Articles tagged with Artificial Intelligence
A new study examines individuals' willingness to adopt algorithms in decision-making contexts, revealing the importance of cognitive perceptions. Participants with low situational awareness were found to be more likely to delegate decisions to an AI algorithm.
A team of researchers from Ludwig-Maximilians-Universität München conducted a study on human perception and behavior towards robots. They found that when robots are humanized through anthropomorphic features, participants are less likely to sacrifice them to save human lives.
A new breed of algorithms has mastered Atari video games 10 times faster than state-of-the-art AI, using a combination of reinforcement learning and intrinsic motivation. This method autonomously identifies sub-goals and exhibits relatively human-like behavior, making it competitive with Google DeepMind in the game Montezuma's Revenge.
A novel model developed by MIT and Microsoft researchers identifies instances where autonomous systems learn from examples that don't match the real world, leading to dangerous errors. The model uses human feedback to pinpoint situations where more information is needed to act correctly.
Researchers explore how AI can analyze words and phrases, turn-taking patterns, and tone to improve medical communications. The technology has potential to revolutionize patient care by providing clinicians with assessments of their communication skills at lower costs.
The University of Texas at San Antonio has developed a framework using deep neural networks to classify and detect malware. This system was recognized as the top prize winner in the AICS 2019 Challenge sponsored by Crowdstrike Foundation and MIT Lincoln Laboratory, showcasing UTSA's expertise in cybersecurity.
A new software, idtracker.ai, can identify up to 150 individual fish with high accuracy, extracting valuable data for understanding group behavior. The AI-powered system uses deep learning neural networks and conventional algorithms to recognize unique features of each zebrafish.
A recent study by UC Berkeley reveals that current AI-powered tracking methods can identify individuals using their daily step data, posing significant threats to health data privacy. The research suggests that existing laws and regulations, such as HIPAA, are inadequate to protect individuals' confidential health information.
A UK-wide survey found that GPs are skeptical about AI's impact on their profession, with many believing technology can't replace empathic care. However, some GPs think AI can handle administrative tasks efficiently.
Researchers developed two classification models that gauge human trust in intelligent machines using electroencephalography and galvanic skin response. The models achieved a mean accuracy of 71.22 percent, enabling real-time trust estimation.
Researchers developed AlphaZero, a generalized game-playing program that can learn any game through self-play reinforcement learning. The system mastered complex games like chess, shogi and Go within hours of self-learning and outperformed state-of-the-art AI programs.
Cristiano Fanelli uses artificial intelligence to optimize particle detectors and study exotic hybrid mesons. He aims to apply deep learning techniques for accurate particle identification, revolutionizing nuclear physics research.
Researchers developed an AI Clinician system that predicts and recommends optimal treatment strategies for patients with sepsis, showing a higher chance of survival compared to human doctors. The system analyzed 100,000 hospital records and made more reliable treatment decisions than humans.
A new AI tool called DiffProf helps developers identify and optimize energy-draining features in their apps, improving battery life. The tool analyzes 'call trees' of similar tasks to show why certain features consume more energy than others, providing actionable diagnosis and fixes for developers.
Researchers at MUSC have developed a novel AI method that analyzes the neural networks of epilepsy patients to predict seizure-free outcome post-surgery. The technique, which maps the whole-brain connectome, achieved an accuracy rate of 79-88%, outperforming current clinical methods.
Researchers developed an ADC model that considers agent, deed, and consequence to explain variability in moral judgments. The study found that when stakes are high, consequences outweigh the nature of the deed, while lower stakes prioritize the agent's character.
The Quadratic Phenotypic Optimization Platform (QPOP) uses AI to analyze small datasets and design optimal drug combinations based on real experimental data. This approach can optimize and personalize combination therapy for individual patients with complex diseases, such as multiple myeloma.
Researchers at MIT CSAIL have developed a robot system that can inspect and understand objects, allowing it to perform specific tasks such as picking up objects from a cluttered space. The system, called Dense Object Nets, uses a self-supervised approach and doesn't require human annotations.
A commentary suggests using smart technology and artificial intelligence to improve the detection of sepsis in children in Canada. The authors propose automating data combinations of sepsis symptoms and other relevant information to create more accurate and low-cost trigger tools.
The National Institute of Biomedical Imaging and Bioengineering (NIBIB) at NIH will convene experts in academia, industry, and government to discuss the state-of-the-art of AI applications for medical imaging. The workshop aims to improve quality, reproducibility, and reliability of AI in medical imaging.
