Articles tagged with Artificial Intelligence
The University of Ottawa's AI + Society Initiative proposes a roadmap for protecting democracies in the age of AI, addressing challenges posed by fake news and deepfakes. Four priority actions aim to support institutions against AI interference on elections and democratic processes.
Researchers created an automated system that enables developers to build algorithms taking advantage of two types of data redundancy, boosting computation speed and reducing energy consumption. The system can optimize machine-learning algorithms for a wide range of applications, including scientific computing.
A team of MIT engineers has developed a training method for multiagent systems that can guarantee their safe operation in crowded environments. The method enables agents to continually map their safety margins, allowing them to scale up to any number of agents while maintaining system safety.
A new study published in The Lancet Digital Health found that AI-supported breast cancer screening increased the detection of early-stage invasive cancers by 24% compared to traditional screening. AI also identified pre-cancerous lesions with a 51% higher rate, without increasing false positives.
The project's recordings help improve voice recognition tools by providing diverse speech patterns to train artificial intelligence models. Microsoft has seen significant improvements in recognizing non-standard English speech, with accuracy gains ranging from 18% to 60%, depending on the speaker's disability.
The TU Graz AI system optimises vehicle components using simulation models and evolutionary algorithms to reduce development time by several months. It considers multiple objectives including production costs, efficiency and package space requirements, as well as CO2 emissions across the entire supply chain.
Researchers developed an AI model that analyzes electrocardiogram (ECG) data to predict biological age and potential cognitive decline. The study, based on UK Biobank data, found a link between higher ECG-age and poorer cognitive performance.
Binghamton University researchers aim to perfect the electrospray deposition process using artificial intelligence to create consistently thin polymer films. The project seeks to overcome challenges in controlling film characteristics and data collection, enabling efficient manufacturing and reducing labor costs.
Researchers developed an AI approach using artificial intelligence to analyze multiple parameters, including medical history, laboratory values, imaging, and genetic analyses. The AI model combined the data to calculate an overall prognosis for each individual patient, making it a potential tool for personalized cancer therapy.
A recent study led by H. Lee Moffitt Cancer Center researchers found that artificial intelligence can help doctors make better decisions when treating cancer. The study showed that collaboration between AI and human judgment improves radiotherapy outcomes for patients with non-small cell lung cancer and hepatocellular carcinoma.
The Keysight-UMA Lab focuses on testing and measurement equipment for 6G networks, while the Victoria Network promotes technology through end-to-end solutions. The facilities aim to collaborate with industry and academic institutions, advancing 6G research in Spain and positioning the University of Malaga as an international benchmark.
The iKNOW system combines VR and AI features to provide a natural learning environment for students with autism. It includes avatar interactions and real-time speech recognition, allowing students to practice social skills in a more relatable setting.
Researchers found that training AI agents in a less noisy environment can lead to better performance than traditional methods. The indoor training effect suggests that constructing simulated environments with specific noise levels can improve AI learning.
A new AI-based analysis of almost 10,000 pregnancies has discovered previously unidentified combinations of risk factors linked to serious negative pregnancy outcomes. The study found that there may be up to a tenfold difference in risk for infants who are currently treated identically under clinical guidelines.
A team of researchers proposes a new generation of language models inspired by human language acquisition. They demonstrate how artificial agents learn linguistic constructions linked to their environment and sensory perceptions, leading to more realistic language understanding.
Researchers developed MUNIS, a deep learning tool that predicts CD8+ T cell epitopes with high accuracy, potentially accelerating vaccine development. The tool was validated using experimental data from influenza, HIV, and EBV, demonstrating its potential to streamline vaccine design.
The University of Bath's new Institute for Digital Security and Behavior aims to put people at the heart of research into digital security and behavior. Researchers will study factors that make individuals, organizations, and societies safer and more secure.
Researchers used AI to identify genetic factors in Parkinson's disease progression and FDA-approved drugs that can be repurposed for treatment. The study found potential risk genes like SNCA and LRRK2, which are known to cause inflammation, and identified candidate drugs such as simvastatin.
The UK parliament has been called upon to develop a framework for the safe and productive use of AI in education, amid concerns about inequalities and digital divides. A key priority is ensuring that policymakers provide training and guidance on AI application, enhancing student learning experiences.
A collaborative study investigates how advanced AI tools can make it easier to evaluate interventions for ageing, providing personalised recommendations. The researchers identified eight critical requirements for effective AI-based evaluations and found that following specific guidelines improved the quality of the recommendations.
