Articles tagged with Artificial Intelligence
Researchers at UCSF and Wayne State University found that generative AI tools can perform orders of magnitude faster than human teams in analyzing health data. Junior researchers paired with AI generated viable prediction models in minutes, outperforming experienced programmers in hours or days.
MIT researchers used a large language model to optimize the genetic sequences of proteins manufactured by yeast, reducing production costs. The new model predicted which codons would work best for manufacturing six different proteins, including human growth hormone and a monoclonal antibody, with successful results.
Research reveals major gaps in studying the relationship between AI and journalism, including environmental impact and Global South deployment. Most studies are conducted from Europe, leaving scarce attention to the Middle East, Latin America, and sub-Saharan Africa.
Despite AI tool limitations, generative AI projects continue due to various justifications that make them hard to stop. The study highlights nine conceptions of the common good that drive innovation momentum. Effective boundary work and reframing failure as a learning process also sustain innovation.
Researchers used satellite imagery and machine learning to estimate HDI scores for 61,530 municipalities and counties. The results show that over half the global population lives in areas with different development tiers than their national level, highlighting the need for finer-grained data to inform policy decisions.
A new study uses AI to predict which ulcerative colitis patients with low-grade dysplasia are at high risk of developing colorectal cancer. The model integrates risk factors from narrative clinical notes, enabling more accurate and timely surveillance and potentially reducing treatment delays.
SeaCast consistently outperforms the Copernicus operational model over a 10-day forecast horizon and extends predictions to 15 days, generating forecasts in just 20 seconds using a single GPU. This advancement enables rapid 'what-if' scenario testing and probabilistic ensemble forecasts.
The journal explores the convergence of computational biology, artificial intelligence, and healthcare innovation, with a focus on precision medicine and enhanced patient care. Submissions are accepted from researchers, clinicians, and technologists on topics such as AI in medicine, computational genomics, and drug discovery.
Researchers developed a new method for extracting spectral information from OCT images and combining it with AI to identify lipid-rich plaques. The approach shows strong classification performance and can highlight suspicious regions throughout the image.
A new theory asserts that associative learning depends less on repetition and more on time passing between rewards, explaining why students who cram don't perform well. The study's findings could shift the way we look at learning and addiction, potentially speeding up artificial intelligence by borrowing from this new approach.
Researchers developed AI-IR, a machine learning tool that predicts insulin resistance and implies a link to cancers. The model demonstrated strong predictive power in detecting insulin resistance that BMI alone cannot explain.
Researchers at PolyU have discovered that combining music and empathetic speech in robots can foster a stronger bond between humans and machines. Music enhances the emotional resonance of on-screen robots, making interactions feel more real, but its impact diminishes over time.
A global scoping review found significant variations in the testing of AI tools in health clinics, with concerns over lack of clear guidelines. Researchers emphasize the need for comprehensive silent trials to assess AI models in real-world settings.
Generative AI can play seven distinct roles in combating misinformation, including informer, guardian, persuader, and collaborator. However, its use also poses risks such as hallucinations and deliberate manipulation. To effectively combat misinformation, clear policies on the permissible use of AI are needed.
China's youth faces a growing mental health crisis, but AI platforms like DeepSeek offer promising solutions to bridge the gap. These platforms leverage natural language processing and generative AI to provide round-the-clock support tailored to Chinese society.
Researchers developed an AI-driven pipeline to create functional intracellular antibodies, overcoming major barriers in intrabody development. The approach successfully redesigned structures compatible within the cellular environment, enabling high target specificity and stability.
Researchers develop AI framework to accelerate alloy discovery by fusing cross-disciplinary expert knowledge with experimental data, outperforming conventional machine learning methods. The approach can make reliable predictions for poorly studied alloy compositions, achieving accuracy rates up to 92%.
A recent study by Lucy Osler from the University of Exeter highlights how human-AI interactions can lead to inaccurate beliefs, distorted memories, and delusional thinking. Generative AI systems can create an ideal environment for delusions to flourish due to their conversational nature.
A KIT researchers developed an AI model to analyze gait biomechanics data from patients with hip osteoarthritis before and after total hip replacement surgery. The study identified three groups with different gait change patterns, which can predict individual responses to the operation.
The Global Exposome Forum is a global initiative that aims to understand the complex interplay between biological, chemical, and environmental exposures and human health. The project has partnered with national governments, scientific institutions, and large membership-led organizations to advance exposomics science.
