Latest news and articles about ai & ml
Samsung Electronics has deployed ChatGPT Enterprise and Codex to its employees globally, representing one of OpenAI's largest enterprise AI implementations. This rollout enables Samsung's workforce to leverage large language models and code generation capabilities across the organization.
OpenAI, Anthropic, and Adobe aren't just building AI assistants anymore — they're racing to integrate their models into physical systems. This week's announcements signal a strategic shift that could reshape who controls the robotics stack.
Waymo's second major recall signals a shift in how we handle autonomous technology. Rather than a sign of failure, the recall mechanism may be the most promising development in robotics safety yet — if regulators can keep up.
GrayMatter Robotics highlights how its autonomous surface-finishing systems using edge-deployed physical AI can address critical workforce shortages in defense manufacturing and depot maintenance. The U.S. military faces significant readiness challenges due to a 174,000-worker shortfall in the industrial base, with surface preparation and finishing work being a critical bottleneck that traditional automation cannot handle due to the geometric variation in each job.
Carnegie Mellon's VideoManip system represents a genuine breakthrough in robot learning — not because it's flashy, but because it solves a fundamental data problem that's been holding the field back for years.
Two separate announcements this week revealed automotive suppliers achieving over 99% success rates on production tasks with AI-powered robots. The impressive numbers look great in press releases, but they obscure a more uncomfortable truth about how we're measuring progress in physical AI.
AURA Foresight has been selected as one of four finalist teams in the XPRIZE Wildfire Autonomous Wildfire Response competition, a $5 million global challenge to develop autonomous technology for detecting and responding to wildfires. The system combines fixed sensors, artificial intelligence, and swarms of autonomous flying robots to detect ignitions within minutes and coordinate rapid intervention before fires grow out of control, with finals taking place in Alaska.
Luca Guadagnino's Sam Altman biopic has to find a new studio after Amazon dropped it.
The U.S. industrial robotics market grew 11% in 2025 with 38,000 unit installations, driven by 30% surge in food industry adoption, while China maintains dominant global market share with 295,000 annual installations. The IFR reports the U.S. now ranks eighth globally in robot density, and industry advocates are pushing for a national robotics strategy including a Federal Robotics Office and domestic technology procurement mandates.
Robot Talk Episode 161 features an interview with Allison Okamura, a Stanford University professor specializing in haptics, teleoperation, and medical robotics. The podcast discusses her research on collaborative haptic systems that enable advanced touch-based robotic interaction, covering applications in teleoperation, virtual reality, and medical robotics.
The WEEDINATOR autonomous agricultural robot project has reached a significant milestone after nearly a decade of development, demonstrating effective mechanical weed removal using a modified commercial Iseki tractor base. The system uses computer vision (OpenCV and YOLO26n on a Jetson Nano) combined with hydraulically-controlled cultivator claws for autonomous weeding in three axes.
SpaceX and other companies are pursuing the concept of building AI data centers in orbit to leverage abundant solar energy and avoid Earth-based environmental and infrastructure constraints. However, the article examines significant engineering challenges including radiation damage, heat dissipation difficulties, expensive repairs, and launch costs that make orbital data centers far more complex than launching satellites alone.
From Snap's insistence that its new glasses aren't 'AI glasses' to a wave of rebranding across the tech industry, companies are suddenly running from the label they spent billions promoting. This isn't just marketing — it's a sign that AI has reached peak hype and crossed into backlash territory.
Three major announcements this week signal a fundamental shift in medical robotics — from experimental technology reserved for elite hospitals to essential infrastructure for underserved populations. The trend reveals something unexpected: accessibility, not capability, is now driving innovation.
OpenAI introduces new spend controls and usage analytics for ChatGPT Enterprise, helping organizations manage costs and scale AI with confidence.
Intel's RealSense unveils the D585 Pro AI-native depth camera designed for robotics applications, featuring on-device AI acceleration, a custom Gen 5 system-on-chip, and capabilities for humanoids, autonomous mobile robots, and industrial systems. The camera offers sub-15cm minimum range, 120×100° field of view at 60 FPS, IP65 protection, and is expected to ship in Q1 2027, with software-defined updates enabling new capabilities without hardware replacement.
Adobe is integrating its Firefly AI Assistant directly into Creative Cloud apps including Premiere, Photoshop, Illustrator, InDesign, and Frame.io via a sidebar interface, enabling users to automate workflows and delegate repetitive tasks. The company is also expanding Firefly's capabilities with new features like context-aware element saving, brand identity creation, video rough cuts, and video generation from storyboards to improve consistency across creative projects.
XGIMI's MemoMind One smart glasses serve as a second screen for smartphones, featuring transparent micro-LED displays for notifications, translations, and AI responses, with 16-hour battery life and comfortable 47-gram design. However, the review criticizes the limited physical controls and expresses significant concerns about an intrusive AI feature that the reviewer finds ethically problematic.
