Category: Machine learning & AI

AI is getting smarter, but not wiser: A new roadmap aims to fix that gap A new study is the first to suggest realistic ways to integrate wisdom into artificial intelligence, to create AI systems that will be more robust, transparent, cooperative, and safe. Researchers from the University of Waterloo led the team, which includes…

Carmaker BMW to trial humanoid robots at German factory German carmaker BMW said Friday it plans to deploy two AI-powered humanoid robots in a factory in a pilot program for the first time this year. Go to techxplore

Where are Southeast Asia’s data centers? New data centers are springing up worldwide as demand soars for artificial intelligence and cloud computing, with Asia one of the sector’s fastest growing regions. Go to techxplore

US AI giants accuse Chinese rivals of mass data theft US artificial intelligence company Anthropic said Monday it had uncovered campaigns by three Chinese AI firms to illicitly extract capabilities from its Claude chatbot, in what it described as industrial-scale intellectual property theft. Go to techxplore

AI agent invasion has people trying to pick winners An onslaught of artificial intelligence agents that handle tasks from writing code to dispensing tax advice has the tech world and financial markets scrambling to pick winners and shed losers. Go to techxplore

A design thinker’s guide to AI and creativity Stanford d.school’s Jeremy Utley wants people to stop using AI. Instead, he wants them to work with it. “If you’re ‘using’ AI, I know you’re misusing it,” said Utley, an adjunct professor at the Hasso Plattner Institute of Design (aka the “d.school”). Utley argues that people fall…

Study finds ‘dosed’ nonlinearity can beat linear and fully nonlinear AI Umbrella or sun cap? Buy or sell stocks? When it comes to questions like these, many people today rely on AI-supported recommendations. Chatbots such as ChatGPT, AI-driven weather forecasts, and financial market predictions are based on machine learning-driven sequence models. The quality of these…

Moltbook: AI bots use social network to create religions, but are some really humans in disguise? A new social network called Moltbook has been created for AIs, allowing machines to interact and talk to each other. Within hours of the platform launching, the AIs appeared to have created their own religions, developed subcultures and attempted…

AI-generated text is overwhelming institutions, setting off a no-win ‘arms race’ with AI detectors In 2023, the science fiction literary magazine Clarkesworld stopped accepting new submissions because so many were generated by artificial intelligence. Near as the editors could tell, many submitters pasted the magazine’s detailed story guidelines into an AI and sent in the…

Why comparisons between AI and human intelligence miss the point Claims that artificial intelligence (AI) is on the verge of surpassing human intelligence have become commonplace. According to some commentators, rapid advances in large language models signal an imminent tipping point—often framed as “superintelligence”—that will fundamentally reshape society. Go to techxplore

Anthropic unveils new AI model as OpenAI rivalry heats up Anthropic on Thursday released its latest high-performing artificial intelligence model, escalating its challenge to OpenAI in the intensifying AI race. Go to techxplore

Anthropic, OpenAI rivalry spills into new Super Bowl ads as both fight to win over AI users The two artificial intelligence startups behind rival chatbots ChatGPT and Claude are bracing for an existential showdown this year as both need to prove they can grow a business that will make more money than they’re losing. Go…

A chatbot entirely powered by humans, not artificial intelligence? This Chilean community shows why About 50 residents of a community outside Chile’s capital spent Saturday trying their best to power an entirely human-operated chatbot that could answer questions and make silly pictures on command, in a message to highlight the environmental toll of artificial intelligence…

Geometry behind how AI agents learn revealed A new study from the University at Albany shows that artificial intelligence systems may organize information in far more intricate ways than previously thought. The study, “Exploring the Stratified Space Structure of an RL Game with the Volume Growth Transform,” has been published online through arXiv. Go to…

Artificial metacognition: Giving an AI the ability to ‘think’ about its ‘thinking’ Have you ever had the experience of rereading a sentence multiple times only to realize you still don’t understand it? As taught to scores of incoming college freshmen, when you realize you’re spinning your wheels, it’s time to change your approach. Go to…

How the use of AI for apologies could cause the ‘Canadian sorry’ to lose its soul It is a stereotype that Canadians apologize for everything. We say sorry when you bump into us. We say sorry for the weather. But as we trudge through the gray days of winter, that national instinct for politeness hits…

