Board-certified cardiothoracic surgeon Dr. Jeremy London, based in Savannah, Georgia, explains why VO2 max and muscle mass are the main indicators of longevity. The first artificial intelligence (AI) stethoscope has gone beyond listening to a heartbeat. Researchers at Imperial College London and Imperial College Healthcare NHS Trust discovered that an AI stethoscope can detect heart failure at an early stage. The TRICORDER study results, published in BMJ Journals, found that the AI-enabled stethoscope can help doctors identify three heart conditions in just 15 seconds. According to the British Heart Foundation (BHF), which partially funded the study, the researchers analyzed data from more than 1.5 million patients, focusing on people with heart failure symptoms like breathlessness, swelling and fatigue.

The chief executive of the UK's leading artificial intelligence agency is stepping down after a staff revolt and government calls for a strategic overhaul. Jean Innes has led the Alan Turing Institute since 2023, but her position has come under pressure amid widespread discontent within the organisation and a demand from its biggest funder, the UK government, for a change in direction. ATI said the search was already under way for a replacement for Innes, who held senior roles in the civil service and technology industry before her appointment. Government sources pointed to a letter sent by the technology secretary, Peter Kyle, to ATI's chair in July that demanded strategic change and indicated a need for new leadership. In the letter, Kyle said the institute should switch its focus to defence and national security and urged "careful consideration" on having an appropriate executive team in place for such a move.

Ukraine has pounded Russia's refineries with deep strikes in the past week, worsening its petrol shortages and causing Moscow to extend a ban on exports of petroleum products. Russia responded with a deadly attack on Kyiv and a barrage of statements portraying Moscow's "special military operation" as a success and Ukraine as teetering on the edge of surrender. Russian President Vladimir Putin used his appearance at the Shanghai Cooperation Organisation (SCO) summit to suggest Russian energy exports to China and India were booming, but reports suggested that Moscow is heavily discounting its crude to hold onto clients. Ukraine's European and regional allies are meeting on Thursday to try to finalise security guarantees in case a ceasefire should come about, while calling on US President Donald Trump to use sanctions to press Putin into direct negotiations with Kyiv. Russia redeployed marines and paratroopers – elite units – from Ukraine's northern Sumy region to the eastern region of Donetsk on September 1, suggesting it may be preparing a renewed push for the city of Pokrovsk, which Ukraine has identified as a Russian key tactical objective in the east since August 2024.

Fox News senior medical analyst Dr. Marc Siegel shares his perspective on whether the mosquito-borne virus in China will spread to the United States and how AI can be detrimental to children's and young adults' mental health on'Fox Report.' A little-known disease is spreading in the U.S., primarily in the state of California, health officials warn. In a new study published in the CDC journal Emerging Infectious Diseases, researchers state that human cases of Chagas disease have been confirmed in eight states, leading them to recommend that the disease is classified as "endemic." "Acknowledging the endemicity of Chagas disease in the United States is crucial for achieving global health goals," the authors wrote. The Centers for Disease Control and Prevention defines a disease as "endemic" when there is a "constant presence and/or usual prevalence" in a population within a specific geographic area -- in other words, the "baseline" level of disease within a community.

Trump to host Big Tech CEOs in White House's redone rose garden U.S. President Donald Trump and his wife, Melania, are hosting an artificial intelligence event for tech CEOs on Thursday at the newly renovated garden. Tech industry leaders including Meta Platforms's Mark Zuckerberg, Apple's Tim Cook, and Microsoft's Satya Nadella are expected to attend an artificial intelligence event hosted by U.S. first lady Melania Trump on Thursday and then join U.S. President Donald Trump for an evening reception in the White House's newly renovated rose garden. Other attendees are expected to include OpenAI's Sam Altman, Alphabet's Sergey Brin and Sundar Pichai, and Microsoft co-founder Bill Gates, a White House official said. Alexandr Wang, who recently joined Meta's AI effort, is also among the invitees. The first lady announced last month that she was launching a presidential challenge to encourage students to use emerging AI technology to find solutions to community challenges. The effort will also encourage educators to adopt AI in the classroom, the White House has said.

Russian forces launched 408 attacks on Ukraine's Zaporizhia region, killing a 62-year-old man in the Polohy district, Regional Governor Ivan Fedorov said. A man was killed in a Russian attack on the Khmelnytskyi region in western Ukraine, the region's governor, Serhii Tiurin, said. Russian attacks injured at least 35 people across Ukraine on Wednesday, including 14 people in the Donetsk area and 14 people in the Kherson region, local officials said, according to the Kyiv Independent news outlet. Russia's Ministry of Defence said that its forces shot down 170 Ukrainian drones, five guided aerial bombs and two rockets in a 24-hour period, according to Russia's state-run TASS news agency. Ukraine's air force said it shot down 430 Russian drones and 21 missiles, the Kyiv Independent reported.

