Recent advances in large language models (LLMs), such as GPT-4, have spurred interest in their potential applications across various fields, including actuarial work. This paper introduces the use of LLMs in actuarial and insurance-related tasks, both as direct contributors to actuarial modelling and as workflow assistants. It provides an overview of LLM concepts and their potential applications in actuarial science and insurance, examining specific areas where LLMs can be beneficial, including a detailed assessment of the claims process. Additionally, a decision framework for determining the suitability of LLMs for specific tasks is presented. Case studies with accompanying code showcase the potential of LLMs to enhance actuarial work. Overall, the results suggest that LLMs can be valuable tools for actuarial tasks involving natural language processing or structuring unstructured data and as workflow and coding assistants. However, their use in actuarial work also presents challenges, particularly regarding professionalism and ethics, for which high-level guidance is provided.

Scientists have started to explore whether novel artificial intelligence (AI) tools based on large language models, such as GPT-4, could support the scientific peer review process. We sought to understand (i) whether AI versus human reviewers are able to distinguish between made-up AI-generated and human-written conference abstracts reporting on actual research, and (ii) how the quality assessments by AI versus human reviewers of the reported research correspond to each other. We conducted a large-scale field experiment during a medium-sized scientific conference, relying on 305 human-written and 20 AI-written abstracts that were reviewed either by AI or 217 human reviewers. The results show that human reviewers and GPTZero were better in discerning (AI vs. human) authorship than GPT-4. Regarding quality assessments, there was rather low agreement between both human–human and human–AI reviewer pairs, but AI reviewers were more aligned with human reviewers in classifying the very best abstracts. This indicates that AI could become a prescreening tool for scientific abstracts. The results are discussed with regard to the future development and use of AI tools during the scientific peer review process.

This article contributes to the empirical and theoretical discourse on the ‘stability–instability paradox’, the idea that while possessing nuclear weapons deters cataclysmic all-out war, it simultaneously increases the likelihood of low-level conflict between nuclear dyads. It critiques the paradox’s dominant interpretation (red-line model), which places undue confidence in the nuclear stalemate – premised on mutually assured destruction – to prevent unintentional nuclear engagement and reduce the perceived risks associated with military actions that fall below the nuclear threshold. Recent scholarship has inadequately examined the unintentional consequences of the paradox in conflicts below the nuclear threshold, particularly those relating to the potential for aggression to escalate uncontrollably. The article employs empirically grounded fictional scenarios to illustrate and critically evaluate, rather than predict, the assumptions underpinning the red-line model of the stability–instability paradox in the context of future artificial intelligence (AI)-enabled warfare. It posits that the strategic cap purportedly offered by a nuclear stalemate is illusory and that low-level military aggression between nuclear-armed states increases the risk of unintentional nuclear detonation.

In an era where artificial intelligence (AI) permeates every facet of our lives, the imperative to steer AI development toward enhancing human wellbeing has never been more critical. However, the development of such positive AI poses substantial challenges due to the current lack of mature methods for addressing the complexities that designing AI for wellbeing poses. This article presents and evaluates the positive AI design method aimed at addressing this gap. The method provides a human-centered process for translating wellbeing aspirations into concrete interventions. First, we explain the method’s key steps: (1) contextualizing, (2) operationalizing, (3) designing, and (4) implementing supported by (5) continuous measurement for iterative feedback cycles. We then present a multi-case study where novice designers applied the method, revealing strengths and weaknesses related to efficacy and usability. Next, an expert evaluation study assessed the quality of the case studies’ outcomes, rating them moderately high for feasibility, desirability, and plausibility of achieving intended wellbeing benefits. Together, these studies provide preliminary validation of the method’s ability to improve AI design, while identifying opportunities for enhancement. Building on these insights, we propose adaptations for future iterations of the method, such as the inclusion of wellbeing-related heuristics, suggesting promising avenues for future work. This human-centered approach shows promise for realizing a vision of “AI for wellbeing” that does not just avoid harm, but actively promotes human flourishing.