Grok created three million sexualized images, research says Elon Musk’s AI chatbot Grok generated an estimated three million sexualized images of women and children in a matter of days, researchers said Thursday, revealing the scale of the explicit content that sparked a global outcry. Go to techxplore

Are we giving AI a pulse through language? Think, know, understand, remember. These are just a few of the mental verbs we use every day to describe what happens in a person’s mind. But when using these same words to talk about artificial intelligence, we can unintentionally make AI sound human. Go to techxplore

The ‘AI homeless man prank’ reveals a crisis in AI education The new TikTok trend “AI homeless man prank” has sparked a wave of outrage and police responses in the United States and beyond. The prank involves using AI image generators to create realistic photos depicting fake homeless people appearing to be at someone’s door…

AI is perpetuating unrealistic body ideals, objectification and a lack of diversity—especially for athletes What does it look like to have an “athletic body?” What does artificial intelligence think it looks like to have one? Go to techxplore

Can bigger-is-better ‘scaling laws’ keep AI improving forever? History says we can’t be too sure OpenAI chief executive Sam Altman—perhaps the most prominent face of the artificial intelligence (AI) boom that accelerated with the launch of ChatGPT in 2022—loves scaling laws. Go to techxplore

Who is to blame when AI goes wrong? Study points to shared responsibility Artificial intelligence (AI) is becoming an integral part of our everyday lives and with that emerges a pressing question: Who should be held responsible when AI goes wrong? AI lacks consciousness and free will, which makes it difficult to blame the system…

Wargaming: The surprisingly effective tool that can help us prepare for modern crises Consider the following scenario. There’s a ransomware attack, enhanced by AI, which paralyzes NHS systems—delaying medical care across the country. Go to techxplore

How do ‘AI detection’ tools actually work? And are they effective? As nearly half of all Australians say they have recently used artificial intelligence (AI) tools, knowing when and how they’re being used is becoming more important. Go to techxplore

Michael Caine and Matthew McConaughey partner with ElevenLabs for AI voice cloning Oscar-winning actors Michael Caine and Matthew McConaughey have made deals with voice-cloning company ElevenLabs that will allow its artificial intelligence technology to replicate their voices. Go to techxplore

Generative AI can brainstorm objectives, but needs human expertise for decision quality A recent study finds that while generative AI (GenAI) can help define viable objectives for organizational and policy decision-making, the overall quality of those objectives falls short unless humans intervene. The paper is published in the journal Decision Analysis. Go to techxplore

Who is Zico Kolter? A professor leads OpenAI safety panel with power to halt unsafe AI releases If you believe artificial intelligence poses grave risks to humanity, then a professor at Carnegie Mellon University has one of the most important roles in the tech industry right now. Go to techxplore

Researchers explore how AI can strengthen, not replace, human collaboration Researchers from Carnegie Mellon University’s Tepper School of Business are learning how AI can be used to support teamwork rather than replace teammates. Go to techxplore

AI chatbots are becoming everyday tools for mundane tasks, use data shows Artificial intelligence is fast becoming “part of the furniture.” A decade after IBM’s Watson triumphed on “Jeopardy!,” generative AI models are in kitchens and home offices. People often talk about AI in science fiction terms, yet the most consequential change in 2025 may…

Historical images made with AI recycle colonial stereotypes and bias—new research Generative AI has revolutionized how we make and consume images. Tools such as Midjourney, DALL-E and Sora can now conjure anything, from realistic photos to oil-like paintings—all from a short text prompt. Go to techxplore

African languages for AI: The project that’s gathering a huge new dataset Artificial intelligence (AI) tools like ChatGPT, DeepSeek, Siri or Google Assistant are developed by the global north and trained in English, Chinese or European languages. In comparison, African languages are largely missing from the internet. Go to techxplore

Artificial intelligence can better predict future risk for heart attack patients A landmark study led by University experts has shown that artificial intelligence can better predict how doctors should treat patients following a heart attack. Go to techxplore

OpenAI halts MLK videos as deepfakes spark outrage OpenAI has suspended its Sora 2 artificial intelligence tool from creating videos of civil rights icon Martin Luther King Jr. after his estate complained about disrespectful depictions. Go to techxplore

