and consistency. Future efforts should focus on developing advanced ChatGPT models explicitly designed to improve medical practice and for those with concerns about symptoms. [ABSTRACT FROM AUTHOR]

Background: Large language models (LLMs) such as ChatGPT-4 and Google Gemini show potential for patient health education, but concerns about their accuracy require careful evaluation. This study evaluates the readability and accuracy of ChatGPT-4 and Google Gemini in answering questions about retinal detachment. Methods: Comparative study analyzing responses from ChatGPT-4 and Google Gemini to 13 retinal detachment questions, categorized by difficulty levels (D1, D2, D3). Masked responses were reviewed by ten vitreoretinal specialists and rated on correctness, errors, thematic accuracy, coherence, and overall quality grading. Analysis included Flesch Readability Ease Score, word and sentence counts. Results: Both Artificial Intelligence tools required college-level understanding for all difficulty levels. Google Gemini was easier to understand (p = 0.03), while ChatGPT-4 provided more correct answers for the more difficult questions (p = 0.0005) with fewer serious errors. ChatGPT-4 scored highest on most challenging questions, showing superior thematic accuracy (p = 0.003). ChatGPT-4 outperformed Google Gemini in 8 of 13 questions, with higher overall quality grades in the easiest (p = 0.03) and hardest levels (p = 0.0002), showing a lower grade as question difficulty increased. Conclusions: ChatGPT-4 and Google Gemini effectively address queries about retinal detachment, offering mostly accurate answers with few critical errors, though patients require higher education for comprehension. The implementation of AI tools may contribute to improving medical care by providing accurate and relevant healthcare information quickly. [ABSTRACT FROM AUTHOR]

Background: This study aimed to answer the research question: How reliable is ChatGPT in automated essay scoring (AES) for oral and maxillofacial surgery (OMS) examinations for dental undergraduate students compared to human assessors? Methods: Sixty-nine undergraduate dental students participated in a closed-book examination comprising two essays at the National University of Singapore. Using pre-created assessment rubrics, three assessors independently performed manual essay scoring, while one separate assessor performed AES using ChatGPT (GPT-4). Data analyses were performed using the intraclass correlation coefficient and Cronbach's α to evaluate the reliability and inter-rater agreement of the test scores among all assessors. The mean scores of manual versus automated scoring were evaluated for similarity and correlations. Results: A strong correlation was observed for Question 1 (r = 0.752–0.848, p < 0.001) and a moderate correlation was observed between AES and all manual scorers for Question 2 (r = 0.527–0.571, p < 0.001). Intraclass correlation coefficients of 0.794–0.858 indicated excellent inter-rater agreement, and Cronbach's α of 0.881–0.932 indicated high reliability. For Question 1, the mean AES scores were similar to those for manual scoring (p > 0.05), and there was a strong correlation between AES and manual scores (r = 0.829, p < 0.001). For Question 2, AES scores were significantly lower than manual scores (p < 0.001), and there was a moderate correlation between AES and manual scores (r = 0.599, p < 0.001). Conclusion: This study shows the potential of ChatGPT for essay marking. However, an appropriate rubric design is essential for optimal reliability. With further validation, the ChatGPT has the potential to aid students in self-assessment or large-scale marking automated processes. [ABSTRACT FROM AUTHOR]

Introduction: In light of the growing interest in utilizing AI for information retrieval, assessing the accuracy and reliability of tools such as chatbots is essential. This study aims to evaluate the efficacy of chatbots in providing accurate information about mobile applications (apps) in the field of audiology. Material and methods: The responses of the Gemini and ChatGPT chatbots to eight open-ended questions posed in Polish and English were compared. Each answer was assessed for correctness. Results: Gemini_ENG achieved the highest correctness with a score of 5 points (62.5%), while ChatGPT_PL scored 2 points (25%), and both Gemini_PL and ChatGPT_ENG scored 1 point (12.5%). Chatbots were most effective in recommending apps for older adults, with three of the four chatbots providing accurate recommendations. However, they struggled when asked to recommend apps for non-English speakers, to describe apps, or to provide direct links, with none of them scoring points in these areas. Conclusions: Chatbots are currently unreliable sources of information about audiology apps. Depending on the language, there is significant variability in response accuracy. A good example is that Gemini_ENG performed far better than Gemini_PL. A major issue for all of them was the frequent fabrication of data, including the creation of nonexistent app names and incorrect links. [ABSTRACT FROM AUTHOR]

