This week’s list of top data news highlights covers November 23, 2024 to November 29, 2024, and includes articles on using machine learning to estimate cybersecurity risks and the world’s first fully robotic lung transplant surgery.
Anthropic, a San Francisco-based AI company, has introduced an open-source protocol that allows AI systems to connect to multiple data sources using a standardized framework. This tool, called the Model Context Protocol (MCP), eliminates the need for developers to write custom code for each data source, streamlining the process of integrating AI assistants with various datasets, tools, and applications. MCP allows AI systems to effectively use information from different tools and datasets without losing track of relevant details, making them more efficient at performing tasks.
2. Discovering Archeological Treasures
Researchers from Japan, France, Germany, and the United States have used AI to discover roughly 300 previously unknown Nazca Lines in Peru, which are a group of sacred geoglyphs marked on the desert. The researchers used an AI algorithm to analyze images that low-flying drones took and identified new geoglyphs, providing new insights into pre-Columbian sacred spaces and symbols in South America.
Intuit, a software company, has introduced new features in QuickBooks, its accounting software, that use AI to convert conversations and documents into invoices or bills, generate automated invoice reminders, and categorize expenses.
4. Estimating Cybersecurity Risk
Researchers at New York University have developed a machine learning algorithm trained on simulated data that can estimate cybersecurity risks in construction projects. The tool can estimate the risk of the five most common types of cyberattacks on construction companies: ransomware, insider attacks, data breaches, phishing, and supply chain attacks.
Researchers at the Brigham and Women’s Hospital in Boston have published a study demonstrating the potential of using health data collected from fitness trackers to detect mood episodes in people with bipolar disorder. The researchers used a machine learning algorithm to detect and identify time intervals between major mood changes, such as mania or depression. Alert systems using such algorithms could monitor patients with bipolar disorder in real-time.
A U.K.-based software company called Land App has launched a digital tool to help farmers decide on the best methods to plant new trees. The tool uses a series of questions to generate initial agroforestry designs for each farm, allowing farmers to visualize how they could incorporate trees into their existing farm space.
Runway, a New York-based AI startup, has launched Frames, a new AI video generator that turns text and video prompts describing cinematic scenes into brief video clips. The tool allows users to describe camera angles, lighting, and movements similar to how a human director would, or upload images to generate short videos in various styles such as cinematic dolly shots, drone footage, or slow motion.
Scientists at the Massachusetts Institute of Technology have generated realistic satellite images visualizing potential flooding using generative AI and a physics-based flood model and data on the intensity of an approaching storm in North Carolina. The tool can help prepare for potential flood impacts from storms by visualizing how different areas will be affected.
9. Performing Lung Transplants
A team of surgeons at NYU Langone Health in New York has successfully completed the world’s first fully robotic double lung transplant surgery using the Da Vinci Xi robotic system from Intuitive, a California-based biotechnology company. Da Vinci Xi contains vibration and tremor controls, 3D display and image processing, and force-sensing technology, which allows surgeons to feel the subtlest pressure on tissue during surgery.
10. Streamlining Cancer Diagnosis
Researchers at Ingenii, a New York-based quantum computing company, have developed a framework that uses quantum techniques to analyze mammography images faster. The model can process images up to ten times faster than classical methods without the need for annotated datasets, allowing doctors to diagnose breast cancer faster than current methods allow.
Image credit: Seiji Seiji
