This week’s list of top data news highlights covers May 4, 2024 to May 10, 2024 and includes articles on analyzing secret intelligence and using generative AI to recreate the voice of a famous country music star.
Google is launching a new cybersecurity product called Google Threat Intelligence that uses a large language model (LLM) to improve the cybersecurity measures that organizations take. The product uses the Gemini 1.5 Pro LLM to analyze and understand malicious software by examining its code and behavior, looking for patterns and indicators that help cybersecurity experts understand how the software works and what threats it poses.
The City of New York has partnered with a Vancouver-based tech firm called UrbanLogiq to create a data hub to support business activity and economic growth in key neighborhoods. The new digital tool provides real-time data on things like storefront vacancies and consumer visitation patterns to enable businesses to offer tailored services based on the unique makeup of each neighborhood.
Researchers at the Massachusetts Institute of Technology (MIT) have designed a radiation detector. The tool uses a machine learning algorithm to process radiation data and map radioactive sources, allowing it to discover where radiation is coming from, and improves upon traditional detectors because it receives information even from a distance, minimizing the need for close exposure to radiation sources.
Apprio, a Washington, D.C. and Orlando-based tech firm that provides administrative solutions to hospitals, has developed a software that can automate the billing process for hospitals. A large hospital chain in Tampa Bay has used the software and estimates the new automated bot has generated $12 million in annual savings.
5. Overcoming Health Barriers to Art
Country music star Randy Travis has used an AI-generated recreation of his voice to release a new track called “Where That Came From.” AI developers trained two AI models on vocal stems plucked from the singer’s discography that dates back to 1985. Travis had not been able to sing since suffering a debilitating stroke ten years ago.
6. Speeding Up Robotics Training
Researchers at the University of Pennsylvania, University of Texas at Austin, and California-based multinational tech company Nvidia, have developed an LLM called DrEureka to help train a robot dog. DrEureka analyzes data on the robot dog and designs training programs to improve the robot’s skills.
A Seattle-based tech startup called Hubble Network has successfully made a Bluetooth connection from a satellite in space to Earth, bridging a distance of over 370 miles. Typically, using Bluetooth for high-bandwidth activities requires a lot of power. Hubble Network uses a lower bandwidth, lower energy version of Bluetooth, which is capable of longer-range transmission and could be useful in wildlife monitoring or agricultural contexts.
8. Detecting Document Forgeries
Researchers at ETH Zurich University have developed a smartphone app that verifies whether a document is authentic or a fake using a QR code linked to an encrypted version of the document. The app analyzes images captured during verification in real-time and flags discrepancies.
9. Improving Secret Intelligence Analysis
Microsoft has developed a generative AI platform exclusively for the U.S. intelligence community to streamline analyzing top secret information. The new tool, which will enter testing phase soon, is the first major LLM that is fully isolated from the Internet to ensure maximum security and protection of sensitive national security data.
10. Reducing Healthcare Barriers
A group of researchers at the University of Hong Kong have used deep learning technology to improve image quality of low-power magnetic resonance imaging (MRI) scans, which doctors use to study the internal organs of a patient. According to the study, integrating deep learning into MRI technology could help produce more affordable machines, making it more accessible to low-income populations.
