Articles tagged with Artificial Intelligence
Security researchers have developed a defense mechanism that protects against cryptanalytic parameter extraction attacks on AI systems. By making neurons in the same layer similar to each other, the defense creates a barrier of similarity that makes it difficult for attacks to proceed.
The book highlights how AI tools enhance crop monitoring, pest control, and water conservation through data-driven insights and predictive modeling. It offers practical guidance for researchers, agricultural professionals, and technologists on applying AI responsibly in rural and emerging economies.
A new study reveals AI can corrupt public opinion surveys at scale, passing quality checks and flipping predicted outcomes with fake responses. The findings have significant implications for scientific research and policy-making, highlighting the need for transparency and verification of human participation.
The European Centre for Medium-Range Weather Forecasts (ECMWF) and several National Meteorological Services across Europe have been honored with the 2025 HPCwire Readers’ and Editors’ Choice Award for
Researchers at RIKEN successfully simulated the Milky Way Galaxy with over 100 billion individual stars, far surpassing previous state-of-the-art models. This achievement demonstrates the power of AI-accelerated simulations in tackling complex multi-scale problems in astrophysics and beyond.
Generative Artificial Intelligence could lead to a renewed emphasis on conversational approaches to teaching, researchers argue. This involves teachers and students exploring problems together, discussing ideas from different angles, and testing concepts through collaborative work.
This book offers a comprehensive exploration of AI-driven analytics in finance, addressing market prediction, fraud detection, and risk assessment. It also discusses AI applications in healthcare and cybersecurity, including disease classification and biometric identification systems.
Researchers at Terasaki Institute and Caltech will use stem cell-based models to identify factors influencing early human development. The goal is to gain insights into infertility, pregnancy loss, and developmental disorders.
Artificial intelligence is being used to improve diagnostic accuracy in molecular pathology, with studies showing 93% accuracy for cancer diagnosis. Researchers are also using AI to analyze chromosomal changes in blood cancer patients and provide personalized treatment plans.
A University of Delaware team, funded by NSF, creates next-generation secure computer chips using AI and game theory. The project aims to predict attacks and set new benchmarks for global hardware security, enabling the design of safer and more trustworthy systems.
Researchers used machine learning to uncover structures in finite simple groups, discovering a new theorem on necessary properties of generators. The 2-generator representation shows simplicity has interesting data structure, shedding light on the role of artificial intelligence in pure math.
The American Psychological Association's new health advisory emphasizes that generative AI chatbots and wellness applications lack scientific evidence to ensure user safety. Despite their low cost and accessibility, these tools are not a substitute for systemic solutions to the mental health crisis.
Researchers at UCSF used AI to analyze clinical notes from multiple providers to improve HRS diagnosis accuracy. The technology offers a unified summary of the care team's consensus, aligning care decisions and expediting treatment plans.
A new study found that while parents are willing to accept AI-generated images in children's stories with certain conditions, they raise concerns about errors in illustrations that might encourage unsafe behavior or lack real-world accuracy. Parents prefer clear notification on the cover of the story if AI was used to create images.
The University of Maine has launched internships in AI and digital twins to prepare students for careers in the growing blue economy. Students will work with real-time data, build virtual replicas of ocean structures, and test complex marine scenarios.
Dr. Christina Richey joins the SETI Institute with extensive experience in research management, strategic collaboration, and science leadership. She will focus on creating sustainable partnerships and funding opportunities to support the institute's long-term growth and mission.
The new statistical method adapts to data structure, resisting outliers and providing greater stability on non-Euclidean spaces. This improves the reliability of analysis in areas like medical imaging, computer vision, and machine learning.
A Dartmouth study finds that AI-powered chatbots can deliver personalized learning to large numbers of students. The researchers created an AI teaching assistant called NeuroBot TA that provides around-the-clock individualized support for students, which they found to be more trusted than general chatbots.
Medical imaging foundation models revolutionize cancer precision medicine by integrating multimodal data, AI algorithms, and large language models. They enhance diagnostic accuracy and clinical interpretability, facilitating personalized therapy and intelligent clinical decision support.
Researchers have summarized recent breakthroughs in theranostic nanomaterials, engineered nanoparticles that can both diagnose and treat TBI. These materials can deliver drugs precisely where damage occurs while monitoring biological changes inside the brain.
JMIR Publications has integrated Signals Manuscript Checks into its publication process, utilizing AI-powered tools to streamline research integrity evaluations. This partnership aims to increase efficiency and quality, ensuring high standards of scientific integrity in published research.