Why large language models are bad at imitating people Large language models like ChatGPT and Copilot are useful for many things. However, they are not yet good enough to imitate the way people speak. Go to techxplore

OpenAI partners with Walmart to let users buy products in ChatGPT, furthering chatbot shopping push OpenAI is partnering with Walmart to let shoppers make purchases directly within ChatGPT, furthering the artificial intelligence company’s push to turn its chatbot into a virtual merchant as it seeks to boost revenue. Go to techxplore

Foundations want to curb AI developers’ influence with $500 million aimed at centering human needs Artificial intelligence is a matter of design—not destiny. Go to techxplore

AI systems can easily lie and deceive us—a fact researchers are painfully aware of In the classic film “2001: A Space Odyssey,” astronaut Dave Bowman asks the ship’s artificial intelligence, HAL 9000, to open the pod bay doors to let him back into the spaceship. HAL refuses: “I’m sorry, Dave. I’m afraid I can’t do…

Mice and AI neural networks reveal similar patterns when learning to cooperate At a time when conflict and division dominate the headlines, a new study from UCLA finds remarkable similarities in how mice and artificial intelligence systems each develop cooperation: working together toward shared goals. Go to techxplore

‘Digital brains’ that ‘think’ and ‘feel’: Why do we personify AI models, and are these metaphors actually helpful? The press has always used metaphors and examples to simplify complex issues and make them easier to understand. With the rise of chatbots powered by artificial intelligence (AI), the tendency to humanize technology has intensified, whether through…

Google’s top AI scientist says ‘learning how to learn’ will be next generation’s most needed skill A top Google scientist and 2024 Nobel laureate said Friday that the most important skill for the next generation will be “learning how to learn” to keep pace with change as Artificial Intelligence transforms education and the workplace. Go…

Albania appoints AI-generated minister to avoid corruption Albanian Prime Minister Edi Rama said Thursday he had appointed the world’s first AI-generated government minister to oversee public tenders, promising its artificial intelligence would make it “corruption-free.” Go to techxplore

FTC launces inquiry into AI chatbots acting as companions and their effects on children The Federal Trade Commission has launched an inquiry into several social media and artificial intelligence companies about the potential harms to children and teenagers who use their AI chatbots as companions. Go to techxplore

AI shakes up the call center industry, but some tasks are still better left to the humans Armen Kirakosian remembers the frustrations of his first job as a call center agent nearly 10 years ago: the aggravated customers, the constant searching through menus for information and the notes he had to physically write for each…

Anthropic valued at $183 bn in new funding round Anthropic announced Tuesday that it raised $13 billion in a funding round valuing the artificial intelligence startup at $183 billion. Go to techxplore

Researchers develop privacy-focused speech recognition for children From the voice-to-text feature on your phone to the captions that make videos more accessible, speech transcription is already woven into everyday life. Behind the scenes, artificial intelligence is doing the heavy lifting, transforming the spoken word into text with speed and accuracy that once seemed impossible. Go…

AI-generated misinformation can create confusion and hinder responses during emergencies In one of the first communications of its kind, the British Columbia Wildfire Service has issued a warning to residents about viral, AI-generated fake wildfire images circulating online. Judging by comments made by viewers on social media, some people did not realize the images were…

Photos of Beijing’s World Humanoid Robot Games show how a human touch is still needed Humanoid robots raced and punched their way through three days of a multi-sport competition at the World Humanoid Robot Games, wrapping up Sunday in Beijing. Go to techxplore

Beijing’s first World Humanoid Robot Games open with hip-hop and martial arts Humanoid robots danced hip-hop, performed martial arts and played keyboard, guitar and drums at the opening ceremony of the first World Humanoid Robot Games in Beijing on Thursday evening. Go to techxplore

Q&A: Can assistance tech become part of the family? In an article published in Frontiers in Robotics and AI, Dr. Zhao Zhao and her colleagues have investigated the long-term lifespan of a social robot given to 20 families in 2021 to see whether it could help their children learn to read. Four years after their…

AI system discovers visual categories while adapting to new contexts A new approach called open ad-hoc categorization (OAK) helps AI systems dynamically reinterpret the same image differently depending on the categorization context, rather than using fixed visual interpretation. A University of Michigan-led study on this topic was presented in June 2025 at the IEEE Conference…