Multimodal imaging is crucial for diagnosis and treatment in paediatric cardiology. However, the proficiency of artificial intelligence chatbots, like ChatGPT-4, in interpreting these images has not been assessed. This cross-sectional study evaluates the precision of ChatGPT-4 in interpreting multimodal images for paediatric cardiology knowledge assessment, including echocardiograms, angiograms, X-rays, and electrocardiograms. One hundred multiple-choice questions with accompanying images from the textbook Pediatric Cardiology Board Review were randomly selected. The chatbot was prompted to answer these questions with and without the accompanying images. Statistical analysis was done using X2, Fisher’s exact, and McNemar tests. Results showed that ChatGPT-4 answered 41% of questions with images correctly, performing best on those with electrocardiograms (54%) and worst on those with angiograms (29%). Without the images, ChatGPT-4’s performance was similar at 37% (difference = 4%, 95% confidence interval (CI) –9.4% to 17.2%, p = 0.56). The chatbot performed significantly better when provided the image of an electrocardiogram than without (difference = 18, 95% CI 4.0% to 31.9%, p < 0.04). In cases of incorrect answers, ChatGPT-4 was more inconsistent with an image than without (difference = 21%, 95% CI 3.5% to 36.9%, p < 0.02). In conclusion, ChatGPT-4 performed poorly in answering image-based multiple-choice questions in paediatric cardiology. Its accuracy in answering questions with images was similar to without, indicating limited multimodal image interpretation capabilities. Substantial training is required before clinical integration can be considered. Further research is needed to assess the clinical reasoning skills and progression of ChatGPT in paediatric cardiology for clinical and academic utility.

The Artificial Intelligence Act (AI Act) of the European Union (EU) claims to be based on a risk-based approach to avoid over-regulation and to respect the principle of legislative proportionality. This paper argues that risk-based regulation is indeed the right approach to AI regulation. At the same time, however, the paper shows that important provisions of the AI Act do not follow a truly risk-based approach. Yet, this is nothing that cannot be fixed. The AI Act provides for sufficient tools to support future-proof legislation and to implement it in line with a genuine risk-based approach. Against this background, the paper analyses how the AI Act should be applied and implemented according to its original intention of a risk-based approach, and what lessons legislators around the world can learn from the AI Act in regulating AI.

Since its inception, Health Technology Assessment (HTA) has typically determined the value of a technology by collecting information derived from randomized clinical trials (RCTs), in line with the principles of evidence-based medicine (EBM). However, data from RCTs did not constitute the sole source of information, as other types of evidence (such as primary qualitative research) have often been utilized. Recent advances in both generating and collecting other types of evidence are broadening the landscape of evidence, adding complexity to the discussion of “robustness of evidence.” What are the consequences of these recent developments for the methodology and conduct of HTA, the HTA community, and its ethical commitments? The aim of this article is to explore some ethical challenges that are emerging in the current evolving evidence landscape, particularly changes in evidence generation and collection (e.g., diversification of data sources), and shifting standards of evidence in the field of HTA (e.g., increasing acceptability of evidence that is thought of as lower quality). Our conclusion is that deciding how to best maintain trustworthiness is common to all these issues.

The analysis of liability aspects facing Artificial Intelligence (‘AI’)-generated outputs under copyright and related rights has been overlooked compared to other issues connected to the development and use of AI. This study fills this gap by exploring pertinent questions under international, EU and UK law. Specifically, the study tackles actionable reproduction, allocation of liability, and availability of defences. The analysis ultimately shows that, while it is clear that each case will need to be decided on its own merits, the generative AI output phase raises several profiles of liability under copyright law. If the goal of policymakers and relevant stakeholders is to ensure the balanced and sustainable development of AI, then the issues related to the generation and dissemination of AI outputs need to be given ample attention and a greater role in the debate than what has been the case so far, whether it is in the context of risk assessment and compliance, licensing initiatives, or in contentious scenarios.