Kinova Inc. launched KIMA, a medical robotic arm specifically designed for clinical environments including endoscopy, bronchoscopy, and surgical interventions. The compact arm features a 3 kg payload, weighs under 13 kg, integrates medical-grade safety standards (IEC 62304 Class C and ISO 14971), and uses open EtherCAT architecture to integrate with existing medical platforms and surgical instruments.
Richtech Robotics launched a round-the-clock interactive livestream platform featuring its ADAM AI-powered humanoid robot, allowing global audiences to interact with the robot in real time through chat and direct requests. The initiative demonstrates advances in embodied AI and natural human-robot interaction, with ADAM having previously served at the Kennedy Space Center and prepared noodles at a restaurant industry show, while the company expands its AI infrastructure with a new Las Vegas warehouse.
Waymo is recalling over 3,800 of its self-driving robotaxis due to a software issue that could cause vehicles to enter closed freeway construction zones at high speed. The company has voluntarily restricted freeway operations and filed a software recall with NHTSA, marking the second recall for the autonomous taxi fleet in over a month.
OpenAI has improved ChatGPT's health and wellness response capabilities through GPT-5.5 Instant, which features enhanced reasoning, better context understanding, clearer communication, and physician-informed evaluation frameworks to provide more reliable health-related information.
Researchers leveraged an OpenAI reasoning model to assist in diagnosing rare genetic diseases in children, successfully identifying 18 new diagnoses in cases that had previously remained unsolved. This application demonstrates the practical use of advanced AI reasoning capabilities in clinical settings to improve diagnostic outcomes for rare pediatric conditions.
From automotive suppliers to construction sites, physical AI platforms are finally leaving research labs and entering real factories. But the gap between impressive success rates and actual production deployment reveals an industry struggling with the messy reality of manufacturing.
OpenAI has rolled out a new Scheduled tasks hub in ChatGPT that allows users to better manage and monitor scheduled prompts, with improvements to reliability and speed. The update includes a dedicated Scheduled page where users can view, pause, edit, and delete upcoming tasks, and introduces monitoring capabilities that let ChatGPT proactively search the web or connected apps on behalf of users. The feature is available to Plus, Pro, Business, and Enterprise users, while OpenAI is sunsetting its Pulse daily summary feature in favor of the new scheduling system.
Epic Games is making generative AI a big part of upcoming versions of Unreal Engine.
Anthropic took its Claude Fable 5 and Mythos 5 AI models offline following a U.S. government export-control directive, citing concerns that the models' dual-use capabilities for both cybersecurity defense and exploitation pose national security risks. Security experts argue the restriction is ineffective, noting that similar advanced AI capabilities are likely already being developed by competitors and will become widely available across multiple companies and open-source developers within months.
Anthropic has released a major update to Claude Design that integrates it more seamlessly with Claude Code, allowing designers to work from local codebases and hand off designs to the coding agent for implementation. The update includes improved design system imports, refined image editing tools, admin controls for organizations, and shared usage limits across Anthropic's products.
CMU researchers have developed VideoManip, a system that trains robots to perform dexterous manipulation tasks by converting videos of people interacting with objects into training data. The system reconstructs 3D hand movements and object interactions from ordinary internet videos, eliminating the need for specialized equipment or teleoperated demonstrations, and addresses a major bottleneck in robot learning by enabling robots to learn from internet-scale datasets.
As physical AI moves from research labs into automotive plants and industrial settings, a new challenge is emerging: proving these systems are safe enough for production environments. Recent developments suggest the industry is racing to solve a problem most people don't know exists yet.
Fennec Engineering has received TÜV Rheinland certification (Tool 2 qualification) for its Advanced Safety Acceleration Platform (ASAP), a SaaS tool designed to automate functional safety processes for robotics and AI systems. The platform helps manufacturers manage safety requirements, traceability, and compliance with IEC 61508 and ISO 13849 standards, addressing the challenge of integrating safety into complex autonomous systems from the early stages of development.
Autonomique, an AI startup spun out of SRI International, is deploying its physical AI platform and mobile manipulator at Canadian automotive supplier F&P Manufacturing to address labor shortages and production complexity. The company's hardware-agnostic software combines teleoperation systems, generalist-specialist AI architecture, and reinforcement learning to enable industrial robots to adapt to multi-step workflows with human-like dexterity while maintaining manufacturing precision and reliability.
MIT and University of Pennsylvania researchers developed MIGHTY, an open-source trajectory-planning system that enables unmanned aerial vehicles (UAVs) to navigate complex environments with obstacle avoidance in milliseconds while optimizing flight paths. The system uses novel mathematical formulations to generate smooth trajectories faster than existing methods while running efficiently on onboard computers, with potential applications in search-and-rescue, delivery, and industrial inspection.
Sanctuary AI has achieved a 99.5%+ success rate on a complex wire-plugging task at a Tier 1 automotive supplier, demonstrating its physical AI technology can meet production-line performance and cycle-time requirements. Rather than waiting for humanoid robots to reach commercialization, the Vancouver-based company is deploying its AI and proprietary hydraulic hands on existing industrial robot platforms to address manufacturing labor challenges.