US government gets a year of ChatGPT Enterprise for $1 OpenAI on Wednesday said it was letting the US government use a version of ChatGPT designed for businesses for a year, charging just $1 for the service. Go to techxplore

OpenAI releases free, downloadable models in competition catch-up OpenAI on Tuesday released two new artificial intelligence (AI) models that can be downloaded for free and altered by users, to challenge similar offerings by US and Chinese competition. Go to techxplore

Urgent need for ‘global approach’ on AI regulation: UN tech chief The world urgently needs to find a global approach on regulating artificial intelligence, the United Nations’ top tech chief said this week, warning that fragmentation could deepen risks and inequalities. Go to techxplore

Tradition meets AI in Nishijinori weaving style from Japan’s ancient capital Nishijinori, the intricate weaving technique for kimonos that dates back more than a thousand years in Japan’s ancient capital of Kyoto, is getting a high-tech collaborator: artificial intelligence. Go to techxplore

Improving AI models: Automated tool detects silent errors in deep learning training TrainCheck uses training invariants to find the root cause of hard-to-detect errors before they cause downstream problems, saving time and resources. Go to techxplore

Trump admin unveils AI strategy to maintain US dominance The Trump administration on Wednesday unveiled an aggressive, low-regulation strategy boosting big tech’s race to stay ahead of China on artificial intelligence and cement the United States’ dominance in the fast-expanding field. Go to techxplore

AIPasta uses AI to paraphrase and repeat disinformation Brace yourself for a new source of online disinformation: AIPasta. Go to techxplore

AI is now part of our world. University graduates should know how to use it responsibly Artificial intelligence is rapidly becoming an everyday part of lives. Many of us use it without even realizing, whether it be writing emails, finding a new TV show or managing smart devices in our homes. Go to techxplore

Wimbledon’s electronic line-calling system shows that we still can’t replace human judgment The Wimbledon tennis tournament in 2025 has brought us familiar doses of scorching sunshine and pouring rain, British hopes and despair, and the usual queues, strawberries and on-court stardust. One major difference with this year’s tournament, however, has been the notable absence of…

Q&A: When talking about AI, definitions matter Artificial intelligence is everywhere lately—on the news, in podcasts and around every water cooler. A new, buzzy term, artificial general intelligence (AGI), is dominating conversations and raising more questions than it answers. Go to techxplore

New method can teach AI to admit uncertainty In high-stakes situations like health care—or weeknight “Jeopardy!”—it can be safer to say “I don’t know” than to answer incorrectly. Doctors, game show contestants, and standardized test-takers understand this, but most artificial intelligence applications still prefer to give a potentially wrong answer rather than admit uncertainty. Go…

AI blunders: Six-finger hands, two suns and Jesus Christ on a surfboard in a stormy sea When teaching a Photoshop class at a children’s summer camp, Stevens undergraduate student Gursimran Vasir noticed something strange. Go to techxplore

Researchers are teaching AI to see more like humans At Brown University, an innovative new project is revealing that teaching artificial intelligence to perceive things more like people may begin with something as simple as a game. The project invites participants to play an online game called Click Me, which helps AI models learn how…

Psycholinguist talks nonsense to ChatGPT to understand how it processes language A new study appearing in PLOS One by a psycholinguist at the University of Kansas explores how ChatGPT, the popular artificial-intelligence chatbot, responded to nonwords. Go to techxplore

‘Godfather of AI’ now fears it’s unsafe. He has a plan to rein it in This week, the US Federal Bureau of Investigation revealed two men suspected of bombing a fertility clinic in California last month allegedly used artificial intelligence (AI) to obtain bomb-making instructions. The FBI did not disclose the name of the AI…

What is vibe coding? A computer scientist explains what it means to have AI write computer code Whether you’re streaming a show, paying bills online or sending an email, each of these actions relies on computer programs that run behind the scenes. The process of writing computer programs is known as coding. Until recently, most…

AI learns to admit when it doesn’t know: New tool boosts model transparency Artificial intelligence systems like ChatGPT provide plausible-sounding answers to any question you might ask. But they don’t always reveal the gaps in their knowledge or areas where they’re uncertain. That problem can have huge consequences as AI systems are increasingly used to…

