https://techxplore.com/rss-feed/computer-sciences-news/ en-us The latest news on computer science, computer science technology, computer science technologies and technology science. Computer scientists at UC Riverside have identified troubling flaws in a new generation of artificial intelligence (AI) agents designed to take over routine computer chores while users are away—sorting emails, organizing files, analyzing data, and handling other everyday digital tasks that might otherwise consume hours. https://techxplore.com/news/2026-05-ambition-ai-agents-tasks-digital.html Computer Sciences Thu, 14 May 2026 11:00:19 EDT news697975201 New work explaining the inner workings of artificial intelligence could provide a way around the threat of AI "model collapse," potentially averting growing numbers of AI hallucinations in the future. https://techxplore.com/news/2026-05-real-world-ai-collapse-analysis.html Computer Sciences Thu, 14 May 2026 10:00:01 EDT news697886395 Over the past few decades, computer scientists have developed increasingly advanced artificial intelligence (AI) systems that can tackle some tasks exceedingly well. These include computer vision models, systems that can rapidly analyze images and categorize them, recognize objects and faces, or make other accurate predictions. https://techxplore.com/news/2026-05-human-pipeline-vision.html Computer Sciences Fri, 08 May 2026 09:20:05 EDT news697448555 Researchers from MIT and elsewhere have developed a more user-friendly and efficient method to help networking engineers identify potential system failures before they cause major problems, like a cloud service outage that leaves millions of users unable to access applications. https://techxplore.com/news/2026-05-stress-method-cloud-algorithms-network.html Computer Sciences Wed, 06 May 2026 14:40:03 EDT news697296421 Artificial intelligence systems based on neural networks—such as ChatGPT, Claude, DeepSeek or Gemini—are extraordinarily powerful, yet their internal workings remain largely a "black box." To better understand how these systems produce their responses, a group of physicists at Harvard University has developed a simplified mathematical model of learning in neural networks that can be analyzed mathematically using the tools of statistical physics. https://techxplore.com/news/2026-05-simple-physics-ai.html Computer Sciences Tue, 05 May 2026 00:00:01 EDT news697101141 Today, artificial intelligence can describe images, recognize objects, and explain complex relationships. The pace of development is remarkable: So-called vision-language models (VLMs) combine text and image understanding in impressive ways. Yet, of all things, they struggle with a seemingly simple task—counting. Researchers at the Institute for Information Systems (iisys) at Hof University of Applied Sciences are now working to address this issue, with a paper posted to the arXiv preprint server. https://techxplore.com/news/2026-05-ai.html Computer Sciences Mon, 04 May 2026 10:20:01 EDT news697106582 Paintings are often made up of thousands of tiny brushstrokes, each going in a certain direction, that are not easily observed by the viewer. A cross-disciplinary research team from the Penn State College of Information Sciences and Technology (IST) and Loughborough University in England has developed an image analysis method that helps to make the underlying brushstroke structure of paintings visible, giving new insight into how artists physically created their works. https://techxplore.com/news/2026-05-vision-iconic-artworks.html Computer Sciences Fri, 01 May 2026 13:00:03 EDT news696858541 Artificial intelligence that cannot explain how it makes decisions—often called "black box" AI—could soon be replaced by more transparent systems, research suggests. A study by Loughborough University, published in Physica D: Nonlinear Phenomena, outlines a new mathematical blueprint for building AI that can reveal how it learns, remembers, and makes decisions. https://techxplore.com/news/2026-04-black-ai-scientists-blueprint-transparent.html Computer Sciences Thu, 30 Apr 2026 13:20:02 EDT news696772801 The new trend of "vibe coding" allows people to program software without writing a single line of code. Now, a new study by ETH Zurich published in the Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems has shown that users who want to develop apps and programs successfully with AI need not only a capacity for clear written expression, but also a basic knowledge of computer science. https://techxplore.com/news/2026-04-skills-people-successfully-ai.html Computer Sciences Thu, 30 Apr 2026 10:20:04 EDT news696762426 In today's hospitals and clinics, a dermatologist may use an artificial intelligence model for classifying skin lesions to assess if the lesion is at risk of developing into a cancer or if it is benign. But if the model is biased toward certain skin tones, it could fail to identify a high-risk patient. https://techxplore.com/news/2026-04-whac-mole-dilemma-smarter-debias.html Computer Sciences Thu, 30 Apr 2026 09:20:03 EDT news696758402 Most contemporary artificial intelligence (AI) systems learn to complete tasks via machine learning and deep learning. Machine learning is a computational approach that allows models to uncover patterns in data that are useful for making