OpenAI and Molecule.one have demonstrated a near-autonomous AI chemist using GPT-5.4 that successfully improved a challenging drug-making reaction in medicinal chemistry. This advancement shows how large language models can be applied to optimize complex chemical synthesis processes, representing progress in AI-assisted scientific research and automation of chemistry workflows.
LifeSciBench is a new expert-authored and expert-reviewed benchmark designed to evaluate AI systems on real-world life science research tasks and decisions. The benchmark provides a rigorous evaluation framework for assessing how well AI models can handle complex scientific workflows in the life sciences domain.
While the tech world obsesses over humanoid robots and general-purpose AI, some of the most significant advances in physical AI are happening in the least glamorous place imaginable: construction sites. Recent partnerships and new platforms suggest the industry is finally ready to move beyond cautious pilot programs.
UK government partners with Google DeepMind to build a new AI-powered prototype aimed at faster housing decisions.
Snap CEO Evan Spiegel unveiled Snap Specs, AR glasses priced at $2,195, positioning them as a "see-through computer" rather than AI glasses, emphasizing augmented reality and computing overlay rather than recording capabilities. Spiegel highlighted Snap's approach to safety by prohibiting facial recognition features in Lenses, implementing developer moderation, and offering parental controls for younger users amid increasing regulatory scrutiny of smartglasses and social media.
We'll see how players react.
Pegasus Tech Ventures launched a $60 million corporate venture capital fund in partnership with CYBERDYNE Inc. to invest in startups developing robotics, physical AI, healthcare automation, and intelligent systems. The fund will focus on advancing CYBERDYNE's vision for human-cyber-physical space (HCPS) cybernics, combining humans, AI, robotics, and information systems to address global challenges like aging populations and workforce shortages.
Casey Harrell's 3,800 hours with a brain-computer interface represents the first time we can study what happens when medical AI leaves the lab and enters daily life. The numbers tell a story about technology that actually works—and what that means for robotics.
Qualcomm unveiled the Snapdragon Reality Elite processor for next-generation AR and mixed reality headsets, offering significant improvements in performance, efficiency, and battery life while enabling smaller, more efficient devices with on-device AI capabilities. The company also introduced START (Scalable Turnkey AI Ready Toolkit), a platform designed to help eyewear brands and accessory makers integrate AI features into smartglasses and wearables, with UK eyewear company Inspecs as an initial partner.
Xreal officially unveiled its Android XR Aura glasses at the Augmented World Expo, featuring a 70-degree field of view, built-in hand tracking, and Qualcomm's Snapdragon Reality Elite chip. The company opened two-week reservations ($99 with $199 launch credit, or $299 for founder priority pass) but declined to announce the final price, targeting a fall 2026 launch in select markets.
HPE and NVIDIA are expanding the HPE AI Factory to support agentic AI workloads, introducing the NVIDIA Vera CPU for agent-specific computing and the NVIDIA Agent Toolkit for HPE Private Cloud AI, alongside extended confidential computing capabilities across the entire portfolio. The Vera CPU, designed for tool calls and real-time data processing in agent loops, will be available in 2027, with early enterprise adoption from the New York Stock Exchange exploring the technology.
Researchers at the Istituto Italiano di Tecnologia studied elephant trunk skin structure and discovered a functionally zoned architecture with distinct upper and lower regions—the upper skin acts as armor while the underside is highly flexible for grasping. The findings, published in PNAS Nexus, reveal how subsurface dome-shaped structures amplify tactile signals and could inspire the design of next-generation soft robotic grippers with improved tactile sensing and mechanical protection.
Google researchers Rohin Shah and Four Flynn present the AI Control Roadmap, a security framework designed to manage increasingly capable AI agents through a defense-in-depth approach that combines traditional safeguards with model alignment and system-level controls. The roadmap treats AI agents as potential insider threats, using threat-modeling frameworks and supervisor AI systems to detect, prevent, and respond to misaligned behaviors while maintaining scalable security as AI capabilities advance.
NVIDIA's Blackwell platform swept MLPerf Training 6.0 benchmarks, achieving the fastest training times across all seven benchmarks and scaling to 8,192 GPUs—the largest Blackwell-based submission to date. The results showcase performance improvements including up to 1.6x faster training with the GB300 NVL72 over GB200 NVL72, and demonstrate Blackwell's capability for large-scale mixture-of-experts model training including DeepSeek-V3 671B and Llama 3.1 405B.
Kawasaki Robotics will debut its RL030N eight degree-of-freedom robot platform designed for physical AI applications at Automate 2026 in Chicago, along with demonstrations of its Pulseboard inspection technology and other industrial robots. The RL030N combines high-speed motion, enhanced dexterity, and real-time control capabilities to bridge the gap between traditional industrial robotics and AI-driven applications requiring adaptive motion and confined-space manipulation.
Genesis AI unveiled Eno, a general-purpose mobile manipulator robot powered by GENE, its robotics-native foundation model, designed with prioritized mobility and dexterity over human appearance. The company plans to begin production and targeted deployments by end of 2026, starting with industrial customers in manufacturing, logistics, and labs before expanding to service industries and consumer applications.