Neuromorphic computing is poised to emerge into full-scale commercial use, driven by the need for energy-efficient solutions. The review article proposes strategies for building large-scale neuromorphic systems that can tackle complex real-world challenges.
A new AI-powered software, EmbryoNet, can automatically detect defects in animal embryos and link them to underlying signalling pathways. This technology has the potential to replace lengthy animal studies in drug research, speeding up conventional processes while reducing costs significantly.
Researchers propose several key features to optimize sparsity, massive parallelism, and hierarchical structure in neural representation for neuromorphic systems. The goal is to achieve energy efficiency and compactness while retaining information at high fidelity.
A new large language model, 'llm-jp-3-172b-instruct3', has been released with approximately 172 billion parameters, trained from scratch using 2.1 trillion tokens of data. This model surpasses GPT-3.5 performance on benchmarks such as 'llm-jp-eval' and 'llm-leaderboard'.
A new study found that AI-generated empathetic responses were preferred over those from humans and expert crisis responders. The researchers suggest that AI can supplement human empathy, but should not replace it entirely due to potential biases and ethical concerns.
Using machine learning and AI, researchers have discovered a way to quickly and economically find high-performance materials for perovskite solar cells. By training an AI model on 101 synthesized molecules, they were able to identify 48 new molecules with potential for above-average efficiency.
A new contrastive learning model has been developed to forecast cyclone rapid intensification (RI) with high accuracy, reducing false alarms by a factor of three compared to existing techniques. The model achieved an impressive 92.3% accuracy when tested on data from the Northwest Pacific between 2020 and 2021.
A new study from the University of Eastern Finland finds that digital platforms and artificial intelligence can boost B2B service companies' competitiveness by enhancing agility, flexibility, and resilience. The study also highlights the importance of organizational capabilities for managing strategic change in volatile markets.
Scientists at the University of Lausanne used AI to simulate the last Alpine glaciation, finding ice covers 35-50% thinner than previous models. The new approach enables unprecedented accuracy and resolution, making it possible to describe complex topography numerically.
The new model, based on a PV-RNN framework, achieves compositionality by combining language with vision, proprioception, working memory, and attention. It requires less computing power than large language models (LLMs) and makes mistakes similar to humans.
BioChatter bridges the gap between large language models and biomedical research by providing a transparent and adaptable framework for custom research tasks. The platform can integrate with knowledge graphs and bioinformatics tools, making it easier for researchers to analyze complex datasets.
A groundbreaking AI tool called Moscot maps single cells in spatial tissues, enabling the observation of millions of cells simultaneously. This allows researchers to analyze complex cellular processes within entire living organs and organisms, providing new insights into diseases like diabetes.
Establishing a permanent lunar presence will depend on the use of AI, robotics, and 3D printing to adaptively respond to challenges. The moon's natural resources, such as regolith, can also reduce the need for Earth-launched supplies.
The study proposes a 'school within a school' approach to enable small changes that can affect entire systems. Researchers highlight the need for transformational changes to better prepare students for the age of Artificial Intelligence.
A new study published in Environmental DNA revealed that insecticides and fungicides are the main factors affecting biodiversity in UK waters. The AI technology analyzed water and biofilm samples from 52 freshwater lakes, identifying key links between pollutants and biodiversity loss.
A recent study by Aalto University found that successful AI adoption depends on employees' emotional and cognitive trust. The research team analyzed a consulting company's attempt to implement an AI tool, which ultimately failed due to employees' discomfort with AI following their digital footprints.
A new study published in Journal of Management Studies reveals that employee perceptions of AI performance and feelings towards AI vary, affecting AI performance and adoption. Four different trust configurations were identified, leading to distinct behaviors and a vicious cycle of biased data inputs degrading AI performance.
A new interposer design uses optical connections to transfer data between chips, enabling high-performance computing and addressing the 'memory wall' limit on AI growth. The technology allows for faster communication, reconfigurable pathways, and real-time network control.
An English literature graduate has developed a new method for large language models to understand and analyze short text chunks, such as those on social media profiles. The method successfully grouped nearly 40,000 Twitter user biographies from accounts tweeting about US President Donald Trump into 10 categories.
Researchers developed ProteinReDiff, an AI-powered method to redesign proteins for improved ligand binding. The approach uses initial protein sequences and ligand SMILES strings, reducing reliance on detailed structural data.
