With the development of generative artificial intelligence and instruction tuning techniques, multimodal large language models (MLLMs) have made impressive progress on general reasoning tasks. Benefiting from the chain-of-thought (CoT) methodology, MLLMs can solve the visual reasoning problem step-by-step. However, existing MLLMs still face significant challenges when applied to pathology visual reasoning tasks: (1) LLMs often underperforms because they lack domain-specific information, which can lead to model hallucinations. (2) The additional reasoning steps in CoT may introduce errors, leading to the divergence of answers. To address these limitations, we propose PathCoT, a novel zero-shot CoT prompting method which integrates the pathology expert-knowledge into the reasoning process of MLLMs and incorporates self-evaluation to mitigate divergence of answers. Specifically, PathCoT guides the MLLM with prior knowledge to perform as pathology experts, and provides comprehensive analysis of the image with their domain-specific knowledge. By incorporating the experts' knowledge, PathCoT can obtain the answers with CoT reasoning. Furthermore, PathCoT incorporates a self-evaluation step that assesses both the results generated directly by MLLMs and those derived through CoT, finally determining the reliable answer. The experimental results on the PathMMU dataset demonstrate the effectiveness of our method on pathology visual understanding and reasoning.

Artificial intelligence holds great promise for expanding access to expert medical knowledge and reasoning. However, most evaluations of language models rely on static vignettes and multiple-choice questions that fail to reflect the complexity and nuance of evidence-based medicine in real-world settings. In clinical practice, physicians iteratively formulate and revise diagnostic hypotheses, adapting each subsequent question and test to what they've just learned, and weigh the evolving evidence before committing to a final diagnosis. To emulate this iterative process, we introduce the Sequential Diagnosis Benchmark, which transforms 304 diagnostically challenging New England Journal of Medicine clinicopathological conference (NEJM-CPC) cases into stepwise diagnostic encounters. A physician or AI begins with a short case abstract and must iteratively request additional details from a gatekeeper model that reveals findings only when explicitly queried. Performance is assessed not just by diagnostic accuracy but also by the cost of physician visits and tests performed. We also present the MAI Diagnostic Orchestrator (MAI-DxO), a model-agnostic orchestrator that simulates a panel of physicians, proposes likely differential diagnoses and strategically selects high-value, cost-effective tests. When paired with OpenAI's o3 model, MAI-DxO achieves 80% diagnostic accuracy--four times higher than the 20% average of generalist physicians. MAI-DxO also reduces diagnostic costs by 20% compared to physicians, and 70% compared to off-the-shelf o3. When configured for maximum accuracy, MAI-DxO achieves 85.5% accuracy. These performance gains with MAI-DxO generalize across models from the OpenAI, Gemini, Claude, Grok, DeepSeek, and Llama families. We highlight how AI systems, when guided to think iteratively and act judiciously, can advance diagnostic precision and cost-effectiveness in clinical care.

Trauma prevalence is vast globally. Evidence-based digital treatments can help, but most require human guidance. Human guides provide tailored instructions and responsiveness to internal cognitive states, but limit scalability. Can generative AI and neurotechnology provide a scalable alternative? Here we test ANTIDOTE, combining AI guidance and pupillometry to automatically deliver and monitor an evidence-based digital treatment, specifically the Imagery Competing Task Intervention (ICTI), to reduce intrusive memories after psychological trauma. One hundred healthy volunteers were exposed to videos of traumatic events and randomly assigned to an intervention or active control condition. As predicted, intervention participants reported significantly fewer intrusive memories over the following week. Post-hoc assessment against clinical rubrics confirmed the AI guide delivered the intervention successfully. Additionally, pupil size tracked intervention engagement and predicted symptom reduction, providing a candidate biomarker of intervention effectiveness. These findings open a path toward rigorous AI-guided digital interventions that can scale to trauma prevalence.

More than 1,000 construction workers die on the job each year in the US, making it the most dangerous industry for fatal slips, trips, and falls. A new AI tool called Safety AI could help to change that. It analyzes the progress made on a construction site each day, and flags conditions that violate Occupational Safety and Health Administration rules, with what its creator Philip Lorenzo claims is 95% accuracy. Lorenzo says Safety AI is the first one of multiple emerging AI construction safety tools to use generative AI to flag safety violations. But as the 95% success rate suggests, Safety AI is not a flawless and all-knowing intelligence.