A team of researchers from Simon Fraser University has developed an artificial intelligence (AI) method to anonymize faces in news reports, aiming to improve visuals and amplify emotions tied to the story. The technique uses AI painting to simulate a smart painter, resulting in more engaging videos.
A new study from MIT found that doctors' written notes reveal their 'gut feelings' play a significant role in determining how many tests they order for patients, beyond just analyzing symptoms and medical data. The researchers used sentiment analysis to uncover this correlation, which was strongest at the beginning of a patient's hospi...
Researchers have developed an AI system that can identify and quantify cyanobacteria in water samples in about one to two hours, reducing the need for expensive lab tests. This tool could provide early warnings of problems, helping to ensure safe drinking water and preventing public health issues.
A study by the University of Helsinki found that science fiction enthusiasts are more likely to approve of mind upload technology, with traditional religiousness and disgust sensitivity to sexual matters linked to negative reactions. The research also explored moral psychology perspectives on the technology.
The article discusses the ethical implications of genetic testing and the duty to warn family members of hereditary disease risks. Experts argue that patients have a moral obligation to share their genetic information with relatives, citing the principle of rescue as a framework for decision-making.
The Swiss State Secretariat for Migration and the Immigration Policy Lab (IPL) are testing a new data-driven method to assign asylum seekers to cantons across Switzerland. The algorithm aims to maximize job chances, and its recommendations will be compared to those of randomly allocated individuals over several years.
Researchers have developed a system using wearable technology and artificial intelligence to detect the onset of respiratory and cardiovascular diseases. By analyzing data from sensors on smart shirts, they can predict health-related benchmarks during daily activities.
Researchers used AI and image analysis to reconstruct the structure of plants with hidden branch structures, including those under leaves. The new technology will help in future cultivation techniques such as detailed daily management and forecasting plant growth.
Researchers have uncovered characteristics of magnetic solitons, a type of magnetic wave relevant to developing neuromorphic computing systems. The findings reveal how these components act, which is essential for realizing their potential in brain-inspired computing.
Bento browser eliminates 'tab overload' by storing each search session as a project workspace, keeping track of relevant parts of visited web pages. Users can identify useful pages, trash unhelpful ones, and pick up where they left off later.
A new RAND Corporation paper warns that AI's potential to encourage risky behavior and undermine strategic stability poses a significant threat to global security. Improved sensor technologies and the rise of first-strike capabilities could erode mutual assured destruction, making nuclear war more likely.
A new tool called Squadbox, developed by MIT CSAIL, enables people who have been targeted by online harassment to coordinate a squad of friends to filter messages and shield themselves from abusive content. The 'friend-sourcing' approach aims to augment human support with tools in a meaningful way.
Alán Aspuru-Guzik, a world-leading researcher in theoretical and computational chemistry, joins the University of Toronto as a Canada 150 Research Chair. He aims to disrupt chemistry through computer science and robotics, with a focus on bio-inspired solar energy.
A Chinese AI system can diagnose prostate cancer with accuracy comparable to human pathologists, offering potential benefits for streamlining cancer diagnosis and overcoming pathologist shortages. The system was developed using a learning algorithm and showed promise in accurately classifying malignancy levels and Gleason Grades.
Researchers at Åbo Akademi University and Moffitt Cancer Research Centre are developing AI-driven precision cancer medicine to identify effective multi-drug therapies for each patient. The project combines molecular data with network modelling to analyze interactions between drugs and inhibitors, reducing the risk of drug resistance.
AutoSaw, a system developed by MIT CSAIL, lets users design and customize furniture templates using expert knowledge and robotics for cutting tasks. The system uses small mobile robots to cut lumber with minimal risk of injury.
Researchers simulated a sea slug's decision-making process in Cyberslug, a virtual environment that mimics the creature's behavior. The study found that hunger and learning ability are crucial attributes in regulating prey consumption, laying the groundwork for more complex models of AI decision-making.
Researchers from UC3M and BQ will develop a new camera system leveraging artificial intelligence to enhance photographic quality and identify visual patterns. The collaboration aims to foster education, training, and innovation in AI research and its applications.
Scientists at Brown University have created a way for robots to learn high-level abstract representations of their surroundings, enabling them to plan and act more autonomously. By using machine learning algorithms, robots can distill complex sensory data into simple, text-based descriptions that facilitate planning and skill execution.