A new study assesses the historical knowledge of AI chatbots like ChatGPT-4 and finds they struggle with nuanced, PhD-level inquiry. The models performed best on legal systems and social complexity but struggled with topics such as discrimination and social mobility.
The Global Conference on Gerophysics aims to bridge statistical mechanics, complex systems theory, and dynamical modeling with biological mechanisms of aging. The conference will bring together leading scientists to promote healthy longevity and reshape our understanding of the aging process.
The paper highlights the need for a critical re-evaluation of how we study digital technologies' impact on young people's mental health. Oxford researchers call for a collaborative evidence-based framework to regulate AI integration into online platforms, ensuring its safety and benefits for children and adolescents.
The workshop explored how to responsibly integrate AI into society, discussing key stakeholders, transparency, accountability, and human-AI collaboration. EU researchers emphasized the importance of addressing ethical considerations in AI design and governance.
The open-access book delves into the impact of AI and autonomous systems on human lives and ways of working, emphasizing their importance in prioritizing human well-being and creativity. Key learnings include the need for ethical design practices and considering AI as part of a larger system to enhance human experiences.
The KAIST team developed a next-generation neuromorphic semiconductor-based integrated system that can learn and correct errors on its own. This technology will revolutionize AI use in everyday devices, making them faster, more private and energy-efficient.
Researchers at NIMS developed a next-generation AI device leveraging ion-controlled spin wave interference in magnetic materials, outperforming conventional devices by up to 10 times. The technology enables energy-efficient computations with minimal degradation when miniaturized, opening doors for various industrial applications.
Researchers found that wearable devices can identify, differentiate, and predict flares in IBD by analyzing circadian patterns of heart rate variability, heart rate, oxygenation, and daily activity. The study showed that these metrics changed up to seven weeks before flares developed, enabling continuous disease monitoring.
A new study has revealed the alarming extent to which glaciers have shrunk over the past 40 years in Svalbard, with an area loss of over 800 km² detected. The research found that more than half of the glaciers undergo seasonal cycles in glacier calving, highlighting the vulnerability of these ice masses to climate change.
The Abundant Intelligences program challenges the current AI trajectory, incorporating Indigenous knowledge systems to create an alternative approach. This initiative seeks to rethink how we conceive of AI, prioritizing human thriving, preserving Indigenous languages, and addressing environmental issues.
A new study by Mass General Brigham investigators found that Black people were less likely than non-Black people to undergo metabolic and bariatric surgery, even though both groups were equally likely to discuss it with health care providers. Men were also less likely to discuss and receive MBS than women.
A new automated job hazard analysis tool promises to significantly reduce workplace accidents and improve safety in the construction industry. The University of South Australia's research team has built a 'knowledge graph' to predict hazards, which can be analysed in real-time to identify potential risks and control measures.
The partnership establishes Ohio's first specialized degree programs and research experiences in quantum computing, cultivating scientific and entrepreneurial talent. Miami University and Cleveland Clinic will accelerate the leading-edge power of quantum computing to shape the future of healthcare and technology.
A new method has improved AI translation of sign language by adding data on hand and facial expressions, as well as skeletal information. This has led to a significant increase in accuracy, making it easier for people with hearing impairments to communicate.
A multicenter study shows trained health professionals with AI assistance can meet diagnostic standards for lung ultrasounds. This technology has the potential to increase access to lung ultrasound in areas without expert personnel.
The 'MarginCall' project aims to create a fully integrated system for real-time, highly accurate assessments of surgical margins during cancer surgeries, reducing the likelihood of cancerous cells being left behind. This approach promises to be faster and more cost-effective than existing techniques.
DNNs have an inbuilt 'Occam's razor,' favouring simpler solutions that fit training data. This bias helps them generalize well on simple patterns but may struggle with complex data, aligning with real-world data characteristics.
A new AI model will analyze combined data from calcium-scoring CT scans, clinical risk factors, and demographics to identify high-risk patients. The project aims to accurately predict cardiovascular events, including heart failure, and provide personalized treatment options.
The University of Texas at Austin appoints Charles 'Charley' Taylor to lead a new Center for Computational Medicine, developing advanced medical applications and modeling. This move strengthens collaborations between Dell Medical School and the Oden Institute, driving innovation in health and technology.
A research team at George Mason University is developing mixed reality and AI technologies to improve first responder training. They aim to reduce training-related risks and casualties by creating interactive scenarios that mimic real-world challenges.