Technology

In the world of computer science, bipartite matching is a fundamental problem that arises in various real-world scenarios such as ridesharing apps, organ donations, and advertising. This problem involves pairing entities from two disjoint sets in a way that maximizes overall efficiency and satisfaction. Cold Spring Harbor Laboratory Associate Professor Saket Navlakha has made significant

In the ongoing COOLBat joint research project, experts from the Fraunhofer Institute for Machine Tools and Forming Technology IWU have collaborated with industry partners to develop cutting-edge battery enclosures for electric vehicles. The primary goal of this project is to enhance the efficiency of electric vehicles by creating lighter enclosures that reduce carbon dioxide emissions

Self-driving vehicles have made significant advancements in recent years, thanks to the integration of artificial intelligence (AI) technology. This technology plays a crucial role in decision-making, sensing, and predictive modeling for autonomous vehicles. However, ongoing research at the University at Buffalo raises concerns regarding the vulnerability of AI systems in self-driving vehicles to malicious attacks.

Google, once synonymous with internet search, has been facing mounting criticisms regarding the quality and user-friendliness of its search results. The company’s founders, Sergey Brin and Larry Page, expressed concerns as far back as 1998 about the potential conflict between the goals of online advertising and the delivery of useful search results. Fast forward to

In Kenya’s bustling capital, Nairobi, children and young people alike are captivated by the sight of robots seamlessly carrying plates of food through a busy eatery. The Robot Cafe, a groundbreaking establishment in Nairobi and East Africa as a whole, has introduced three robots – Claire, R24, and Nadia – to serve meals alongside human

In a recent interview, Microsoft’s Vice President, Vik Singh, highlighted a crucial aspect missing in current generative AI models – the ability to admit when they are unsure and seek help. Singh emphasized the necessity for AI models, such as Copilot, to acknowledge their limitations and avoid providing inaccurate or misleading information to users. Despite

Reasoning is a vital process through which human beings analyze information to draw conclusions or solve problems. It can be broadly categorized into two main types: deductive reasoning and inductive reasoning. Deductive reasoning involves starting from a general rule or premise and applying it to draw conclusions about specific cases. On the other hand, inductive

Deep learning models have been the cornerstone of modern artificial intelligence, providing powerful tools for a wide range of applications. However, a team of researchers from the Alberta Machine Intelligence Institute (Amii) has uncovered a troubling issue that has been lurking in the shadows of deep learning: the loss of plasticity in continual learning. The

The research published in APL Machine Learning has brought together the worlds of optical illusions, quantum mechanics, and neural networks in a groundbreaking way. By leveraging “quantum tunneling,” the researcher was able to design a neural network that mimics human perception of optical illusions such as the Necker cube and Rubin’s vase. This neural network

In a recent development, artificial intelligence researchers have taken significant steps to address a troubling issue within the AI community. More than 2,000 web links to suspected child sexual abuse imagery were discovered in a dataset used to train popular AI image-generator tools. The dataset in question, known as the LAION research dataset, has been

Researchers from the University of Toronto have conducted a study that sheds light on how using virtual and augmented reality (VR and AR) technologies can temporarily alter the way individuals perceive and interact with the real world. The research, recently published in Scientific Reports, revealed that individuals exhibited different movement patterns in VR and AR,

Artificial Intelligence (AI) has become an integral part of our lives, shaping the way we interact with technology and influencing various aspects of society. However, a recent study conducted by a team of AI researchers has shed light on a concerning issue – the presence of covert racism in Language Models (LLMs). This revelation highlights

Ethiopia recently announced that it has significantly increased its electricity production from the Grand Ethiopian Renaissance Dam (GERD) on the Blue Nile by activating two additional turbines. This multi-billion-dollar project has been a subject of contention between Ethiopia and downstream nations like Egypt and Sudan. The completion of the dam’s construction has marked a transition

Robots have the potential to greatly assist humans in their daily tasks, but in order to do so effectively, they need to be able to manipulate objects with precision and dexterity. This is where the development of cost-effective robotic grippers comes into play. Recently, researchers at Improbable AI Lab and MIT have introduced a new

In recent years, the development of deep neural networks (DNNs) has seen significant progress in addressing real-world challenges. However, studies have shown that these models can exhibit unfairness, especially when it comes to recognizing features based on skin color. This raises questions about the impact of training data and hardware platforms on AI fairness. Research