Without writing a single line of code by a human, an example Monte Carlo simulation-based application for stochastic dependence modeling with copulas is developed through pair programming involving a human partner and a large language model (LLM) fine-tuned for conversations. This process encompasses interacting with ChatGPT using both natural language and mathematical formalism. Under the careful supervision of a human expert, this interaction facilitated the creation of functioning code in MATLAB, Python, and R. The code performs a variety of tasks including sampling from a given copula model, evaluating the model's density, conducting maximum likelihood estimation, optimizing for parallel computing on CPUs and GPUs, and visualizing the computed results. In contrast to other emerging studies that assess the accuracy of LLMs like ChatGPT on tasks from a selected area, this work rather investigates ways how to achieve a successful solution of a standard statistical task in a collaboration of a human expert and artificial intelligence (AI). Particularly, through careful prompt engineering, we separate successful solutions generated by ChatGPT from unsuccessful ones, resulting in a comprehensive list of related pros and cons. It is demonstrated that if the typical pitfalls are avoided, we can substantially benefit from collaborating with an AI partner. For example, we show that if ChatGPT is not able to provide a correct solution due to a lack of or incorrect knowledge, the human-expert can feed it with the correct knowledge, e.g., in the form of mathematical theorems and formulas, and make it to apply the gained knowledge in order to provide a correct solution. Such ability presents an attractive opportunity to achieve a programmed solution even for users with rather limited knowledge of programming techniques. [ABSTRACT FROM AUTHOR]

Computer systems and applications generate large amounts of logs to measure and record information, which is vital to protect the systems from malicious attacks and useful for repairing faults, especially with the rapid development of distributed computing. Among various logs, the anomaly log is beneficial for operations and maintenance (O&M) personnel to locate faults and improve efficiency. In this paper, we utilize a large language model, ChatGPT, for the log parser task. We choose the BERT model, a self-supervised framework for log anomaly detection. BERT, an embedded transformer encoder, with a self-attention mechanism can better handle context-dependent tasks such as anomaly log detection. Meanwhile, it is based on the masked language model task and next sentence prediction task in the pretraining period to capture the normal log sequence pattern. The experimental results on two log datasets show that the BERT model combined with an LLM performed better than other classical models such as Deelog and Loganomaly. [ABSTRACT FROM AUTHOR]

The recent rise in telemedicine, notably during the COVID-19 pandemic, highlights the potential of integrating artificial intelligence tools in healthcare. This study assessed the effectiveness of ChatGPT versus medical oncologists in the telemedicine-based management of metastatic prostate cancer. In this retrospective study, 102 patients who met inclusion criteria were analyzed to compare the competencies of ChatGPT and oncologists in telemedicine consultations. ChatGPT's role in pre-charting and determining the need for in-person consultations was evaluated. The primary outcome was the concordance between ChatGPT and oncologists in treatment decisions. Results showed a moderate concordance (Cohen's Kappa = 0.43, p < 0.001). The number of diagnoses made by both parties was not significantly different (median number of diagnoses: 5 vs. 5, p = 0.12). In conclusion, ChatGPT exhibited moderate agreement with oncologists in management via telemedicine, indicating the need for further research to explore its healthcare applications. [ABSTRACT FROM AUTHOR]