The robots of tomorrow should be endowed with the ability to adapt to drastic and unpredicted changes in their environment and interactions with humans. Such adaptations, however, cannot be boundless: the robot must stay trustworthy. So, the adaptations should not be just a recovery into a degraded functionality. Instead, they must be true adaptations: the robot must change its behaviour while maintaining or even increasing its expected performance and staying at least as safe and robust as before. The RoboSAPIENS project will focus on autonomous robotic software adaptations and will lay the foundations for ensuring that they are carried out in an intrinsically trustworthy, safe and efficient manner, thereby reconciling open-ended self-adaptation with safety by design. RoboSAPIENS will transform these foundations into ‘first time right’-design tools and platforms and will validate and demonstrate them.

This paper aims at exploring the dynamic interplay between advanced technological developments in AI and Big Data and the sustained relevance of theoretical frameworks in scientific inquiry. It questions whether the abundance of data in the AI era reduces the necessity for theory or, conversely, enhances its importance. Arguing for a synergistic approach, the paper emphasizes the need for integrating computational capabilities with theoretical insight to uncover deeper truths within extensive datasets. The discussion extends into computational social science, where elements from sociology, psychology, and economics converge. The application of these interdisciplinary theories in the context of AI is critically examined, highlighting the need for methodological diversity and addressing the ethical implications of AI-driven research. The paper concludes by identifying future trends and challenges in AI and computational social science, offering a call to action for the scientific community, policymakers, and society. Being positioned at the intersection of AI, data science, and social theory, this paper illuminates the complexities of our digital era and inspires a re-evaluation of the methodologies and ethics guiding our pursuit of knowledge.

During the past 30 yr an impasse has developed in the discovery and commercialization of synthetic herbicides with new molecular targets and novel chemistries. Similarly, there has been little success with bioherbicides, both microbial and chemical. These bioherbicides are needed to combat fast-growing herbicide resistance and to fulfill the need for more environmentally and toxicologically safe herbicides. In response to this substantial and growing opportunity, numerous start-up companies are utilizing novel approaches to provide new tools for weed management. These diverse new tools broaden the scope of discovery, encompassing advanced computational, bioinformatic, and imaging platforms; plant genome–editing and targeted protein degradation technologies; and machine learning and artificial intelligence (AI)-based strategies. This review contains summaries of the presentations of 10 such companies that took part in a symposium held at the WSSA annual meeting in 2024. Four of the companies are developing microbial bioherbicides or natural product–based herbicides, and the other six are using advanced technologies, such as AI, to accelerate the discovery of herbicides with novel molecular target sites or to develop non-GMO, herbicide-resistant crops.

With technological developments presenting tremendous opportunities, rapid developments in data-driven artificial intelligence (AI) and machine learning (ML) have the potential to significantly transform humanitarian forensic action. Yet, their role in the forensic identification of dead bodies remains unexamined. The correct and early identification of dead bodies is not only important to afford the deceased their honour and dignity and to ensure that their families know the fate of their loved ones, but also has broader implications for human rights and international humanitarian law (IHL). This article examines the opportunities and challenges of AI and ML in advancing honour and dignity in death for armed conflict victims in humanitarian forensic action under IHL. It argues that the application of AI and ML in humanitarian forensic action has the potential to revolutionize and support forensic practitioners in the identification of human remains. This will consequently guarantee that families know the fate of their loved ones and that the deceased are afforded dignified burials according to their religious and cultural rites. The article proposes recommendations for the future use of AI and ML in humanitarian forensic action, including the adoption of a legally binding international instrument governing their use, the development of guidelines for their use, the training of forensic actors in IHL and human rights law, and the use of new technologies in humanitarian action.

We developed and trained a custom AI text classifier to assist Latin language learners in internalising idioms, with a focus on Ciceronian composition. Idioms are mostly language-, sometimes author-specific expressions that cannot be translated directly into another language. The classifier provides interactive feedback on learners' output, effectively gamifying the process of writing in Latin and motivating students to produce more linguistic output. This not only helps learners build confidence, but also acquire a ‘toolbox’ of idioms useful for both output- and input-based activities. We present the results of our experiment conducted across a number of secondary classes in Switzerland. Our tool is freely available at https://latin-ia.hepl.ch/classifier.