predictions. Deep learning, on the other hand, is a subset of machine learning that entails the use of multi-layered neural networks, which can autonomously extract features and learn complex patterns from unstructured data, sometimes with little or no human supervision. https://techxplore.com/news/2026-04-brain-approach-ai-overconfidence.html Computer Sciences Wed, 29 Apr 2026 09:00:02 EDT news696520698 Confidence is persuasive. In artificial intelligence systems, it is often misleading. Today's most capable reasoning models share a trait with the loudest voice in the room: They deliver every answer with the same unshakable certainty, whether they're right or guessing. Researchers at MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) have now traced that overconfidence to a specific flaw in how these models are trained, and developed a method that fixes it without giving up any accuracy. The team's research is published on the arXiv preprint server. https://techxplore.com/news/2026-04-ai-im-cases-calibration-errors.html Computer Sciences Wed, 22 Apr 2026 16:20:04 EDT news696092821 Large language models (LLMs) can teach other algorithms unwanted traits, which can persist even when training data has been scrubbed of the original trait, according to new research published in Nature. In one example, a model seems to transmit a preference for owls to other models via hidden signals in data. The findings demonstrate that more thorough safety checks are needed when producing LLMs. https://techxplore.com/news/2026-04-ai-chatbot-student-owls.html Computer Sciences Wed, 15 Apr 2026 15:40:04 EDT news695485082 Researchers at North Carolina State University have developed a new AI-assisted tool that helps computer architects boost processor performance by improving memory management. The tool, called CacheMind, is the first computer architecture simulator capable of answering arbitrary, interactive questions about complex hardware-software interactions. https://techxplore.com/news/2026-04-cachemind-chip-tuning-conversation-exposing.html Computer Sciences Wed, 15 Apr 2026 11:40:02 EDT news695470804 Perfect AI alignment with human values and interests is mathematically impossible, according to a study, but behavioral diversity among AI agents offers the promise of some control. Published in PNAS Nexus, Hector Zenil and colleagues used Gödel's incompleteness theorem and Turing's undecidability result for the Halting Problem to show that any LLM complex enough to exhibit general intelligence or superintelligence will also be computationally irreducible and produce unpredictable behavior, making forced alignment impossible. https://techxplore.com/news/2026-04-alignment-ai-human-values-mathematically.html Computer Sciences Tue, 14 Apr 2026 12:20:08 EDT news695387643 What if ChatGPT answered with the name of a minister from a year ago when asked, "Who was the minister inaugurated last month?" This is a prime example of the limitations of AI that fails to properly reflect the latest information. A KAIST research team has developed a new evaluation technology that automatically reflects changing real-world information while catching "temporal errors" that may appear correct on the surface. This is expected to drastically improve AI reliability. https://techxplore.com/news/2026-04-ai-temporal-errors-reliability-medical.html Computer Sciences Tue, 14 Apr 2026 08:40:04 EDT news695374741 In a world where artificial intelligence is quietly shaping who gets hired, who receives loans, and even how medical decisions are made, a new question is emerging: How does AI judge us? A new study by Prof. Yaniv Dover and Valeria Lerman from Hebrew University suggests the answer is both reassuring and deeply unsettling. The study is published in the journal Proceedings of the Royal Society A Mathematical Physical and Engineering Science. https://techxplore.com/news/2026-04-revealing-hidden-logic-ai-judgments.html Computer Sciences Mon, 13 Apr 2026 17:00:01 EDT news695318161 Published in Nature Communications, researchers from St. Olaf College and Syracuse University built a computer made entirely of mechanical components that can perform simple computations without electricity or batteries. https://techxplore.com/news/2026-04-mechanical-odd.html Engineering Mon, 13 Apr 2026 13:40:03 EDT news695304905 It wasn't long ago that news headlines claimed that AI might soon assist radiologists in interpreting X-rays of broken bones and analyzing mammograms. We are still far from the destination, as a new study has brought to light the mirage effect, where AI creates detailed descriptions of images that do not exist. https://techxplore.com/news/2026-04-ai-fake-visual-images-dont.html Computer Sciences Sun, 12 Apr 2026 13:00:02 EDT news694946192 There is an exciting future on the horizon—one in which your thoughts could directly control electronic devices you use every day. In many ways, that future is already here, enabled by neural interfaces—engineered devices designed to exchange information with the body's nervous system. From consumer wearables to clinical devices, electronics controlled by neural interfaces are making their way into the marketplace and medical practice. These technologies are demonstrating potential for augmenting, and even restoring, human capabilities in