This study conducts a comparative analysis of the quality of English-to-Chinese (E-C) and Chinese-to-English (C-E) translation of legal texts between Chat Generative Pre-trained Transformer (ChatGPT) and four online Neural Machine Translation (NMT) systems. The analysis includes both quantitative and qualitative evaluations. The results suggest that both ChatGPT and the NMT systems achieve satisfactory performance in translating legal texts from Chinese to English. Although the quality of ChatGPT’s C-E legal translation is slightly lower than that of the NMT systems, the difference is not statistically significant. However, neither ChatGPT nor the NMT systems meet a passing standard for E-C translation of legal texts, with the NMT systems showing better overall performance. Overall, ChatGPT and the NMT systems perform better at translating legal texts from Chinese to English compared to E-C translation. For E-C legal translation, ChatGPT’s quality is lower compared to the NMT systems. While the types of errors are similar in both systems, ChatGPT tends to exhibit more errors, some of which are more severe. This study serves as a reference for those choosing translation tools for E-C and C-E legal texts. [ABSTRACT FROM AUTHOR]

Numerous techniques, including problem-solving, seeking clarification, and creating questions, have been employed to utilize generative Artificial Intelligence (AI) in education. This study investigates the possibility of using Generate AI (GenAI) to grade Structured Query Language (SQL) queries automatically. Three models were used which are ChatGPT, Gemini, and Copilot. The study uses an experimental approach to assess how well the models perform in evaluating student responses by comparing the models' accuracy with those of human experts. The results showed that despite some inconsistencies, GenAI holds great promise for streamlining. Thus, further research is required in light of inconsistent GenAI performance. If these issues were resolved, GenAI can be utilized in education. However, human oversight and ethical issues must always come first. [ABSTRACT FROM AUTHOR]

This paper presents the architecture of a multimodal human–robot interaction control platform that leverages the advanced language capabilities of ChatGPT to facilitate more natural and engaging conversations between humans and robots. Implemented on the Pepper humanoid robot, the platform aims to enhance communication by providing a richer and more intuitive interface. The motivation behind this study is to enhance robot performance in human interaction through cutting-edge natural language processing technology, thereby improving public attitudes toward robots, fostering the development and application of robotic technology, and reducing the negative attitudes often associated with human–robot interactions. To validate the system, we conducted experiments measuring negative attitude robot scale and their robot anxiety scale scores before and after interacting with the robot. Statistical analysis of the data revealed a significant improvement in the participants' attitudes and a notable reduction in anxiety following the interaction, indicating that the system holds promise for fostering more positive human–robot relationships. [ABSTRACT FROM AUTHOR]

With the rapid advancement of Internet technology and the increasing volume of police reports, relying solely on extensive human labor and traditional natural language processing methods for key element extraction has become impractical. Applying advanced technologies such as large language models to improve the effectiveness of police report extraction has become an inevitable trend in the field of police data analysis. This study addresses the characteristics of Chinese police reports and the need to extract key elements by employing large language models specific to the public security domain for entity extraction. Several lightweight (6/7b) open-source large language models were tested as base models. To enhance model performance, LoRA fine-tuning was employed, combined with data engineering approaches. A zero-shot data augmentation method based on ChatGPT and prompt engineering techniques tailored for police reports were proposed to further improve model performance. The key police report data from a certain city in 2019 were used as a sample for testing. Compared to the base models, prompt engineering improved the F1 score by approximately 3%, while fine-tuning led to an increase of 10–50% in the F1 score. After fine-tuning and comparing different base models, the Baichuan model demonstrated the best overall performance in extracting key elements from police reports. Using the data augmentation method to double the data size resulted in an additional 4% increase in the F1 score, achieving optimal model performance. Compared to the fine-tuned universal information extraction (UIE) large language model, the police report entity extraction model constructed in this study improved the F1 score for each element by approximately 5%, with a 42% improvement in the F1 score for the "organization" element. Finally, ChatGPT was employed to align the extracted entities, resulting in a high-quality entity extraction outcome. [ABSTRACT FROM AUTHOR]

