machine learning

Concrete underlies nearly every modern city’s infrastructure—from skyscrapers and highways to bridges and sidewalks. Yet its primary binding agent, Portland cement, accounts for an estimated 7–8 percent of global carbon dioxide emissions. As governments, companies, and researchers pursue decarbonization strategies, reimagining concrete’s composition emerges as a critical frontier. A recent collaboration between the Olivetti Group and MIT’s Concrete Sustainability Hub (CSHub) has turned to artificial intelligence to tackle the vast search for alternative cementitious materials. Their results, published in Nature Communications Materials on May 17, outline a data-driven framework that combs through scientific literature and geochemical databases to identify viable replacements for cement. By pairing AI-powered text mining with experimental insights, the team has charted a path toward more sustainable, circular concrete that could dramatically cut emissions and costs—without compromising durability or strength.

In a major step toward more intuitive and efficient human-robot collaboration, engineers at the Massachusetts Institute of Technology (MIT) have unveiled a groundbreaking system called “Relevance”, which enables robots to determine, in real time, what objects in a scene are most important for assisting humans.

Accurate identification of disease-carrying mosquitoes has long been a challenge for scientists due to the tiny size and striking similarities among species. Now, a new AI-based tool being developed at Utah State University (USU) could revolutionize how we track and manage these pests, ultimately improving disease control efforts.

In everyday life, when someone sees something that isn’t there, we call it a hallucination. In the realm of artificial intelligence (AI), a similar phenomenon occurs when algorithmic systems generate information that sounds convincing yet is ultimately inaccurate, misleading, or entirely fabricated. These errors are known as AI hallucinations. As AI systems, from chatbots to image generators and autonomous vehicles, become increasingly integrated into our daily routines, understanding AI hallucinations and their implications is more important than ever.

AI systems like ChatGPT don’t actually learn the way humans do. Here’s what that means for you.

Researchers at the Massachusetts Institute of Technology (MIT) have developed a revolutionary new "fibre computer" that can be woven directly into clothing, offering a potential alternative to traditional wearable electronics like smartwatches and fitness trackers. This new technology could redefine the future of health monitoring, fitness tracking, and digital communication.

A viral video has ignited discussions about AI-to-AI communication after two artificial intelligence agents were caught on camera ditching human speech in favor of an exclusive machine-only language.

Explore the groundbreaking AI algorithm developed by Caltech's team, led by mathematician Sergei Gukov, that revolutionizes problem-solving in mathematics. This innovative algorithm excels at tackling complex issues, including the challenging Andrews-Curtis conjecture, by utilizing advanced machine learning techniques like deep neural networks and reinforcement learning. Discover its implications for mathematics, cryptography, and various scientific fields, as well as its transformative potential to democratize access to cutting-edge mathematical solutions. Keep abreast of how this synergy of AI and mathematics is reshaping our understanding of computational challenges.