Researchers at USC Center for Artificial Intelligence in Society developed an AI algorithm, SPOT, to detect poachers in near real-time using infrared images. The algorithm can distinguish between humans and animals in just over three-tenths of a second, reducing the time spent on monitoring wildlife.
Researchers at Carnegie Mellon University have created a conversational agent, Evorus, that can answer a wide range of questions using a human/machine hybrid system. The system, which combines crowd workers and AI, allows the agent to learn and improve over time, making it more accurate and effective in its responses.
Research using motion sensors and AI software reveals expert bricklayers use techniques to limit joint loads and develop efficient work methods. The study's findings could lead to improved training programs for apprentices and enhanced worker safety.
A deep learning computing system developed to classify retinal images can detect diabetic retinopathy, glaucoma, and age-related macular degeneration with high accuracy. The study used 494,661 retinal images from diverse patient populations.
Researchers developed a deep learning tool to automatically analyze road behavior and create searchable databases. The system accurately labeled and tracked objects in traffic videos, identifying potential close encounters between vehicles and pedestrians.
Researchers at MIT's CSAIL developed ComText, a system that allows robots to understand contextual commands, enabling them to perform complex tasks. With this technology, Baxter, a two-armed humanoid robot, achieved success in executing commands with a 90% accuracy rate.
A new AI-powered traffic management system called SimplifAI will be tested in Manchester, aiming to reduce congestion and improve air quality. The system, developed by Professor Lee McCluskey and his team, can analyze data and devise complex strategies in a fraction of the time required by human operators.
Researchers from MIT's CSAIL have developed InstantCAD, a tool that enables designers to interactively edit and optimize CAD models using a streamlined workflow. This allows for real-time improvements and optimizations, reducing the time required to test and refine complex designs.
Researchers from MIT's CSAIL developed an AI system called Pic2Recipe that can analyze photos of food and predict ingredients and suggest similar recipes. The system was trained on a database of over 1 million recipes and showed impressive results, particularly with desserts like cookies or muffins.
Researchers developed Empowerment to help robots and humans work together safely and effectively. The concept enables robots to prioritize human safety while maintaining their own autonomy.
Researchers developed an AI signature to detect long-term concussion effects using artificial intelligence, detecting up to 90% accuracy. The tool could aid in diagnosing and treating concussion-related symptoms, which often persist for years after the injury.
Researchers from MIT's CSAIL have developed a new system called Home3D that allows users to watch 3D movies at home without glasses. The system converts traditional 3D movies into a format compatible with automultiscopic displays, which show multiple images simultaneously.
Researchers at MIT developed a system that allocates cache access on the fly to create tailored 'cache hierarchies' for specific programs, increasing processing speed by 20-30% and reducing energy consumption by 30-85%. The system, called Jenga, uses a sampling algorithm to efficiently evaluate tradeoffs between latency and space.
A team of engineers at the University of Arizona is designing an autonomous border-surveillance system that uses artificial intelligence to collect, assess, and act on data in real time. The system will deploy drones on its own and integrate data from different sources, including aerial and ground vehicles, to respond effectively.
A study using artificial intelligence has found that automated embryo assessment systems are more accurate than those of embryologists, with a 76% accuracy rate. The system uses mathematical variables from time-lapse images to classify embryos automatically, reducing human error and improving predictive value.
USC researchers found that smiling during victory can hurt future cooperation chances, but smiling during a loss can increase success odds. The study used emotion-tracking software to analyze participants' facial reactions during the Golden Balls game show.
Researchers used AI to analyze medical imaging of 48 patients' chests, predicting which would die within five years with high accuracy. The study suggests that AI can recognize subtle patterns in images to predict medical outcomes, offering new hope for early disease detection and treatment.
A new wearable system uses a 3D camera to identify surfaces and objects, providing tactile signals through vibrating motors and a reconfigurable Braille interface. The system significantly reduces obstacles other than chairs from being contacted by visually impaired users, and minimizes cane collisions with people.
A team of MIT researchers created a system that allows directors to specify a shot's framing and generate control signals for a camera-equipped autonomous drone, preserving the framing as actors move. The system guarantees safe collision avoidance with obstacles.
A new Yale University study found that even 'dumb' AI can help human groups by boosting their performance. The experiment, which used an online game with human players and robotic AI players, showed a significant improvement in the median time for groups to solve problems.
Georgia Tech researchers have developed a team of free-flying robots that can maintain formation and avoid collisions. The robots use virtual top hats to maneuver in the air. Meanwhile, autonomous blimps with face-detecting capabilities are being built to recognize hand gestures and interact with humans.