Unlearning in large language models (LLMs) involves precisely removing specific information from a pre-trained model. This is crucial to ensure safety of LLMs by deleting private data or harmful knowledge acquired during pre-training. However, existing unlearning methods often fall short when subjected to thorough evaluation. To overcome this, we introduce JensUn, where we leverage the Jensen-Shannon Divergence as the training objective for both forget and retain sets for more stable and effective unlearning dynamics compared to commonly used loss functions. In extensive experiments, JensUn achieves better forget-utility trade-off than competing methods, and even demonstrates strong resilience to benign relearning. Additionally, for a precise unlearning evaluation, we introduce LKF, a curated dataset of lesser-known facts that provides a realistic unlearning scenario. Finally, to comprehensively test unlearning methods, we propose (i) employing an LLM as semantic judge instead of the standard ROUGE score, and (ii) using worst-case unlearning evaluation over various paraphrases and input formats. Our improved evaluation framework reveals that many existing methods are less effective than previously thought.

In this study we analyzed average sleep durations across 61 countries to examine the impact of Daylight Saving Time (DST) practices. Key metrics influencing sleep were identified, and statistical correlation analysis was applied to explore relationships among these factors. Countries were grouped based on DST observance, and visualizations compared sleep patterns between DST and non-DST regions. Results show that, on average, countries observing DST tend to report longer sleep durations than those that do not. A more detailed pattern emerged when accounting for latitude: at lower latitudes, DST-observing countries reported shorter sleep durations compared to non-DST countries, while at higher latitudes, DST-observing countries reported longer average sleep durations. These findings suggest that the influence of DST on sleep may be moderated by geographical location.

This paper addresses the critical gap in evaluating bias in multilingual Large Language Models (LLMs), with a specific focus on Spanish language within culturally-aware Latin American contexts. Despite widespread global deployment, current evaluations remain predominantly US-English-centric, leaving potential harms in other linguistic and cultural contexts largely underexamined. We introduce a novel, culturally-grounded framework for detecting social biases in instruction-tuned LLMs. Our approach adapts the underspecified question methodology from the BBQ dataset by incorporating culturally-specific expressions and sayings that encode regional stereotypes across four social categories: gender, race, socioeconomic class, and national origin. Using more than 4,000 prompts, we propose a new metric that combines accuracy with the direction of error to effectively balance model performance and bias alignment in both ambiguous and disambiguated contexts. To our knowledge, our work presents the first systematic evaluation examining how leading commercial LLMs respond to culturally specific bias in the Spanish language, revealing varying patterns of bias manifestation across state-of-the-art models. We also contribute evidence that bias mitigation techniques optimized for English do not effectively transfer to Spanish tasks, and that bias patterns remain largely consistent across different sampling temperatures. Our modular framework offers a natural extension to new stereotypes, bias categories, or languages and cultural contexts, representing a significant step toward more equitable and culturally-aware evaluation of AI systems in the diverse linguistic environments where they operate.

-- Intuitive T eleoperation interfaces are essential for mobile manipulation robots to ensure high quality data collection while reducing operator workload. A strong sense of embodiment combined with minimal physical and cognitive demands not only enhances the user experience during large-scale data collection, but also helps maintain data quality over extended periods. This becomes especially crucial for challenging long-horizon mobile manipulation tasks that require whole-body coordination. We compare two distinct robot control paradigms: a coupled embodiment integrating arm manipulation and base navigation functions, and a decoupled embodiment treating these systems as separate control entities. Additionally, we evaluate two visual feedback mechanisms: immersive virtual reality and conventional screen-based visualization of the robot's field of view. These configurations were systematically assessed across a complex, multi-stage task sequence requiring integrated planning and execution. Our results show that the use of VR as a feedback modality increases task completion time, cognitive workload, and perceived effort of the teleoperator . Coupling manipulation and navigation leads to a comparable workload on the user as decoupling the embodiments, while preliminary experiments suggest that data acquired by coupled teleoperation leads to better imitation learning performance. Our holistic view on intuitive teleoperation interfaces provides valuable insight into collecting high-quality, high-dimensional mobile manipulation data at scale with the human operator in mind.