profound ways. https://techxplore.com/news/2026-04-neural-interfaces-game-theory-wearables.html Computer Sciences Sun, 12 Apr 2026 07:00:01 EDT news695187648 Training a large artificial intelligence model is expensive, not just in dollars, but in time, energy, and computational resources. Traditionally, obtaining a smaller, faster model either requires training a massive one first and then trimming it down, or training a small one from scratch and accepting weaker performance. https://techxplore.com/news/2026-04-compression-technique-ai-leaner-faster.html Computer Sciences Thu, 09 Apr 2026 15:20:01 EDT news694960323 A major problem with quantum computers is memory, as the information they contain can be quickly lost. Quantum computers are not yet fully reliable—they are far too unstable. However, all around the world, people are trying to improve them—some of whom are based in Norway. https://techxplore.com/news/2026-04-quantum-memory-problem.html Computer Sciences Sat, 04 Apr 2026 10:00:01 EDT news694354592 Artificial intelligence is increasingly being used to help optimize decision-making in high-stakes settings. For instance, an autonomous system can identify a power distribution strategy that minimizes costs while keeping voltages stable. https://techxplore.com/news/2026-04-ai-method-flags-fairness-autonomous.html Computer Sciences Thu, 02 Apr 2026 12:20:04 EDT news694349461 Although AI is not intentionally biased, it can inherit biases from the data fed into it, learning and repeating them until the system becomes inherently unfair. This is complicated by the problem of identifying where the AI system introduced the bias, as most AI systems display their final decision without showing the steps that made it. Unfair patterns may go unnoticed simply because they are hard to identify. https://techxplore.com/news/2026-04-fair-decisions-fuzzy-ai-fairness.html Computer Sciences Thu, 02 Apr 2026 11:40:03 EDT news694344001 The number of scientific papers is growing so rapidly that scientists are no longer able to keep track of all of them, even in their own research area. Researchers from the Karlsruhe Institute of Technology (KIT), in collaboration with scientific partners, have shown how new research ideas can still be obtained from this wealth of information. Using artificial intelligence (AI), they systematically analyzed materials science publications to identify potential new avenues of research. Their results have been published in Nature Machine Intelligence. https://techxplore.com/news/2026-04-ai-science-papers-trends-years.html Computer Sciences Wed, 01 Apr 2026 15:20:05 EDT news694274532 Nano- and microplastics are increasingly being detected in the human body. However, their detection remains challenging, often relying on invasive techniques and specialized equipment. Researchers at the Institute of Computer Science at the University of Tartu are developing a device that can measure plastic in the human body. Their research is published in the journal Proceedings of the 27th International Workshop on Mobile Computing Systems and Applications. https://techxplore.com/news/2026-03-smartwatch-device-plastic-particles-human.html Computer Sciences Mon, 30 Mar 2026 16:40:04 EDT news694106881 Even with all the recent advances in the ability of large language models (like ChatGPT) to help us think, research, summarize, and learn complex and technical texts, how do they fare in understanding storytelling and literature? These questions around interpretive nuance remain. https://techxplore.com/news/2026-03-ai-literature.html Computer Sciences Fri, 27 Mar 2026 09:00:01 EDT news693820201 Large language models (LLMs), artificial intelligence systems that can process and generate texts in different languages, are now used daily by many people worldwide. As these models can rapidly source information and create convincing content for specific purposes, they are now also used in some professional settings or for gathering legal, medical, or financial information. https://techxplore.com/news/2026-03-highly-ai-agents-deception.html Computer Sciences Sat, 21 Mar 2026 11:00:05 EDT news692979468 Large language models (LLMs) can generate credible but inaccurate responses, so researchers have developed uncertainty quantification methods to check the reliability of predictions. One popular method involves submitting the same prompt multiple times to see if the model generates the same answer. But this method measures self-confidence, and even the most impressive LLM might be confidently wrong. Overconfidence can mislead users about the accuracy of a prediction, which might result in devastating consequences in high-stakes settings like health care or finance. https://techxplore.com/news/2026-03-method-overconfident-large-language.html Computer Sciences Thu, 19 Mar 2026 09:00:03 EDT news693128401 Sheepdogs, bred to control large groups of sheep in open fields, have demonstrated their skills in competitions dating back to the 1870s. In these contests, a handler directs a trained dog with whistle signals to guide a small group of sheep across a field and sometimes split the flock cleanly into two groups. But sheep do not always cooperate. https://techxplore.com/news/2026-03-sheepdogs-reveal-robot-swarms.html Robotics Wed, 18 Mar 2026 08:40:01 EDT news693041882