OpenAI has publicly rolled out its GPT-5.6 model after government approval and unveiled 'ChatGPT Work,' a new AI agent combining ChatGPT and Codex for non-technical users. This new offering aims to streamline professional tasks by integrating with various apps and creating diverse content.
Four new US microreactors have achieved "criticality," a symbolic milestone in nuclear energy development, surpassing a federal goal. While this marks a positive step for emerging clean energy technologies, significant technical and regulatory hurdles remain before these reactors can contribute to the power grid.
Researchers have developed a novel framework that unifies calibration with computation in quantum computers, enabling them to learn from their errors and operate continuously. This breakthrough, utilizing reinforcement learning, significantly improves the logical stability and performance of quantum systems.
Scientists from the Wellcome Sanger Institute have developed Phylo-Plex, a new computational method for cost-effective and scalable DNA sequencing of pathogens. This innovation promises to transform how laboratories with limited resources track disease outbreaks and monitor antibiotic resistance worldwide.
EmTech AI 2026 highlighted the profound transformation of artificial intelligence into a foundational platform, reshaping industries and societies. Discussions ranged from new AI models and their impact on the labor market to advancements in brain implants and software development.
Researchers have developed a novel "gradient solvation electrolyte" using a targeted ligand anti-solvent (TLAS) to significantly enhance the stability and cycle life of high-voltage lithium metal batteries. This breakthrough addresses critical issues of oxidative decomposition and interphase deterioration, paving the way for more efficient energy storage.
SpaceXAI has released Grok 4.5, described by Elon Musk as an "Opus-class model" that is faster, more token-efficient, and lower cost than competitors. This new model aims to automate various complex tasks with double the token efficiency of other leading AI models.
Meta is reportedly developing "super sensing" smart glasses that continuously record audio and snap photos, using metadata to power its AI. This raises significant privacy concerns, especially as the recording indicator will remain off in continuous sensing mode.
Google's SynthID system achieved a significant victory by debunking a widely circulated AI-generated hoax image of Senator Mitch McConnell. This marks a crucial success for anti-deepfake technology, demonstrating the effectiveness of invisible watermarks in identifying synthetic media.
A new study systematically evaluates the feasibility of deploying advanced humanoid robots for laparoscopic surgical tasks, assessing their precision, control, and safety in real-world scenarios, including live animal surgeries.
Waymo has officially launched fully driverless operations in Las Vegas, allowing riders to experience vehicles without a human safety driver. The Alphabet subsidiary plans to expand this service to Denver, San Diego, and Tampa, pushing towards its goal of one million paid rides per week by 2026.
OpenAI has unveiled new voice models, GPT-Live-1 and GPT-Live-1 mini, designed to offer more natural and interactive live conversations through full-duplex communication. These models allow users to speak and listen simultaneously, enhancing turn-taking and enabling features like live translation.
New research suggests that advanced orbital detectors could soon identify and track nuclear warheads in space, offering a critical tool for global security and arms control. This development could deter the weaponization of space and foster greater transparency among nations.
While global responses to critical mineral supply shocks focus on increasing extraction and diversifying sources, a crucial underlying constraint is often overlooked: intellectual property. Addressing this technological barrier is essential for truly widening access to these vital materials.
A simulated war game exposed the chilling reality of a catastrophic cyberattack on US water utilities, revealing widespread societal collapse from food shortages to hospital failures. Experts warn that China's "Volt Typhoon" group is actively pre-positioning to disrupt critical civilian infrastructure.
AI chip maker SambaNova Systems has raised $1 billion in its Series F round, pushing its valuation to $11 billion just five months after its last major funding. The company also announced a strategic partnership with JPMorgan Chase for on-premises AI inference solutions.
Microsoft is reportedly shifting its AI strategy to reduce costs by increasingly deploying its own in-house AI models, lessening reliance on third-party solutions from OpenAI and Anthropic. This move, already evident in applications like Excel and Word, reflects a broader industry trend towards optimizing AI expenditures amidst rising operational costs.
A new study reveals that California's premier wine-growing regions like Napa and Sonoma face severe threats from global warming and increasing wildfire risks, potentially shifting the industry's heartland to new areas such as Mendocino and Monterey. Proactive adaptation strategies, including careful varietal selection, are crucial for the future of viticulture in the state.
Quantinuum has unveiled a 98-qubit trapped-ion quantum computer with all-to-all connectivity, marking a significant leap in quantum computing capabilities. This innovation promises to revolutionize fields from drug discovery to AI, despite a minor publisher correction regarding an author's name.
Sam Altman's proposal for the US government to take a 5% stake in OpenAI aims to share AI wealth and create a safety net, while a leaked Treasury report warns of an AI market bubble.
The iconic Hubble Space Telescope, after over three decades of groundbreaking discoveries, faces orbital decay, prompting a global effort to preserve its legacy. NASA and private entities are exploring options, including reboost missions, to extend its operational life or ensure a safe deorbit.
Meta's new AI image model, Muse Image, now automatically uses public Instagram photos for generative AI remixes. Users must manually opt out through settings, as no notifications are sent for content usage and existing AI images will not be deleted.
Modern life relies heavily on precise satellite timing signals, but these are vulnerable to cyber threats. Scientists are now exploring quantum entanglement to create highly secure ground-to-satellite links, promising unprecedented protection for critical timing infrastructure.
Anthropic's Claude Cowork AI agent is now available on smartphones and web browsers, capable of running tasks independently without an active desktop session. This marks a significant shift towards always-on, autonomous AI agents controlled via mobile, enhancing productivity and flexibility for users.