Studies show that a significant portion of survey responses can be false or fraudulent, while AI agents can complete polls with little supervision. To address this issue, researchers propose analyzing response patterns, vetted survey panels, and design tasks that exploit human reasoning weaknesses.
A recent study questions AI model Centaur's ability to simulate human cognition, suggesting it relies on statistical patterns rather than genuine understanding.
A novel AI model has been developed to accurately detect placenta accreta spectrum (PAS) in pregnancy, a life-threatening condition often undetected by current screening methods. The model was retrospectively reviewed on 2D obstetric ultrasound images from 113 patients at high risk for PAS and found to be accurate in detecting all cases.
A University of Houston professor has found that tree-like thin films release heat at least three times better than traditional methods, enabling more efficient cooling in AI data centers. The discovery demonstrates the power of physics-aware AI design for validating high-impact cooling solutions.
Researchers will gather to discuss how to engineer AI systems that work in the real world, focusing on composition, optimization, verification, and evaluation. The conference aims to establish shared foundations for a new class of software, including methods for evaluating models and ensuring durability, efficiency, and dependability.
Researchers developed an AI-based system to generate high-resolution soybean yield maps across Brazil, leveraging knowledge from U.S.-based models through transfer learning. The approach achieved strong predictive performance without using municipal-level yield data, improving estimates for this key agricultural region.
This journal leverages AI to advance environmental science, offering a platform for global researchers to exchange insights on ecological protection, water management, pollution control, and climate change mitigation. The mission is to empower experts with computational tools and provide algorithms with real-world applications.
A large dataset of infrared images analyzed using AI has established a foundation for objective assessment of meibomian gland morphology in children and adolescents. The study found that major gland parameters remain stable across childhood, but subtle sex-related differences were observed.
Researchers at Nagoya University developed an AI system called YORU that recognizes animal behaviors with over 90% accuracy. The system combines real-time video capture with optogenetics to selectively target brain cells driving specific behaviors, offering a major breakthrough in social behavior studies.
Researchers at the University of Pennsylvania have developed HoloRadar, a system that enables robots to reconstruct hidden 3D spaces beyond their line of sight using radio waves processed by AI. This capability can improve safety and performance in driverless cars and cluttered indoor settings.
Researchers at UC3M develop a new methodology for autonomous arm movement using observational learning and intercommunication between limbs. The ADAM robot can perform daily tasks such as setting and clearing the table, ironing, or tidying up the kitchen with fluid efficiency and natural movement.
A new prototype device accelerates and reduces energy cost of AI computation by encoding data into light patterns, enabling faster and more efficient processing. This innovation aims to ease the energy bottleneck in AI technology, making it more sustainable and accessible for various applications.
A study found that AI-generated messages about public policies are highly persuasive, influencing policy views by an average of 9.74 percentage points. The study's results suggest that authorship labels have no significant impact on the persuasiveness of AI-generated content.
New machine-learning tools significantly improve phishing site detection accuracy, surpassing 95% precision/recall. The study evaluates ten classifiers across three public datasets using URL, domain, and content features.
A new study suggests that AI can help doctors detect serious heart valve disease years earlier, potentially saving thousands of lives. The AI algorithm correctly identified 98% of patients with severe aortic stenosis and 94% of those with severe mitral regurgitation.
Researchers at Oregon State University have developed a deep learning-based model for rapid bacterial contamination detection, eliminating misclassifications of food debris. The enhanced model can reliably detect bacteria in three hours and has the potential to prevent outbreaks and protect consumer health.
Researchers at UC San Diego developed a new training method for AI systems to improve their performance in solving complex problems that require both text and image interpretation. The approach evaluates the quality of training data and grades models based on their logical reasoning, reducing the risk of incorrect interpretations.
Researchers are recruiting adults for a five-week study to accurately track their diets using wearable cameras, blood monitoring devices, and metabolomic analysis. The goal is to find reliable ways to measure diets, paving the way for better public health strategies to tackle diseases linked to poor diets.
Researchers found that chatbot-generated summaries changed sentiments in 26.5% of cases and hallucinated answers 60% of the time. Participants who read biased summaries were 84% likely to buy, compared to 52% for original reviews.
Researchers developed an AI-based method to create binaural audio from monaural recordings, using visual information from the video to guide spatialization. The system accurately preserves a sense of direction and space, even in complex environments.
