While generative AI may accelerate workplace productivity, researchers warn it could be acting as a barrier to deep, critical thinking. Experts suggest that a long-term dependency on these tools may result in a diminished capacity for independent problem-solving. Using AI chatbots to answer complex questions, analyze large datasets, or refine job cover letters has become commonplace. However, some specialists fear that outsourcing these cognitive tasks to AI means the human brain is being underutilized, potentially leading to a decline in intellectual agility. Insights from MIT: A Decline in Cognitive Engagement Earlier this year, the Massachusetts Institute of Technology (MIT) published a study involving 54 students from MIT and surrounding universities. Read More Russian Economy Under Strain but Putin Remains Defiant: CNN Analysis Researchers used electrodes and EEG (electroencephalography) to monitor brain activity while participants performed writing tasks. The findings were striking: Lower Brain Activity: Participants who used ChatGPT to write essays showed lower cognitive engagement compared to those who did not. Memory Gaps: AI users struggled to cite specific parts of their own work, a task that non-users performed easily. Skill Erosion: Researchers noted that this indicates a potential decline in learning efficiency, as users relied on AI for summarizing questions, finding sources, and correcting grammar. While participants used AI for brainstorming, many noted that the technology was less effective at generating truly original ideas than humans. Workplace Dependency and Student Concerns Separate research conducted by Carnegie Mellon University and Microsoft, involving 319 office workers across 900 different tasks, reached a similar conclusion. The study found that individuals with high levels of trust in AI tools tended to apply less of their own critical thinking, especially in data analysis and rule verification. Academic institutions are seeing a similar trend: The OUP Survey: An October survey by Oxford University Press (OUP) among UK school students revealed that 6 out of 10 students believe AI has negatively impacted their study skills. The Counter-Argument: Despite the concerns, Dr. Alexandra Tomescu, a generative AI expert at OUP, noted that 9 out of 10 students reported that AI helped them improve at least one skill, such as creativity or revision. The Road Ahead OpenAI CEO Sam Altman recently reported that ChatGPT now has over 800 million weekly users. To assist students, the platform even released a curated list of 100 prompts. However, many academics remain cautious. Professor Wayne Holmes of University College London (UCL) argues that simply providing prompts is not enough. He maintains that more rigorous research into the long-term cognitive impacts of AI is essential before the technology is fully integrated into educational and professional standards.

Shares in China's MetaX, which makes high-end chips used to power AI systems, rocketed more than 750 percent on its debut in Shanghai on Wednesday, fuelled by hopes of the country's potential to challenge US powerhouse Nvidia. The rally in MetaX Integrated Circuits Shanghai came less than two weeks after another Chinese chipmaker, Moore Threads, surged 425 percent following a blockbuster US$1.1 billion initial public offering. Both firms make the advanced chips known as graphics processing units (GPUs) which run and train artificial intelligence tools. The GPU sector is dominated by Nvidia, now the world's most valuable company thanks to frenzied global demand and optimism for AI. MetaX and Moore Threads are much smaller firms but analysts say investors are buying into the prospect of China ramping up its chip industry to eventually rival the US giant. Shares in MetaX -- which had been priced at 104.66 yuan (US$14.86) in an IPO that raised US$585.8 million -- soared as much as 755 percent to 895.00 yuan, before dropping slightly to 730.34 yuan. China and the United States are competing for dominance in AI, and Beijing has reportedly been encouraging tech companies to use homegrown chips owing to Washington's export restrictions on top-end Nvidia models. However, US President Donald Trump last week said he had agreed with Chinese counterpart Xi Jinping to allow Nvidia to export its advanced H200 chips to the country. H200s are roughly 18 months behind the US company's most state-of-the-art offerings, which will still be off-limits to China. But the move marked a shift away from export curbs imposed by Trump's predecessor Joe Biden. Those restrictions were first implemented in 2022 over national security concerns such as the Chinese military gaining a technological advantage over the United States. Rui Ma, founder of the Tech Buzz China newsletter, told AFP this month that "export controls have created a protected high-end segment" for chips in China. "Combined with policy support and much deeper capital pools, domestic GPU and AI accelerator companies now have far more favourable conditions than in prior cycles," she said. "As a result, firms such as Moore Threads, Cambricon, MetaX, Biren, and Enflame are drawing significant investor interest."

