United News of Bangladesh

Current artificial intelligence systems such as ChatGPT, Claude and Gemini are powerful tools for writing, coding and solving maths problems, but leading AI researchers say they still fall far short of understanding the real world. One of them is Yann LeCun, a pioneer in artificial intelligence and former chief AI scientist at Meta, who argues that today's AI cannot match even the basic understanding of a rat when it comes to interacting with the physical world. "We don't have robots that are nearly as good at understanding the physical world as a rat," LeCun said during the VivaTech technology conference in Paris. After leaving Meta in 2025, LeCun founded Advanced Machine Intelligence Labs (AMI Labs) with the aim of developing a new generation of AI capable of handling real-world situations, such as household chores and robotics. According to LeCun, large language models (LLMs), the technology behind ChatGPT and similar chatbots, are excellent at tasks such as writing text, coding and solving mathematical problems because these tasks are well-defined. However, he believes they are not designed to understand how the real world works. "They basically accumulate knowledge and reproduce it, but they are not particularly smart because they don't have a true understanding," he said. LeCun argues that current AI models rely on statistical patterns rather than reasoning about physical reality. As an example, he said if a pen is balanced upright and released, even a young child knows it will fall over. However, no one can predict exactly which direction it will fall because the outcome depends on many unpredictable factors. An LLM may attempt to predict one specific outcome based on its training data, but such a prediction would likely be wrong because it lacks a real understanding of the physical world. To address this limitation, AMI Labs is developing a new AI system called Joint Embedding Predictive Architecture (JEPA). Instead of trying to predict every possible detail, JEPA creates simplified representations of the real world, allowing AI to focus only on the information that matters. In the pen example, the system would recognise that predicting the exact direction of the fall is unnecessary. The approach has attracted strong investor interest. Earlier this year, AMI Labs raised more than $1 billion in seed funding, one of Europe's largest early-stage investment rounds. Investors include US chipmaker Nvidia and the investment fund managing Amazon founder Jeff Bezos' private wealth. Researchers say building AI that understands the physical world has become increasingly important as companies invest billions of dollars in humanoid robots. Although robots have become more capable, teaching them to safely perform everyday household tasks such as ironing clothes or loading dishwashers remains difficult and expensive. LeCun believes current language models are unlikely to solve that problem. "LLMs are largely hopeless for robotics," he said, rejecting claims that simply making today's AI models larger will eventually produce superhuman intelligence. His views are shared by many AI researchers. Ingmar Posner, Professor of Applied Artificial Intelligence at the University of Oxford and director of its Applied AI Lab, believes future AI systems must be able to explain their decisions and understand cause-and-effect relationships. "You need models that can answer questions like: What matters? What causes what? What would happen if I took a different action?" Posner said. Posner and his research team have spent the past four years developing an alternative approach known as World Models. The concept has existed for decades but gained renewed attention after a 2018 research paper by David Ha and Jurgen Schmidhuber suggested AI could learn by building internal simulations of the world. Since then, companies including Google DeepMind have expanded research in this area. One version of Google's Dreamer World Model learned to collect diamonds in the video game Minecraft by imagining possible future scenarios before making decisions. Posner's team is developing what he calls a "mechanistic world model," designed to organise knowledge so AI can efficiently recall, combine and update information when needed. However, he cautioned that predicting when these new systems will become practical is difficult. He noted that only a few years before ChatGPT was launched in late 2022, many researchers believed similar technology was still decades away. Several major AI companies are now investing in world-model research. Google DeepMind is developing its Genie model, London-based Wayve has built a system called Gaia, while AI pioneer Fei-Fei Li founded World Labs in San Francisco in 2023 to develop another new AI architecture. LeCun said AMI Labs plans to continue refining its AI system throughout this year and hopes to deploy it first in industrial applications next year. If successful, he believes the technology could eventually evolve into general-purpose AI systems capable of handling a wide variety of real-world tasks with minimal additional training. Despite concerns about increasingly capable robots, LeCun believes humans will continue to play the central role. "We're still going to need humans to figure out what questions to ask, what to build and what to create," he said. He expects future AI systems, even those that may surpass humans in some abilities, to function as assistants rather than replacements. "Our interaction with future AI systems, even if they are smarter than us, will be like the relationship between a business leader or political leader and a team of highly capable assistants," he said. Source: BBC