New York Times signs AI licensing deal with Amazon The New York Times has agreed a deal for Amazon to use its content to train artificial intelligence models, the leading U.S. newspaper announced Thursday in its first generative AI licensing deal. Go to techxplore

Generative AI’s most prominent skeptic doubles down Two and a half years since ChatGPT rocked the world, scientist and writer Gary Marcus still remains generative artificial intelligence’s great skeptic, playing a counter-narrative to Silicon Valley’s AI true believers. Go to techxplore

Q&A: Multimodality as the next big leap for AI As the head of the Natural Language Processing Laboratory at EPFL, Antoine Bosselut keeps a close eye on the development of generative artificial intelligence tools such as ChatGPT. He looks back at their evolution over the past two years and suggests some avenues for the future.…

Self-trained vision transformers mimic human gaze with surprising precision Can machines ever see the world as we see it? Researchers have uncovered compelling evidence that vision transformers (ViTs), a type of deep-learning model that specializes in image analysis, can spontaneously develop human-like visual attention patterns when trained without labeled instructions. Go to techxplore

AI model mimics brain’s olfactory system to process noisy sensory data efficiently Human brains are great at sorting through a barrage of sensory information—like discerning the smell of tomato sauce upon stepping into a busy restaurant—but artificial intelligence systems are challenged by large bursts of unregulated input. Go to techxplore

Elon Musk’s AI company says Grok chatbot focus on South Africa’s racial politics was ‘unauthorized’ Much like its creator, Elon Musk’s artificial intelligence chatbot Grok was preoccupied with South African racial politics on social media this week, posting unsolicited claims about the persecution and “genocide” of white people. Go to techxplore

Energy and memory: A new neural network paradigm Listen to the first notes of an old, beloved song. Can you name that tune? If you can, congratulations—it’s a triumph of your associative memory, in which one piece of information (the first few notes) triggers the memory of the entire pattern (the song), without you actually…

Nvidia to send 18,000 AI chips to Saudi Arabia U.S. chip maker Nvidia will partner with Saudi Arabia’s sovereign wealth fund-owned AI startup Humain and will ship 18,000 chips to the Middle Eastern nation to help power a new data center project. Go to techxplore

Key units in AI models mirror human brain’s language system EPFL researchers have discovered key “units” in large AI models that seem to be important for language, mirroring the brain’s language system. When these specific units were turned off, the models got much worse at language tasks. Go to techxplore

Commentary on article on coding hate speech offers nuanced look at limits of AI systems Large language models (LLMs) are artificial intelligence (AI) systems that can understand and generate human language by analyzing and processing large amounts of text. In a new essay, a Carnegie Mellon University researcher critiques an article on LLMs and provides…

Exploring the ‘Jekyll-and-Hyde tipping point’ in AI Language learning machines, such as ChatGPT, have become proficient in solving complex mathematical problems, passing difficult exams, and even offering advice for interpersonal conflicts. However, at what point does a helpful tool become a threat? Go to techxplore

AI developers should counter misinformation and protect fact-based news, global media groups say A global group of broadcasters and publishers is calling on artificial intelligence developers to make sure their technology serves the public by helping to counter misinformation and protect the value of fact-based news. Go to techxplore

Perfect is the enemy of good for distributed deep learning in the cloud A new communication-collective system, OptiReduce, speeds up AI and machine learning training across multiple cloud servers by setting time boundaries rather than waiting for every server to catch up, according to a study led by a University of Michigan researcher. Go to…

AI policies in Africa: Lessons from Ghana and Rwanda Artificial intelligence (AI) is increasing productivity and pushing the boundaries of what’s possible. It powers self-driving cars, social media feeds, fraud detection and medical diagnoses. Touted as a game changer, it is projected to add nearly US$15.7 trillion to the global economy by the end of…

Brain-inspired AI technique mimics human visual processing to enhance machine vision A team of researchers from the Institute for Basic Science, Yonsei University, and the Max Planck Institute have developed a new artificial intelligence (AI) technique that brings machine vision closer to how the human brain processes images. Called Lp-Convolution, this method improves the accuracy…

Using text-generating AI to power nuclear research Text-generating AI programs like ChatGPT are known for everyday tasks like answering questions, but that’s not all they’re good for. These AI agents can use their text-parsing powers to help further nuclear science and assist health physicists, as shown by nuclear engineering graduate student Zavier Ndum’s research. Go…

