Home BlogWeekly News 10 Bits: The Data News Hotlist

10 Bits: The Data News Hotlist

by Aswin Prabhakar

This week’s list of top data news highlights covers February 3, 2024 to February 9, 2024 and includes articles on identifying dogs using neural networks to assess the personalities of dogs and improving recycling accuracy with AI.

1. Predicting Antidepressant Responses

Researchers at Amsterdam University Medical Center and Radboud University Medical Center in the Netherlands have used AI to predict antidepressant effectiveness within a week, significantly cutting down the six to eight weeks it typically takes. By analyzing brain scans and clinical data, the researchers hope to enable faster, personalized treatment, reducing unnecessary prescriptions and enhancing patient care.

2. Automating Complex Tasks

Google has introduced Gemini Ultra, its most advanced large language model to date, replacing Bard. Gemini Ultra offers enhanced capabilities across various Google Workspace apps. The model can complete more complex tasks than previous models, such as coding and following nuanced instructions.

3. Enhancing NFL Player Safety

The National Football League (NFL) is leveraging AI to enhance player safety with its Digital Athlete platform. This platform uses machine learning and computer vision to predict injury risks by analyzing game and practice footage, player positions, and movements. By integrating data from RFID tags, optical tracking cameras, and other sources, it aims to develop personalized training programs and inform league rule changes to mitigate injury risks.

4. Deciphering Ancient Scrolls

A team of international researchers has used AI to decipher texts from 2,000-year-old scrolls buried by Mount Vesuvius, revealing secrets held since 79 CE. The team used digital scanning and machine learning to read texts previously hidden in carbonized papyri, offering insights into ancient philosophies and cultures.

5. Improving Transit Transparency

The Maryland Transit Administration (MTA) has unveiled a new data dashboard, enhancing transparency with real-time updates on bus and train reliability. This tool displays updated metrics on ridership and service delivery, aiming to improve the customer experience by providing more accurate and accessible information.

6. Recycling with AI

Bollegraaf, a leading recycling plant builder from the Netherlands, and Greyparrot, an AI startup based in the UK, are equipping facilities worldwide with AI to enhance sorting accuracy. This collaboration aims to improve recycling rates and reduce contamination, making recycled materials more valuable and supporting global sustainability efforts.

7. Predicting Chemical Reactions

Researchers at Friedrich Schiller University of Jena in Germany have adapted an AI chatbot to provide insights into chemistry, using it to predict molecular properties and reaction yields. They hope this approach will enable labs with fewer resources to better conduct chemistry research by requiring less data for fine-tuning and achieving accuracy comparable to specialized tools.

8. Approving Government Permits 

The Canadian province of British Columbia is adopting low-code platforms and open-source software to speed up permit approvals. This approach ensures that online applications are automatically updated to reflect latest legislative requirements, and integrated with back-end processes to make the permit process more straightforward and reduce errors.

9. Evaluating Working Dogs’ Personalities

A multi-disciplinary research team has developed an AI algorithm that automates the process of evaluating working dogs’ personalities to accurately assess which animals are likely to succeed in aiding law enforcement and people with disabilities over the long term.

10. Reducing Data Needs for AI

Researchers at the Massachusetts Institute of Technology (MIT) have discovered a method to develop AI models using less data. By applying a concept from mathematics called Weyl’s law, which takes advantage of the symmetry in data, they significantly reduce the data needed for AI training.

Image credits: Unsplash user Matt Nelson


You may also like

Show Buttons
Hide Buttons