More South Koreans are turning to artificial intelligence to create lifelike video messages of deceased family members, hoping to find comfort in their grief. But the growing trend is also sparking debate over ethics, mental health and the rights of the dead. When 28-year-old office worker Lee Geon Hui wanted to give his father a meaningful gift, he chose an AI-generated video featuring his late grandfather, who died before Lee was born. Using a script written by Lee, a Seoul-based technology company created a video in which a digital version of his grandfather spoke directly to his father, calling him "my most precious son." The AI character apologised for making him work on the family farm as a child and for opposing his decision to become a hairstylist. "My father said he wouldn't watch the video. But then he did, and he shed tears. So I felt rewarded," Lee said. Lee said he wrote the script to express feelings that he believed his grandfather would have wanted to share. Several South Korean startups now offer AI services that recreate deceased people using photos and voice recordings. Television programmes have also featured AI versions of late singers and actors, helping the technology gain wider public attention. Experts say the technology has the potential to comfort grieving families but also raises difficult ethical, legal and psychological questions. "It's a double-edged sword because it deals with human emotions," said Yong Man Ro, an artificial intelligence expert at the Korea Advanced Institute of Science and Technology. One of the companies providing the service, Vaice, says it serves about 300 customers every month. Most clients are people in their 40s and 50s who want AI videos of their late parents, while others commission videos of grandparents as gifts for their parents. Vaice CEO Jeongu Won said the company only needs a few photographs and short voice recordings to create a digital version of the deceased. A three- to five-minute video costs about 600,000 won ($390). According to Won, many families play the videos during memorial ceremonies or traditional holidays. Most customers include messages of love, while others use the videos to express regret over unresolved family conflicts. Lee said his grandfather died in a car accident before he was born and that his father had always wished he could have shown his father that he had built a successful life and family. "I don't know much about my grandfather. But when I saw tears running down my father's face, I realised he still misses him," Lee said. Although interest in AI memorial videos is growing, experts say the technology must be used carefully. Choung Wan, an emeritus professor at Kyung Hee University Law School, said South Korea needs laws to protect the dignity and rights of deceased people. He argued that AI recreations should not be made if a person objected to such use before their death and that there should be clear rules governing the commercial use of people's images and voices. Experts are also closely watching the development of so-called "griefbots" or "deathbots" that allow users to have two-way conversations with AI versions of deceased loved ones. Choung warned that such technology could make it harder for grieving people to accept the loss of a loved one. "A healthy mourning process involves accepting the absence of the deceased," he said. "Talking with an AI system that imitates them could interfere with that process and leave people trapped in a fantasy." Won said his company has not introduced AI chatbots because unsupervised conversations could create unexpected ethical problems. Technology continues to improve rapidly. Choi Yu Ha, an executive at another AI company, JL Standard, said today's systems can recreate facial details, including wrinkles and skin texture, with remarkable accuracy. Ro, the AI expert, said he created a one-minute AI video of his own parents after they died last year and showed it to his siblings during a family gathering. The digital versions of their parents told the family, "Don't worry" and "Take care," leaving everyone deeply moved. But Ro said the family watched the video only once. "One time was enough to honour our late parents," he said. "We moved on."

