physical AI

Data centers are increasingly shaping the physical American landscape, with large-scale facilities being built across the country to meet growing demand for cloud computing and artificial intelligence. The expansion is drawing attention to their land use, energy consumption, and impact on local communities, as commentators debate whether the benefits outweigh the environmental and infrastructure costs.

Nvidia is reportedly pursuing glass substrate packaging for its AI chips, a manufacturing shift that could improve performance and density as conventional materials hit physical limits. The move could reshape how advanced semiconductors are packaged and give Nvidia an edge in the race to power large AI systems, though details of timing and partners remain unclear.

Researchers are combining artificial intelligence with physics-based models to push imaging beyond what cameras can directly capture, from reconstructing hidden scenes to automatically colorizing anime footage. In a new Q&A, experts explain how machine learning can fill in missing visual information while physical laws keep the results grounded, pointing toward faster, cheaper imaging tools for science and media alike.

A Harvard study has concluded that an AI tutor outperformed traditional in-class physics teaching. The findings suggest students using the AI tutor achieved better learning outcomes than those in conventional classroom settings. The result is drawing attention as educators debate the role of artificial intelligence in higher education and whether such tools should be adopted more widely in science courses.

South Korean robotics company Angel Robotics has launched PHAI-X1, a wearable robot platform aimed at physical AI research. The system is intended to give researchers a hardware base for developing and testing AI that interacts with the physical world, such as exoskeleton control and human-assistance applications. Details on pricing and availability were not included in the initial announcement.

Rystad Energy warns that surging power demand from data centers, driven by AI and cloud computing growth, is colliding with the limits of electricity grids and generation capacity. The firm's message is blunt: no amount of investment or scheduling can outrun the physical constraints on power supply, and grid bottlenecks will increasingly shape where and how fast data centers can be built.

Bloomberg Tech reports that AMD has acquired World Labs, the spatial-intelligence startup founded by AI researcher Fei-Fei Li, in a deal valued at $8.2 billion. The move signals AMD's push beyond chips into AI systems that understand 3D space and the physical world, a technology many see as a possible stepping stone toward more general artificial intelligence and advances in robotics.

Tohoku University announced research on an AI model for materials discovery that is grounded in physics rather than relying purely on data patterns. The approach aims to make predictions of new materials more reliable and reproducible, addressing a common weakness of purely statistical machine learning in chemistry and materials science. The university released the work through its research news service.

AMD is acquiring a startup outside its core chipmaking business in an all-stock deal valued at $8.2 billion, a move driven by CEO Lisa Su's push into physical AI, technology that connects artificial intelligence to robots, sensors and real-world systems. The deal signals AMD's ambition to compete beyond traditional semiconductors as demand for AI applied to hardware and machines accelerates across the industry.

Analysis in the medical device industry examines how regulators must adapt as AI moves beyond software and into surgical robots that physically operate on patients. The piece argues that existing approval frameworks, built for diagnostic tools and imaging, may not be suited to autonomous or semi-autonomous systems whose behaviour can change after deployment, raising questions about oversight, accountability and post-market monitoring.

Tech workers in San Francisco connected ChatGPT to a Toyota Corolla and had it navigate a parking lot using an off-the-shelf language model with no prior driving training. Only one model managed to complete the course. The experiment is drawing attention for showing both the potential and the limits of using general-purpose AI for real-world control tasks.

Real estate services firm JLL argues that physical AI — robots and intelligent machines working in buildings — is set to become the sector's next major transformation, following earlier waves of digitalisation and smart-building technology. The claim points to automation reshaping how properties are managed, operated and valued, though the company has released few details beyond the headline assertion.

Online car retail is being transformed by artificial intelligence and automation, streamning browsing, pricing and financing for buyers. Yet one part of the process remains stubbornly offline: buyers still want to test drive or inspect a car in person before committing. Industry observers point to this gap as a limit on fully digital car buying, prompting retailers to blend AI-driven convenience with physical showroom experiences.

Robotics company Astribot is bringing its T1 robot to North America, presenting its physical AI technology stack built around a 'design for AI' approach. The announcement, distributed via a press release, positions the company as it expands from its home market into the North American robotics sector, where interest in AI-powered humanoid and manipulation robots is growing fast. Details beyond the launch itself remain limited in the available reporting.

AMD has agreed to acquire the startup founded by AI pioneer Fei-Fei Li in a deal worth $8.2 billion, part of a broader push into physical AI technology. The acquisition signals AMD's ambition to expand beyond chips for data centers into AI systems that operate in the real world, intensifying competition with Nvidia and other rivals in the rapidly growing physical AI sector.

Industry commentary argues that physical AI — robotics and intelligent machines moving from pilots to real-world deployment — still lacks a repeatable, reliable path to production. Writers in the manufacturing press say companies need standardized processes to take AI-powered systems out of demos and into factories at scale, echoing broader debate about how fast embodied AI can be industrialized.