Humanoid robots are back in the spotlight at a Silicon Valley summit, fueled by recent advances in artificial intelligence, despite lingering doubts about their high cost and complexity. Once considered an unattractive investment, the ambition to create machines that move and work like humans has been revived by AI developments.  This renewed interest was evident at the Humanoids Summit in Mountain View, where over 2,000 attendees, including engineers from Disney, Google, and various startups, gathered to showcase emerging technologies and strategize on speeding up development. Summit founder and venture capitalist Modar Alaoui noted that many researchers now believe humanoid robots or other physical forms of AI, could eventually become commonplace, although a firm timeline remains uncertain. Read More Data Centers: A View from the Inside Despite the enthusiasm, widespread skepticism remains among experts. They caution that major technical hurdles must be overcome before robots can reliably function as workers in homes or offices. Cosima du Pasquier, founder of Haptica Robotics, specifically pointed to significant research gaps that still need addressing, particularly in areas like dexterity and touch. Globally, China currently leads the sector, supported by government incentives and a national goal to establish a humanoid robotics ecosystem by 2025, according to McKinsey & Company.  While Chinese-made robots dominated displays at the summit, US firms are leveraging generative AI advancements to improve how their robots understand and navigate complex environments. Veteran roboticists warn that fully capable humanoid robots are still a distant reality, raising questions about whether current investments will truly yield the promised breakthroughs. Source: AP 

The expansion of data centers to power the AI boom has more people wondering: what exactly is in a data center? AFP got a chance to take a look at what is inside. Concrete warehouse Data centers are the physical infrastructure that make our digital lives possible, yet most people have never seen one up close or understand how they operate. Roughly 12,000 data centers are in operation in the world, with about half in the US, according to Cloudscene, a data center directory. At its most basic, a data center is a concrete warehouse filled with thousands of computer servers working in tandem. Traditional facilities span one or two floors divided into vast rooms, though newer ones rise higher. A facility may serve a single company or be shared by several clients. The servers sit in standardized 19-inch (48 cm) racks -- essentially metal closets lined up in rows. A large data center can house tens of thousands of servers running simultaneously, generating enormous heat and consuming significant energy for both power and cooling. High-speed networking equipment -- switches, routers, and fiber optic cables -- connects everything, moving terabytes of data per second. Stay close Having a data center close to end users improves speed, which is critical for things like trading and gaming where immediacy is paramount. Ashburn, Virginia, which has the highest concentration of data centers in the world, offers ideal conditions as it is located only about 30 miles from the US capital, Washington. However, building in densely populated areas costs more and faces local resistance. Companies increasingly turn to rural locations where land is cheaper and zoning less restrictive. But distance adds to loading times -- that brief delay when a page loads or a feed refreshes. To balance cost and performance, operators typically house core infrastructure -- or the training of AI models -- in affordable rural regions while keeping equipment that handles time-sensitive requests closer to urban centers. Stay Cool Inside these bunker-like buildings, a single server rack generates as much heat as several household ovens running nonstop. Cooling consumes roughly 40 percent of a data center's total energy. The most advanced chips -- GPUs (graphics processing units) used for AI -- can reach temperatures exceeding 90øC, threatening performance and causing permanent damage during extended operation. They are also much heavier than lower performing chips. Traditional facilities use computer room air conditioners with heat blasting out of mounted vents on on rooftops - but this is not fit for GPUs that mainly turn to water for cooling. Modern facilities are beginning to deploy "free cooling" that uses outside air when temperatures allow, and different water-based approaches: liquid cooling systems that pump coolant directly to components or evaporative cooling that works like perspiration on skin. Today massive amounts of water are still required for direct and indirect cooling in data centers. In 2014, US data centers used 21.2 billion liters of water, and that number rose to 66 billion liters in 2023, according to federal estimates. Where's the power? Power supply -- and the high voltage transmission lines needed to source it -- is key for a data center and is only growing with facilities that run the powerful GPUs. "One of the biggest challenges for a lot of our customers is they buy the chips and then they don't know where to go," Chris Sharp, Chief Technology Officer at Digital Realty, which operates data centers around the world, told AFP. The big tech giants, caught up in the AI arms race, have spent tens of billions of dollars in just months towards building suitable structures for GPUs. Operators rely on the existing power grid but are increasingly seeking to secure their own resources -- called "behind-the-meter" -- for greater security and to limit rate increases for all users. Solar panels or gas turbines are sometimes installed, and many are also awaiting the arrival of the first small modular reactors (SMRs), a nuclear energy technology currently under development. Most data centers have to run 24/7 and every critical system has backups in case of power outages. This can come through massive battery banks or diesel generators. The best facilities guarantee power 99.995 percent of the time.