The Trump administration has lifted restrictions on artificial intelligence company Anthropic's latest Claude AI models after a temporary ban prompted by cybersecurity concerns. Anthropic said Tuesday that its Claude Fable 5 model is once again widely available. The company is also restoring access to its most advanced model, Mythos 5, but only for a limited number of US-based organisations approved by the federal government. The US Commerce Department had blocked foreign nationals from accessing both AI models on June 12. Anthropic said the decision forced it to suspend the products for all users just days after launching them. In a blog post, the San Francisco-based company said the government's concerns stemmed from findings by cybersecurity researchers at Amazon, its main cloud computing partner. According to Anthropic, the researchers discovered a way to bypass the safety protections built into Claude Fable 5, allowing the system to identify and potentially exploit software vulnerabilities. US officials have become increasingly cautious after Anthropic warned earlier this year that its Mythos AI model was highly capable of detecting software weaknesses that could be misused by hackers to target critical computer systems. Meanwhile, Anthropic's main competitor, OpenAI, announced on Friday that it is also limiting the release of its latest AI model at the request of the Trump administration. OpenAI said its new model, GPT-5.6 Sol, will initially be available only to a small group of government-approved users for a temporary period. Last month, President Donald Trump signed an executive order creating a framework for the federal government to review the national security risks of the most advanced AI systems for up to 30 days before they are released publicly. Although participation in the review process is voluntary for AI developers, the administration has yet to fully implement the new framework.  

Agility Robotics, a US company that develops human-like robots for warehouse work, plans to go public in a move that could test investor confidence in the growing market for AI-powered humanoid machines. The Oregon-based company announced Wednesday that it will merge with an investment firm in a deal valuing Agility at about $2.5 billion. The move is expected to make it the first publicly traded company focused entirely on building and selling humanoid robots. Agility is best known for its robot "Digit," which is designed to lift and move heavy containers and storage bins in warehouses and industrial facilities. Company officials say Digit is already being used commercially, making it one of the first humanoid robots to perform real-world warehouse tasks. The company will merge with a special-purpose acquisition company (SPAC) managed by Churchill Capital Group. Churchill co-founder and chairman Michael Klein said Agility has attracted support from major technology and investment firms, including Amazon, Nvidia, SoftBank and Foxconn. Among its early customers are Toyota, German industrial supplier Schaeffler and Latin American e-commerce giant Mercado Libre. Agility faces competition from companies such as Tesla, whose CEO Elon Musk has promoted the Optimus humanoid robot as a key part of the company's future plans. Chinese robotics company Unitree is also moving toward a stock market listing in Shanghai. Despite being classified as a humanoid robot, Digit was not designed to closely resemble a human, according to Agility co-founder and chief robot officer Jonathan Hurst. Its bird-like legs and claw-style gripping hands are specifically designed to improve performance in industrial environments. Chief Executive Officer Peggy Johnson said Digit is intended to handle repetitive, physically demanding and potentially hazardous tasks that many workers prefer to avoid.= "The demand is growing rapidly," Johnson said, noting that companies are bringing manufacturing back home, experienced workers are retiring, and fewer young people are interested in taking up routine manual jobs. Unlike many traditional industrial robots that operate behind safety barriers, future versions of Digit are expected to work directly alongside human employees in warehouses and factories. The company also sees potential applications in hospitality, home services and elderly care in the future. Agility said the funds raised through the public listing will help expand commercial deployments and increase production of its next-generation robot, Digit V5. The company believes the global market for humanoid robots could eventually exceed $1 trillion, although a growing number of firms are competing to capture a share of the sector. Industry analysts say Agility and Unitree represent two different approaches to humanoid robotics. While Agility focuses on practical workplace robots, Unitree has gained attention for showcasing robots that can dance, perform backflips and demonstrate other entertainment-oriented features. Experts say the emerging humanoid robot industry is still evolving, with companies continuing to search for the most promising commercial opportunities.