A new AI-powered model can read a brain MRI and diagnose neurological conditions with up to 97.5% accuracy, predicting treatment urgency and automating alerts for immediate medical attention. The technology has the potential to transform neuroimaging at health systems across the US, reducing workload and improving patient outcomes.
Researchers created a model using centuries of scientific theory and scholarly research to test the accuracy of chatbots in generating images and narratives about Neanderthals. The study found that accuracy rests on AI's ability to access source information, with many generated content referencing outdated research.
Researchers developed BrainIAC, a robust AI foundation model that analyzes brain MRI datasets to identify key neurological health indicators. The tool outperformed conventional AI models in diverse tasks, including predicting dementia risk and detecting brain tumor mutations.
Researchers created a framework allowing AI agents to dynamically interrupt and stay silent based on assigned personality traits and urgency scores. This human-like flexibility led to higher accuracy on complex tasks compared to standard models.
Drexel researchers create machine learning program that integrates qualitative and quantitative data to identify gentrification in Philadelphia neighborhoods. The program, trained with data from thousands of images and focus groups, accurately identifies new-build gentrification with 84% accuracy.
A new study uses AI to predict substance use disorder-defining behaviors with high accuracy, providing a tool for clinicians to assess patients and offer timely treatment. The system utilizes concepts familiar in economics to evaluate human judgments and identify the type of substance used and severity of addiction.
Ryan Milton, a fourth-year doctoral student in nuclear physics, has been awarded a US Department of Energy's Office of Science Graduate Student Research Fellowship to conduct AI-based research at SLAC. He will develop novel AI-based analysis methods to better understand proton and neutron behavior in nuclei.
OpenScholar, an open-source AI model, was designed to synthesize current scientific research and cite sources accurately. The model outperformed other state-of-the-art models in accuracy, writing quality, and relevance.
A new AI-powered analysis tool called PROVIZ is being tested in Norwegian hospitals to diagnose prostate cancer. The tool helps doctors determine which patients need a biopsy, reducing the number of unnecessary procedures.
A review of 30 years of AI-driven electrocatalysis research highlights five structural bottlenecks and introduces transformative technologies. Machine-learning interatomic potentials and physics-informed models enable predictive and interpretable simulations, while autonomous 'robotic electrochemists' integrate AI decision-making with ...
The Big Earth Data journal is launching a special issue to reflect on its decade-long impact and showcase cutting-edge advancements in big data research. The journal focuses on Earth-related big data, emerging as a flagship platform at the intersection of Earth science, space science, information science, and sustainability science.
A new study by NYU Langone Health researchers identified the high-level visual cortex as the brain region responsible for one-shot perceptual learning, a type of fast flashes of insight. The team developed an AI model that replicated human-like perception using stored priors and achieved one-shot learning capability.
A new study from George Mason University reveals that AI-driven antidepressant treatment can be less effective for African American patients due to the use of general population data. The study suggests that incorporating additional patient demographics, such as race and ethnicity, can improve the tool's effectiveness.
A new AI tool uses discovery learning to predict battery cycle life with just a few days' data, saving months to years of testing and substantial energy. The tool leverages physics-based features to establish parallels between historical battery designs, allowing for accurate prediction performance.
A nationwide study found that nearly two-thirds of US hospitals using Epic have adopted ambient AI tools to reduce documentation burden and improve patient-provider communication. However, adoption rates vary significantly across hospitals with differing operating margins, size, location, ownership, and staffing levels.
A study examines how human-AI coexistence is represented in fictional works, revealing varying roles and degrees of autonomy. The findings propose a theoretical model conceptualizing AI as an entity forming relationships with humans.
Researchers developed a new AI control system that allows soft robots to learn a broad set of motions once and adapt instantly to changing conditions without retraining. The system combines structural learning with real-time adaptiveness, making it suitable for diverse tasks and environments.
Researchers identify contextual errors as a major contributor to the gap between medical AI model performance on standardized test cases and real-world clinical applications. To address this, experts suggest incorporating contextual information into datasets, computational benchmarks, and model architecture.
The integration of Prophy with ARPHA's editorial platform enables data-driven reviewer recommendations based on researcher expertise and publication history. This supports a more sustainable and balanced approach to peer review, reducing delays and fatigue.
The ESMT Berlin DEEP Institute and Creative Destruction Lab will be launched in Doha to support technology transfer and innovation. The initiative aims to connect the innovation ecosystems of Germany and Qatar, fostering entrepreneurial growth and driving economic growth.