The exponential growth of Knowledge Graphs necessitates effective and efficient methods for their exploration and understanding. Frequently Asked Questions (FAQ) is a service that typically presents a list of questions and answers related to a specific topic, and which is intended to help people understand that topic. Although FAQ has already shown its value on large websites and is widely used, to the best of our knowledge it has not yet been exploited for Knowledge Graphs. In this paper, we present ULYSSES, the first system for automatically constructing FAQ lists for large Knowledge Graphs. Our method consists of three key steps. First, we select the most frequent queries by exploiting the available query logs. Next, we answer the selected queries, using the original graph. Finally, we construct textual descriptions of both the queries and the corresponding answers, exploring state-of-the-art transformer models, i.e., ChatGPT 3.5 and Gemini 1.5 Pro. We evaluate the results of each model, using a human-constructed FAQ list, contributing a unique dataset to the domain and showing the benefits of our approach. [ABSTRACT FROM AUTHOR]

Large Language Models (LLMs), based on their highly developed ability to comprehend and generate human-like text, promise to revolutionize all aspects of society. These LLMs facilitate complex language understanding, translation, content generation, and problem-solving, enabled by vast historical data processing and fine-tuning. Throughout the past year, with the initial release of ChatGPT to the public, many papers have appeared on how to exploit LLMs for the ways we operate and interact with intelligent transportation systems. In this study, we review more than 130 papers on the subject and group them according to their major contributions into the following five categories: autonomous driving, safety, tourism, traffic, and others. Based on the aggregated proposals and findings in the extant literature, this paper concludes with a set of challenges and research recommendations, hopefully contributing to guide research in this young, yet extremely active research domain. [ABSTRACT FROM AUTHOR]

Despite significant advancements in Artificial Intelligence (AI) and Large Language Models (LLMs), detecting and mitigating bias remains a critical challenge, particularly on social media platforms like X (formerly Twitter), to address the prevalent cyberbullying on these platforms. This research investigates the effectiveness of leading LLMs in generating synthetic biased and cyberbullying data and evaluates the proficiency of transformer AI models in detecting bias and cyberbullying within both authentic and synthetic contexts. The study involves semantic analysis and feature engineering on a dataset of over 48,000 sentences related to cyberbullying collected from Twitter (before it became X). Utilizing state-of-the-art LLMs and AI tools such as ChatGPT-4, Pi AI, Claude 3 Opus, and Gemini-1.5, synthetic biased, cyberbullying, and neutral data were generated to deepen the understanding of bias in human-generated data. AI models including DeBERTa, Longformer, BigBird, HateBERT, MobileBERT, DistilBERT, BERT, RoBERTa, ELECTRA, and XLNet were initially trained to classify Twitter cyberbullying data and subsequently fine-tuned, optimized, and experimentally quantized. This study focuses on intersectional cyberbullying and multilabel classification to detect both bias and cyberbullying. Additionally, it proposes two prototype applications: one that detects cyberbullying using an intersectional approach and the innovative CyberBulliedBiasedBot that combines the generation and detection of biased and cyberbullying content. [ABSTRACT FROM AUTHOR]

Large language models, e.g., Chat Generative Pre-Trained Transformer (also known as ChatGPT), are currently contributing enormously to making artificial intelligence even more popular, especially among the general population. However, such chatbot models were developed as tools to support natural language communication between humans. Problematically, it is very much a "statistical correlation machine" (correlation instead of causality), and there are indeed ethical concerns associated with the use of AI language models including ChatGPT, such as bias, privacy, and abuse. This paper highlights specific ethical concerns about ChatGPT and articulates key challenges when ChatGPT is used in various applications. Practical recommendations for different stakeholders of ChatGPT are also proposed that can serve as checklist guidelines for those applying ChatGPT in their applications. These best practice examples are expected to motivate the ethical use of ChatGPT. [ABSTRACT FROM AUTHOR]

Since ChatGPT exploded onto the scene in October 2022, the media has been breathless with their discussion about how it is going to impact many aspects of life and work. In order to bring this new reality of generative AI into the Introduction to Management Information Systems classroom at both the undergraduate and graduate levels, we have developed two short cases that can be completed in or out of the classroom. These cases allow students to use generative AI for an interesting and useful purpose and develop prompts to deliver the output. Students can begin to learn what generative AI can do for them, and how it can shape their future careers. We tested the perceived effectiveness of the project on students using a pre/post survey. The results are analyzed using a paired t-test and demonstrate that students increased their understanding of generative AI and prompt engineering. [ABSTRACT FROM AUTHOR]