A new AI model can predict a person's risk of developing breast cancer up to five years in advance by analyzing mammograms. This could make screening much more effective. Every year, around 2.3 million new cases of breast cancer are diagnosed worldwide. The disease "caused an estimated 670,000 deaths globally in 2022," according to the World Health Organization (WHO). "Breast cancer is the most common cause of cancer death in women — despite mammography screening," said Christiane Kuhl, the director of the Department of Diagnostic and Interventional Radiology at RWTH Aachen University Hospital. "The explanation is that mammography still fails to detect many cases of breast cancer, or is not detecting them early enough," she said, explaining that fast-growing aggressive tumors in particular were often not picked up and these were precisely those that killed many women. Now there is a new algorithm that could revolutionize breast cancer screening. The AI (artificial intelligence) model can very accurately classify a person's risk of developing the disease within the next five years solely by analyzing mammogram images. In one study, women whom the algorithm identified as having a high risk of developing the disease were significantly more likely to develop breast cancer than those identified as having a "normal risk," Kuhl said. "To be specific, these women were four times more likely to develop breast cancer than those whose AI score was low," she explained. "With this newly-developed AI model, we can predict with much greater precision that a woman will develop breast cancer in the next five years — on the basis of mammograms that are normal and show no signs of breast cancer." Mammograms every two years after 50 Currently, all women in Germany aged 50 to 75 are offered breast cancer screening mammograms every two years. But the risk of developing the disease and thus the need for effective early detection varies considerably from one woman to another. Kuhl believes that the "one-size-fits-all" principle is outdated and advocates for tailor-made breast cancer screening, considering the variations in mammography accuracy. The denser the breast tissue, the higher the risk of disease — and the less accurate mammography's prediction of this risk. But many women are unaware of this, said Kuhl. In the US it has long been a requirement that women be informed of glandular tissue density in their breasts and the "masking risk" that means breast cancer might not be detected by mammography. MRI technology is more reliable than mammography For several years now, it has been recommended that women with very dense breast tissue have an MRI (magnetic resonance imaging) which is better at detecting breast cancer at an early stage. MRI technology is a medical imaging technique that uses strong magnetic fields and radio waves, but not X-rays, to produce very detailed cross-sectional images of the body. Although it is very reliable, it is much more expensive than mammography or ultrasound technology, which are both less dependable. In order to identify women who need an MRI for early detection, the Clairity Consortium — an international association of 46 research institutions in North and South America and Germany — has developed an AI model called "Clairity Breast." The algorithm was trained on more than 420,000 mammograms from Europe, South and North America. Unlike traditional risk models for breast cancer, the AI model does not need any data about family history, genetics, or lifestyle. It calculates the probability of developing breast cancer on the basis of mammograms and classifies women into risk categories using established thresholds. The AI model cannot only detect how much glandular tissue there is but also its texture and how it is arranged, which is another breast cancer risk factor. "Only about 10% of women have such extremely dense glandular tissue," said Kuhl. "The vast majority of women who develop breast cancer and receive a late diagnosis have less dense tissue." What is important about the new technology, in her view, is that the "AI model can decide within seconds whether a woman needs an MRI for early detection or not." Should comprehensive breast cancer screening start earlier? Comprehensive breast cancer screening starts at the age of 50 in most countries because the risk of developing the disease increases significantly with age and the benefits of doing this are proven. But Kuhl said that provided the AI model worked, "younger women would benefit from early detection" as though they are less likely to develop breast cancer than older women, if they do they are more likely to develop aggressive tumors. Moreover, mammography is particularly problematic for younger women, she said. "Young women are more likely to have dense glandular tissue, which makes early detection using mammography particularly difficult." But generally Kuhl did not think it would make sense to lower the general screening age. "If we simply lower the age at which women are invited for a screening, we will not alter the fundamental problem." She was in favor of a two-step approach: " First, mammography for early detection; then an AI analysis to determine the risk of disease over the next five years." If the algorithm detected a particularly high risk, then an MRI would have to be offered. "A mammogram is no longer necessary for these women," Kuhl said.