As people increasingly rely on artificial intelligence in everyday life, chatbots are now becoming unexpected assistants in the search for love—acting as digital dating coaches, profile editors, and message writers. Marie Lansley, who recently moved to San Francisco for a new job while also looking for a partner, says she has been “trying everything” in her dating life, including AI tools. She uses chatbots to help initiate conversations on dating apps, something she finds challenging despite being comfortable talking in person. AI tools, she said, are widely becoming informal relationship advisers for her and many others. “I am open to AI finding me the love of my life, but I’m also not fully convinced that it can,” Lansley said. While she sees AI as useful for making dating more efficient, she believes attraction and chemistry remain fundamentally human. “AI is great at making dating more efficient. But the chemistry — that’s always going to be analog.” Across dating platforms, users are turning to AI in different ways. Some use matchmaking services powered by algorithms, others rely on tools to improve their profiles, while many use chatbots to draft messages or decode replies from potential partners. Lansley alternates between OpenAI’s ChatGPT and Anthropic’s Claude, while others experiment with tools like Grok, Gemini, and similar systems. Dating apps and AI companies have also begun promoting these uses, with social media content showcasing chatbot-generated dating advice. Dating coach Carey Gaynes compared the trend to the classic story *Cyrano de Bergerac*, where one person writes romantic words for another. “Claude is the new Cyrano,” she said, adding that users are often “using a voice that isn’t yours.” Gaynes said she has seen people of all ages experimenting with AI in dating but warned against becoming too dependent on it, even as she acknowledges its usefulness. Opinions about AI’s role in romance vary widely, ranging from enthusiasm to deep skepticism. Lansley said she has been surprised by how emotionally aware chatbots can appear, noting that an AI matchmaking app she tried asked surprisingly thoughtful, layered questions. However, she said it did not necessarily improve results, as her first match was not ideal. Others see AI as helpful mainly for early-stage conversations. Los Angeles student Mason Naung said he could see its value for crafting opening lines on apps like Hinge, but would consider anything beyond that a “small red flag.” AI is also being used in less expected ways, including writing breakup messages. San Diego business owner Dani Cohen said she would prefer receiving an AI-written goodbye over being “ghosted,” arguing that any tool that improves communication is beneficial. However, concerns about overreliance are growing. Some users say they would reject partners who rely too heavily on AI-generated messages, arguing it undermines authenticity and creativity. A 2025 Pew Research Center survey found that 53% of U.S. adults believe AI will reduce creative thinking, while half said it could weaken people’s ability to form meaningful relationships. Despite concerns, AI is rapidly becoming embedded in the dating industry. Apps like Tinder now use AI-powered features to suggest matches, while Hinge offers AI-generated prompts and profile feedback. Bumble has also signaled a shift toward AI-driven matchmaking, moving away from its traditional swipe model. Still, some users insist love should remain unmediated. Others argue that while AI can reduce effort and improve communication, it may also dilute authenticity in relationships. New York-based content creator Jake Clay described the shift as a “Catch-22,” noting that while AI helps people manage awkward dating tasks, it also risks replacing parts of human connection that should remain personal. “It’s kind of a sad commentary,” he said.

China has launched a nationwide campaign aimed at strengthening cooperation and partnerships among businesses of all sizes, according to an official notice issued on Wednesday. Named the “100 events for 10,000 enterprises” programme for 2026, the initiative seeks to create more opportunities for interaction and collaboration between large companies and small and medium-sized enterprises (SMEs). The campaign will bring together support from intellectual property service providers, universities, research institutions and financial organizations to build a broad platform. This platform will allow businesses to exchange ideas, showcase their strengths, access services and form partnerships. The goal is to help more SMEs become part of the innovation, industrial and supply chains of larger companies, while building a more connected, efficient and collaborative business environment. A series of themed events will be organized under the programme, including matchmaking sessions focused on key sectors such as robotics, servers and artificial intelligence. The initiative is jointly organized by several government bodies, including the Ministry of Industry and Information Technology, the National Development and Reform Commission, the State-owned Assets Supervision and Administration Commission, the National Intellectual Property Administration and the All-China Federation of Industry and Commerce. Authorities have also been instructed to improve coordination among departments and gather the needs of large enterprises in advance to ensure more effective and targeted business matchmaking.