To combat the shortcuts and risk-taking, Lorenzo is working on a tool for the San Francisco–based company DroneDeploy, which sells software that creates daily digital models of work progress from videos and images, known in the trade as "reality capture." The tool, called Safety AI, analyzes each day's reality capture imagery and flags conditions that violate Occupational Safety and Health Administration (OSHA) rules, with what he claims is 95% accuracy. That means that for any safety risk the software flags, there is 95% certainty that the flag is accurate and relates to a specific OSHA regulation. Launched in October 2024, it's now being deployed on hundreds of construction sites in the US, Lorenzo says, and versions specific to the building regulations in countries including Canada, the UK, South Korea, and Australia have also been deployed. Safety AI is one of multiple AI construction safety tools that have emerged in recent years, from Silicon Valley to Hong Kong to Jerusalem.

Despite the increasing use of large language models (LLMs) in education, concerns have emerged about their potential to reduce deep thinking and active learning. This study investigates the impact of generative artificial intelligence (AI) tools, specifically ChatGPT, on the cognitive engagement of students during academic writing tasks. The study employed an experimental design with participants randomly assigned to either an AI-assisted (ChatGPT) or a non-assisted (control) condition. Participants completed a structured argumentative writing task followed by a cognitive engagement scale (CES), the CES-AI, developed to assess mental effort, attention, deep processing, and strategic thinking. The results revealed significantly lower cognitive engagement scores in the ChatGPT group compared to the control group. These findings suggest that AI assistance may lead to cognitive offloading. The study contributes to the growing body of literature on the psychological implications of AI in education and raises important questions about the integration of such tools into academic practice. It calls for pedagogical strategies that promote active, reflective engagement with AI-generated content to avoid compromising self-regulated learning and deep cognitive involvement of students.

The aim of this paper is to review the use of GenAI in scientometrics, and to begin a debate on the broader implications for the field. First, we provide an introduction on GenAI's generative and probabilistic nature as rooted in distributional linguistics. And we relate this to the debate on the extent to which GenAI might be able to mimic human 'reasoning'. Second, we leverage this distinction for a critical engagement with recent experiments using GenAI in scientometrics, including topic labelling, the analysis of citation contexts, predictive applications, scholars' profiling, and research assessment. GenAI shows promise in tasks where language generation dominates, such as labelling, but faces limitations in tasks that require stable semantics, pragmatic reasoning, or structured domain knowledge. However, these results might become quickly outdated. Our recommendation is, therefore, to always strive to systematically compare the performance of different GenAI models for specific tasks. Third, we inquire whether, by generating large amounts of scientific language, GenAI might have a fundamental impact on our field by affecting textual characteristics used to measure science, such as authors, words, and references. We argue that careful empirical work and theoretical reflection will be essential to remain capable of interpreting the evolving patterns of knowledge production.

Generative AI systems have revolutionized human interaction by enabling natural language-based coding and problem solving. However, the inherent ambiguity of natural language often leads to imprecise instructions, forcing users to iteratively test, correct, and resubmit their prompts. We propose an iterative approach that systematically narrows down these ambiguities through a structured series of clarification questions and alternative solution proposals, illustrated with input/output examples as well. Once every uncertainty is resolved, a final, precise solution is generated. Evaluated on a diverse dataset spanning coding, data analysis, and creative writing, our method demonstrates superior accuracy, competitive resolution times, and higher user satisfaction compared to conventional one-shot solutions, which typically require multiple manual iterations to achieve a correct output.

Computer programming is undergoing a true transformation driven by powerful new tools for automatic source code generation based on large language models. This transformation is also manifesting in introductory programming courses at universities around the world, generating an in-depth debate about how programming content should be taught, learned, and assessed in the context of generative artificial intelligence. This article aims, on the one hand, to review the most relevant studies on this issue, highlighting the advantages and disadvantages identified in the specialized literature. On the other hand, it proposes enriching teaching and learning methodologies by focusing on code comprehension and execution rather than on mere coding or program functionality. In particular, it advocates for the use of visual representations of code and visual simulations of its execution as effective tools for teaching, learning, and assessing programming, thus fostering a deeper understanding among students. Finally, the opinions of students who took the object-oriented programming course are presented to provide preliminary context supporting the incorporation of visual simulations in Java (or other languages) as part of the training process.

As the basis of generative AI, an autoregressive model requires the generation of a new token depending on all the previously generated tokens, which brings high quality but also restricts the model to generate tokens one by one, forming a bottleneck limiting the generation speed. In this paper, we propose a new decoder architecture that efficiently generates text in parallel for context-aware generation tasks. Our proposed pipelined decoder initiates the generation of multiple subsequences simultaneously, and, at each time-step, it generates a new token for each subsequence to realize parallelism. Experiments on multiple text generation tasks, including question answering, text summarization, and keyphrase generation, show that our pipelined decoder significantly improves the generation speed without a significant loss of generation quality or additional memory consumption.