OpenAI’s robotics lead stepping down over ethical concerns caught my attention, as it shows the growing tension between innovation and responsibility in AI development. Caitlin Kalinowski’s principled resignation demonstrates that even cutting-edge engineering feats must respect clear boundaries, especially when national security and surveillance are involved. While OpenAI’s rapid Pentagon deal may seem opportunistic, it reminds us that ethical debates in technology are complex and still ongoing. Thus, as engineers, we must balance progress with careful oversight, and this story reminds us that this principle should guide our work as much as technical prowess.

Read the full article here

Samsung SDI’s introduction of a pouch-type solid-state battery prototype for humanoid robots marks a major step for robotics and AI-powered machines. This battery addresses the challenge of handling sudden power spikes, which is crucial for robot performance and safety. By moving beyond traditional lithium-ion designs, Samsung is enabling higher energy density and improved reliability in compact systems. The new pouch format also supports flexible integration into tight spaces, potentially benefiting a wide range of AI platforms and next-generation devices. This development could very well shape the future of high-performance power solutions across multiple industries.

Read the full article here

Meta’s acquisition of Moltbook, the social network built for AI agents, is a fascinating indicator of where AI development may be heading. By bringing Moltbook and its OpenClaw technology under its roof, Meta is positioning itself at the forefront of autonomous agent interaction; a space that’s drawing significant industry attention. This decision by Meta further proves how major tech companies are doubling down on infrastructure for AI-to-AI communication, which could reshape how automation supports both individuals and businesses. With competition heating up, developments like this are worth watching, as they could set new standards for intelligent systems and their integration into daily life.

Read the full article here

Sandbar’s $23 million Series A for its Stream smart ring caught my eye, as it could change how we as humans might interact future technology. The idea of a wearable focused purely on voice notes, without the health tracking clutter, seems both practical and refreshingly targeted (compared to failed attempts like the Human AI Pin). I also appreciate that the design encourages intentional use, with the mic only active when you lift your hand. With early adopters embracing the device and Sandbar planning deeper AI integration, it’s quite exciting to see hardware innovation address everyday needs in such a streamlined way.

Read the full article here

Mind Robotics’ $500 million Series A, led by Accel and Andreessen Horowitz, marks a potential breakthrough for industrial automation. The company is tackling the limitations of traditional robotics by developing AI-powered systems capable of dexterous and adaptive tasks; crucial for modern manufacturing. With Rivian as a strategic partner, Mind Robotics gains access to real-world production data, accelerating both model training and deployment. This substantial investment indicates a growing confidence amongst investors in robotics as a solution for labor shortages and global competitiveness, something that could very well spark a new industrial revolution.

Read the full article here

Arduino’s recent launch of the Ventuno Q edge AI platform brings engineers into a new era for robotics and embedded AI. Powered by Qualcomm’s Dragonwing IQ8 processors and paired with a real-time STM32H5 microcontroller, the board delivers both powerful AI capabilities and precise hardware control on a single device. With support for Linux, a robust development environment, and broad hardware compatibility, Ventuno Q is positioned to accelerate innovation in offline robotics, machine vision, and automation. Thus, this new platform continues to prove Arduino’s tradition of making advanced technology more accessible, opening up new possibilities for developers in education, research, and industry.

Read the full article here

The first Lego Smart Brick sets, launched with a Star Wars theme, showed both promise and initial limitations. While the hardware includes sensors such as a microphone, ambient light, and fine distance measurement, many of these features are not yet enabled and may require future software updates. Currently, the sets make use of NFC readers and basic motion and color sensors for interactive play. Battery life is limited to about 45 minutes of active use, and the non-replaceable battery raises questions about long-term sustainability. So, are these bricks the future of play, or just another un-needed integration of technology?

Read the full article here

MIT’s new multimaterial 3D-printing platform demonstrates a major step forward in hardware manufacturing, as it can print a complete, functional electric linear motor in just a few hours. The system uses five distinct material types, including conductors and magnetic materials, to fabricate devices that traditionally require complex assembly and multiple supply chains. While the prototype motor matches or exceeds conventional builds, the technology is still in early stages, especially for more demanding applications like rotating motors. This approach could make hardware production faster, more flexible, and less dependent on traditional manufacturing infrastructure in the future.

Read the full article here

Researchers at Nankai University have unveiled a promising hydrofluorocarbon-based electrolyte for lithium metal batteries, achieving a notable energy density of 700 Wh/kg at room temperature and around 400 Wh/kg at -50°C. By moving away from traditional oxygen-based solvents, this new approach enables better charge transfer and performance in extreme cold—an area where conventional batteries struggle. The development not only addresses key limitations in current lithium-ion technology but also opens the door for energy-dense, scalable batteries suitable for harsh environments, making it a significant advance for sectors requiring reliable low-temperature operation.

Read the full article here

Researchers at Washington State University have created a 3D-printed, beating model of the left side of the human heart, offering a reusable and patient-specific training tool for surgeons and medical students. Unlike traditional models or simulations, this replica mimics both the anatomy and movement of a real heart, using soft materials and pneumatic actuators for lifelike motion. The model allows for practice of complex procedures, including valve repairs, with real-time feedback from built-in sensors. This advancement could significantly enhance surgical training by providing more realistic and customizable practice scenarios without relying on animal or cadaver models.

Read the full article here

As someone who appreciates the openness and flexibility of RISC-V, I hate to agree with the author’s assessment of the current status of the ISA: performance is still a significant hurdle. When you see build times for RISC-V lagging far behind other architectures, often by hours, it’s clear the hardware ecosystem needs to mature before RISC-V can be a first-class citizen in distributions like Fedora. While upcoming SoCs like the UltraRISC UR-DP1000 and SpacemiT K3 offer some promise, the current reality is that speed and manageability just aren’t there yet for mainstream, rack-mounted server use. I’m rooting for RISC-V, but it has to get faster if it’s going to compete.

Read the full article here