This study explores the polarization of news content shared on Facebook compared with email using data from the New York Times' Most Emailed and Most Shared lists over 2.5 years. Employing latent Dirichlet allocation and large language models (LLMs), we find that highly polarized articles are more likely to be shared on Facebook (versus email), even after accounting for factors like topics, emotion, and article age. Additionally, distinct topic preferences emerge, with social issues dominating Facebook shares and lifestyle topics prevalent in emails. Contrary to expectations, political polarization of articles shared on Facebook did not escalate post-2020 election. We introduce a novel approach to measuring polarization of text content that leverages generative artificial intelligence models, like ChatGPT, and it is both scalable and cost effective. This research contributes to the evolving intersection of LLMs, social media, and polarization studies, shedding light on descriptive patterns of content dissemination across different digital channels. This paper was accepted by Duncan Simester, marketing. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2023.04134. [ABSTRACT FROM AUTHOR]

This paper analyses the phenomenology and epistemology of chatbots such as ChatGPT and Bard. The computational architecture underpinning these chatbots are large language models (LLMs), which are generative artificial intelligence (AI) systems trained on a massive dataset of text extracted from the Web. We conceptualise these LLMs as multifunctional computational cognitive artifacts, used for various cognitive tasks such as translating, summarizing, answering questions, information-seeking, and much more. Phenomenologically, LLMs can be experienced as a "quasi-other"; when that happens, users anthropomorphise them. For most users, current LLMs are black boxes, i.e., for the most part, they lack data transparency and algorithmic transparency. They can, however, be phenomenologically and informationally transparent, in which case there is an interactional flow. Anthropomorphising and interactional flow can, in some users, create an attitude of (unwarranted) trust towards the output LLMs generate. We conclude this paper by drawing on the epistemology of trust and testimony to examine the epistemic implications of these dimensions. Whilst LLMs generally generate accurate responses, we observe two epistemic pitfalls. Ideally, users should be able to match the level of trust that they place in LLMs to the degree that LLMs are trustworthy. However, both their data and algorithmic opacity and their phenomenological and informational transparency can make it difficult for users to calibrate their trust correctly. The effects of these limitations are twofold: users may adopt unwarranted attitudes of trust towards the outputs of LLMs (which is particularly problematic when LLMs hallucinate), and the trustworthiness of LLMs may be undermined. [ABSTRACT FROM AUTHOR]

How good a research scientist is ChatGPT? We systematically probed the capabilities of GPT-3.5 and GPT-4 across four central components of the scientific process: as a Research Librarian, Research Ethicist, Data Generator, and Novel Data Predictor, using psychological science as a testing field. In Study 1 (Research Librarian), unlike human researchers, GPT-3.5 and GPT-4 hallucinated, authoritatively generating fictional references 36.0% and 5.4% of the time, respectively, although GPT-4 exhibited an evolving capacity to acknowledge its fictions. In Study 2 (Research Ethicist), GPT-4 (though not GPT-3.5) proved capable of detecting violations like p-hacking in fictional research protocols, correcting 88.6% of blatantly presented issues, and 72.6% of subtly presented issues. In Study 3 (Data Generator), both models consistently replicated patterns of cultural bias previously discovered in large language corpora, indicating that ChatGPT can simulate known results, an antecedent to usefulness for both data generation and skills like hypothesis generation. Contrastingly, in Study 4 (Novel Data Predictor), neither model was successful at predicting new results absent in their training data, and neither appeared to leverage substantially new information when predicting more vs. less novel outcomes. Together, these results suggest that GPT is a flawed but rapidly improving librarian, a decent research ethicist already, capable of data generation in simple domains with known characteristics but poor at predicting novel patterns of empirical data to aid future experimentation. [ABSTRACT FROM AUTHOR]