A recent FAU Engineering study leverages quantum computing to enhance the accuracy of chronic kidney disease (CKD) diagnosis. The research team developed and compared two automated systems: Classical Support Vector Machine (CSVM) and Quantum Support Vector Machine (QSVM). CSVM achieved remarkable 98.75% accuracy, while QSVM reached 87....
A recent study finds that artificial intelligence has a negligible effect on global greenhouse gas emissions. The researchers' analysis of US energy consumption and AI use across industries revealed that the energy usage from AI in the US equals the energy consumption for all of Iceland, but not noticeable on a global scale.
Researchers from the University of Liège examined the capabilities of Midjourney and DALL·E to produce images from simple sentences, revealing both aesthetic feats and beginner's mistakes. The study highlights the limitations of translation between verbal and visual language.
Large language models systematically rate speakers of German dialects less favorably than those using Standard German, associating dialects with negative traits. The bias grows when dialects are explicitly mentioned, and larger models display even stronger biases.
Researchers develop a new AI approach to simulate the Earth system, enabling better estimates of local impacts on ecosystems and societies. The WOW project aims to couple AI models across scales of space and time, revealing hidden connections in the climate system.
The ESMO Congress on Artificial Intelligence and Digital Oncology features a diverse range of presentations on multimodal AI for clinical decision support, AI-based biomarkers for precision oncology, and digital health in practice. The congress also explores ethics, regulation, and patient perspectives.
A new report from Northwestern University offers six practical strategies to improve the doctor-patient bedside encounter in an era dominated by AI. By employing these strategies, clinicians can strengthen patient-physician relationships, combat inequities, and reduce burnout, ultimately leading to better patient outcomes.
The University of Osaka and NEC have demonstrated a wide-area distributed campus AI processing platform, enabling on-demand use of GPUs from compute servers via an optical network. This platform leverages NEC's ExpEther technology to connect and disconnect devices across campus, improving efficiency and optimizing GPU utilization.
Researchers developed CellWhisperer, an AI method and software tool that links gene expression with descriptive text across millions of biological samples. It provides a virtual AI-based colleague to support biologists in their research, making biomedical data exploration easier and more exciting.
A new study finds that generative AI can help define viable objectives for organizational and policy decision-making, but the overall quality of those objectives falls short without human intervention. The research suggests that AI tools are valuable brainstorming partners, but sound decision analysis still requires a 'human in the loop'.
Omni-modal language models integrate perception, understanding, and reasoning within a unified framework, bringing AI closer to human cognition. Lightweight adaptation strategies improve deployment efficiency in real-time medical and industrial scenarios.
A new study by Penn State researchers uses an AI model to analyze over 400 small-group conversations and finds that discussions improve critical thinking in 4th-5th graders. The approach, based on Quality Talk, enables the analysis of large amounts of qualitative data quickly and efficiently.
The Hong Kong Polytechnic University (PolyU) has achieved a breakthrough in perovskite/silicon tandem solar cells, focusing on improving efficiency, stability and scalability. The team aims to raise the energy conversion efficiency from 34% to 40%, while promoting industry-academia-research collaboration.
Researchers studied how large language models (LLMs) operate and found they use built-in patterns for tracking word positions to form internal beliefs and make social inferences. By understanding these patterns, it may be possible to make LLMs more efficient by activating only a subset of parameters.
A five-minute training can significantly improve people's ability to identify fake faces created by artificial intelligence, according to new research. The study found that super-recognisers correctly identified fake faces 64% of the time, while typical participants scored 51% accuracy.
Researchers developed an AI model that outperforms radiologists in detecting hidden objects on chest scans, particularly for cases of foreign body aspiration. The model achieves high precision and recall rates, helping doctors diagnose complex conditions more reliably.
The study found that AI's rapid growth would annually put 24-44 million metric tons of CO2 into the atmosphere, draining 731-1125 million cubic meters of water. A roadmap suggests using smart siting, faster grid decarbonization and operational efficiency to cut impacts by 73% (carbon dioxide) and 86% (water).
A PhD student at Lehigh University is working with Siemens to develop real-time monitoring and control tools for hyperscale data centers. The goal is to create a localized power network that can operate independently of the main grid, reducing power demands from artificial intelligence and increasing energy efficiency.
Researchers found that large language models exhibit systematic biases when evaluating texts, but only when the source or author is revealed. The studies showed a high level of agreement among models when no information was provided, but decreased agreement and even bias emerged when fictional sources were used.