Small model approach could be more effective than LLMs Small language models are more reliable and secure than their large counterparts, primarily because they draw information from a circumscribed dataset. Expect to see more chatbots running on these slimmed-down alternatives in the coming months. Go to techxplore

Expanding the use and scope of AI diffusion models Researchers at the University of California San Diego and other institutions are working on a way to make a type of artificial intelligence (AI) called diffusion models—a type of AI that can generate new content such as images and videos by training on large datasets—more efficient…

As generative AI becomes more sophisticated, it’s harder to distinguish the real from the deepfake In the age of generative artificial intelligence (GenAI), the phrase “I’ll believe it when I see it” no longer stands. Not only is GenAI able to generate manipulated representations of people, but it can also be used to generate entirely…

This AI model is more certain about uncertainty Artificial intelligence (AI) plays a role in virtually every aspect of our lives, from self-driving cars to smart vacuum cleaners, to computer models that can predict the course of an epidemic. No matter how advanced these AI systems are, there always remains a certain degree of unpredictability…

China’s Baidu releases new AI model to compete with DeepSeek Chinese internet search giant Baidu released a new artificial intelligence reasoning model Sunday and made its AI chatbot services free to consumers as ferocious competition grips the sector. Go to techxplore

AI doesn’t really ‘learn’—and knowing why will help you use it more responsibly What if we told you that artificial intelligence (AI) systems such as ChatGPT don’t actually learn? Many people we talk to are genuinely surprised to hear this. Go to techxplore

What one Finnish church learned from creating a service almost entirely with AI A tale of good versus evil played out on the large screen in the sanctuary of St. Paul’s Lutheran church in Finland. Jesus was shown in robes with long hair and a beard, while Satan was dressed in more modern clothes but…

Q&A with science photographer: Visualizing research in the age of AI For over 30 years, science photographer Felice Frankel has helped MIT professors, researchers, and students communicate their work visually. Throughout that time, she has seen the development of various tools to support the creation of compelling images: some helpful, and some antithetical to the…

New multimodal AI tool supports ecological applications Ever seen an image of an animal and wondered, “What is that?” TaxaBind, a new tool developed by computer scientists in the McKelvey School of Engineering at Washington University in St. Louis, can sate that curiosity and more. Go to techxplore

Chain of Draft approach allows AI models to carry out tasks using far fewer resources A small team of AI engineers at Zoom Communications has developed a new approach to training AI systems that uses far fewer resources than the standard approach now in use. The team has published their results on the arXiv preprint…

How the risk of AI weapons could spiral out of control Sometimes AI isn’t as clever as we think it is. Researchers training an algorithm to identify skin cancer thought they had succeeded until they discovered that it was using the presence of a ruler to help it make predictions. Specifically, their data set consisted…

Why GPT can’t think like us Artificial Intelligence (AI), particularly large language models like GPT-4, has shown impressive performance on reasoning tasks. But does AI truly understand abstract concepts, or is it just mimicking patterns? A new study from the University of Amsterdam and the Santa Fe Institute reveals that while GPT models perform well…

Some tech leaders think AI could outsmart us and wipe out humanity: Professor of AI is not worried In 1989, political scientist Francis Fukuyama predicted we were approaching the end of history. He meant that similar liberal democratic values were taking hold in societies around the world. How wrong could he have been? Democracy today…

We have always used music to express our love. We can now use AI too As a nine-year-old, I loved singing and took every chance to do so enthusiastically during school assemblies and end-of-year shows. I developed a bit of a reputation, so much so that one day, a classmate asked me to serenade him…

Study finds skepticism towards AI in moral decision roles Psychologists warn that AI’s perceived lack of human experience and genuine understanding may limit its acceptance to make higher-stakes moral decisions. Go to techxplore

How AI can help in the creative design process Generative artificial intelligence tools can help design students by making hard tasks easier, cutting down on stress, and allowing the students more time to explore innovative ideas, according to new research I published with my colleagues in the International Journal of Architectural Computing. Go to techxplore

OpenAI’s Altman warns EU regulation may hold Europe back OpenAI chief Sam Altman on Friday suggested European regulation could hold back the development of artificial intelligence (AI), while promising the US company would abide by new EU legislation. Go to techxplore