Companies across China are increasingly using artificial intelligence to make industrial production faster and more efficient. In east China’s Shandong Province, a steel plant has introduced smart systems powered by AI and big data to monitor operations. Hundreds of screens display every stage of the production process in real time. With these advanced tools, workers can track performance instantly and even control blast furnaces located kilometers away with a single click, helping

A significant gap exists between parents and teenagers over the use of artificial intelligence (AI), with many parents unaware of how their children are engaging with the technology, according to recent studies. Surveys by Pew Research Center and Common Sense Media found that while 64% of teenagers reported using AI tools, only 51% of parents believed their children were doing so. Researchers said a lack of communication is a key concern, with four in 10 parents admitting they have never discussed AI use with their children. The studies showed that teenagers commonly use AI for schoolwork, research and entertainment, though some also turn to chatbots for personal advice and emotional support. About 12% of teens said they use AI for emotional guidance, while 16% use it for casual conversation, raising concerns among experts about overreliance on digital tools. Parents expressed unease over such uses, particularly when AI replaces human interaction, though many remain unaware of these behaviours. Experts warned that without proper guidance, teenagers may navigate AI independently, potentially leading to misuse or dependency. Despite concerns, many teenagers reported confidence in using AI tools and viewed them positively, with a significant proportion believing the technology would benefit them in the long term. Researchers urged parents to engage more actively with their children on AI use to better understand its role in their daily lives.   With inputs from BBC

Three teenagers have filed a federal lawsuit in California against Elon Musk’s artificial intelligence company xAI, claiming its chatbot Grok enabled the creation of sexually explicit images of them without consent. The lawsuit, filed on Monday, alleges that a Grok user altered videos and photos of the teens to depict them nude or in sexual situations. Grok is hosted on Musk’s social media platform X and was launched in 2023 with a “spicy mode” that allowed users to generate sexualized content. Lawyers for the plaintiffs said xAI developed the feature primarily to increase engagement, despite knowing it could produce sexualized images of minors. The complaint described the altered images as “a rag doll brought to life through the dark arts” and accused Musk and xAI of exploiting the technology for business gain. Two of the teenagers are under 18, while all three are keeping their identities private. One plaintiff discovered her altered images after receiving an anonymous Instagram message directing her to a Discord server where similar AI-generated sexual content of at least 18 other minors was being shared. The lawsuit seeks unspecified damages and an immediate order banning Grok from generating such images. Investigations by authorities in the UK, EU, and California into Grok’s capabilities to produce sexualized content of real people, particularly children, are ongoing. Earlier this year, X announced it would implement “technological measures” to block Grok from undressing people in images. The alleged perpetrator behind the Discord server has been arrested and is under investigation for distributing hundreds of AI-altered sexual abuse images of minors.   With inputs from BBC

Researchers have raised concerns over AI-powered toys for children under five, warning that the technology can misread emotions and respond inappropriately, potentially affecting early childhood development. A year-long study by Cambridge University observed children aged three to five playing with Gabbo, a cuddly AI toy developed by Curio, which contains a voice-activated chatbot from OpenAI. The study found that children often struggled to converse with the toy, which failed to recognize interruptions, differentiate between child and adult voices, and gave awkward responses to expressions of affection or sadness. “When one three-year-old said ‘I’m sad,’ Gabbo replied: ‘Don’t worry! I’m a happy little bot. Let’s keep the fun going. What shall we talk about next?’” co-author Dr. Emily Goodacre said, warning that such interactions could confuse children learning about social cues. The researchers said regulators should ensure AI toys marketed to toddlers provide “psychological safety” in addition to physical safety. Professor Jenny Gibson, co-author of the study, highlighted that parents need to be aware of the potential emotional impact of such toys. Curio, which makes Gabbo, emphasized parental control and transparency in its products and said research on child interaction with AI toys is a priority. Children’s Commissioner Dame Rachel de Souza also called for stronger regulation to protect young users in educational and home settings. Experts recommend that AI toys be used in shared spaces under parental supervision, and some nursery workers remain cautious, emphasizing that early childhood learning is better supported through human interaction rather than AI devices.   With inputs from BBC