The study reveals how chemical sensors discriminate among various odorant molecules using explainable AI, enabling the selection of receptor materials for high-performance sensors. This breakthrough advances understanding of human olfaction and paves the way for practical application of artificial olfaction technology.
The inaugural Margot and Tom Pritzker Prize for AI in Science Research Excellence recognizes outstanding contributions to the intersection of artificial intelligence and scientific research. Winners Kyle Cranmer and Deborah Marks were honored for their pioneering work in simulation-based inference and language modeling methods.
The $12.25 million NSF grant will fund the development of an advanced system prototype and tools for automation, remote data control, and scientific workflow management. The project aims to support the establishment of data centers in the region and provide access to these resources to researchers nationwide.
Researchers from Chung-Ang University have developed a novel AI-based approach for producing high-fidelity and defect-aware ultrasonic images, outperforming traditional techniques. This technology has the potential to revolutionize non-destructive testing in industries such as semiconductors, energy, and automotive.
The joint research hub aims to bridge academic research and scalable infrastructure for open AI systems, with a focus on decentralised computing architectures and secure AI marketplaces. Early proof-of-concepts are expected within two years, targeting pilot applications in finance, healthcare, and smart infrastructure.
Recent research in pathological foundation models has demonstrated robust representation learning capabilities and generalization performance. These models enable multi-task transfer with minimal data, significantly enhancing clinical utility and generalizability.
The American College of Cardiology and OpenEvidence are collaborating to integrate AI technology into clinical workflows, enhancing patient care and outcomes. The partnership aims to translate scientific discovery into clinical implementation, providing clinicians with timely and accurate medical evidence.
KVzip reduces chatbot response time and memory cost while maintaining accuracy, achieving 3–4× memory reduction and approximately 2× faster response times. The technology also demonstrates scalability to extremely long contexts and has been integrated into NVIDIA's open-source library.
The Biohub initiative combines world-class compute capacity, frontier AI research and engineering, and state-of-the-art experimental and imaging technologies to accelerate science. The combined team will develop datasets, laboratory technologies, and modeling innovations needed to build next-generation AI systems for biological discovery.
A new study reveals AI can generate highly realistic images of real people that are hard to tell apart from authentic photos. Despite familiarity with the subject's appearance, participants struggled to distinguish between genuine and fake images.
A new AI model uses machine learning to predict drug toxicity in humans by identifying biological differences between cells, mice, and humans. The model improved predictive power over existing state-of-the-art models and demonstrated practicality in predicting market withdrawal due to toxicity.
The ITU delegation will work with governments and private-sector leaders to leverage digital transformation as a catalyst for climate solutions. The Green Digital Action initiative aims to address the sector's environmental impact through practical tools and tangible results.
A new study published in the Chinese Neurosurgical Journal explores an AI tool that identifies medulloblastoma subgroups based on magnetic resonance imaging scans. The model achieved impressive accuracy in predicting molecular subtypes and genetic risk factors, with 91% accuracy for TP53 mutations and 87% accuracy for chromosome 11 loss.
The Mount Sinai Health System will implement Microsoft Dragon Copilot, an AI-powered voice assistant designed to streamline clinical documentation and automate administrative tasks. The rollout aims to reduce burnout, improve patient experiences, and enhance clinical efficiency across the system.
Researchers developed a protein language model that can generate functional antibodies against specific viral antigens. This approach enables rapid generation of antibodies against emerging health threats without requiring blood samples or antigen proteins.
Researchers created an AI algorithm that can identify obstructive sleep apnea (OSA) using ECG test results, making it faster and cheaper to spot the condition. The innovation reveals a higher visibility of OSA on ECG in females compared to males, suggesting they may suffer more cardiovascular damage.
The FAU College of Engineering and Computer Science has established the 'Ubicquia Innovation Center for Intelligent Infrastructure' to develop transformative technologies. The center will empower students and faculty to create AI-First solutions for a smarter, more connected world.
A Michigan State University-led study examines how well AI personas can detect human deception and compares their performance to humans. The results show that AI is more lie-biased and less accurate than humans, highlighting the need for improvement before generative AI can be used for deception detection.
Researchers found that non-expert users' intuitive prompts triggered biased responses as effectively as expert 'jailbreak' techniques. The study used a competition to identify biases in generative AI models and suggests strategies for mitigating issues.
A team of researchers proposes a theoretical framework to induce non-reciprocal interactions in solid-state systems using light. By applying dissipation-engineering scheme, they predict a non-equilibrium phase transition called non-reciprocal phase transition, leading to a spontaneous and continuous rotation of magnetization.