This week’s list of top data news highlights covers May 25, 2024 to May 31, 2024 and includes articles on a robot that guides people with visual impairments in public spaces and using a deep learning algorithm to detect Earth-like planets outside the solar system.
Researchers at Duke University have developed a machine learning model to help medical personnel better read electroencephalography (EEG) charts of patients, which display the electrical activity of a brain and help determine when unconscious individuals are experiencing seizure-like episodes or are in danger of having a seizure. The researchers trained the model to identify patterns in EEGs using samples from over 2,700 patients, creating a visual categorization system that improves decision accuracy. This model could save thousands of lives by enhancing the diagnosis of seizure-like events.
Paris-based AI company Mistral AI has launched a coding assistant called Codestral that developers trained on over 80 programming languages, including popular ones like Java, Python, and C++. Codestral can complete coding functions, write tests, and fill in partial code, enhancing developer productivity and reducing errors.
3. Guiding the Visually Impaired
Scientists at the University of Glasgow are developing RoboGuide, an AI-powered robot dog that can guide people with visual impairments in navigating public spaces. RoboGuide uses advanced sensors and a large language model to map environments, avoid obstacles, and provide interactive guidance.
4. Establishing Quantum Networks
Scientists at Harvard University have partnered with Amazon Web Services to establish Boston’s first quantum network for transmitting digital information between two quantum processors. The network is the longest fiber quantum network so far, running between Boston and Cambridge in Massachusetts.
The Pennsylvania Department of Health has launched a new data dashboard to help residents navigate the seasonal increase in tick bites, which can carry Lyme disease. This dashboard features data on tick borne disease activity, prevention tips, and safe tick removal practices.
A San Francisco-based startup called Autobiographer has launched an app that uses generative AI that writes biographies for users by capturing, preserving, and sharing their life stories through voice-to-text content generation. Users can record their stories, discuss their feelings, and share everyday memories, and then the app will use Anthropic’s Claude-3 model to turn the voice prompts into a cohesive written prose.
A group of international researchers has published a new study demonstrating how a deep learning algorithm can help find new exoplanets, which are Earth-like planets outside the solar system. The algorithm used data on radial velocity, which measures an object’s motions towards or away from an observer, to reduce noise from stellar activity and successfully detect several exoplanets.
Perplexity, an AI search platform, has launched a new tool called Pages that generates written reports based on user prompts. Users enter a topic or question they want the report or webpage to cover, and Perplexity’s AI model searches the Internet for relevant information, organizes it into sections, cites sources, and adds visuals.
Indonesia’s government has launched a new app called INA Digital that will serve as a single, centralized digital platform for citizens to sign in and access various government services, such as obtaining driver’s licenses, health services, and police permits for public assembly.
10. Creating Ultra-Thin Sensors
Researchers at the University of Cambridge have developed ultra-fine sensors, 50 times thinner than human hair, which can monitor human health and environmental changes without being noticeable. The researchers designed the ultra-fine sensors to be placed directly on the body and collect data on body signals like heartbeats by detecting and transmitting electrical activity to connected devices for monitoring and analysis.