Alibaba on Tuesday introduced RynnBrain, an artificial intelligence model built to support robotics, as competition intensifies among global technology companies to define the software foundations of physical AI. The launch reflects how leading AI developers are moving beyond text and images toward systems that can interpret and act within real-world environments.

RynnBrain is designed to help robots perceive their surroundings, identify objects, and coordinate movement accordingly. In a demonstration released by Alibaba’s DAMO Academy, a robot identifies pieces of fruit and places them into a basket. While visually unremarkable, the task requires a tight integration of perception, reasoning, and motion control, areas that have long constrained the commercial deployment of autonomous machines.

Physical AI as a Strategic Priority

Robotics is increasingly grouped under the broader category of physical AI, which includes machines such as industrial robots and self-driving vehicles that rely on artificial intelligence to operate in dynamic environments. China has made physical AI a strategic focus, viewing it as a critical arena in its technological competition with the United States.

Industry leaders have echoed the scale of the opportunity. Nvidia chief executive Jensen Huang has described AI and robotics as a multitrillion-dollar growth market, a framing that helps explain why major model developers are investing in systems that extend beyond digital tasks. For Alibaba, RynnBrain represents an entry point into this emerging layer of the AI stack, complementing its existing work on large language models under the Qwen brand.

Competing World Models Take Shape

Alibaba’s move places it alongside a growing list of companies developing so-called world models, AI systems intended to help machines understand and simulate the physical world. Nvidia has introduced several robotics-focused models under its Cosmos platform, while Google DeepMind has developed Gemini Robotics-ER, aimed at embodied reasoning and control. Tesla is pursuing a similar approach internally for its Optimus humanoid robot.

The competition is particularly pronounced in humanoid robotics, where China is widely viewed as moving faster than the U.S. toward scaled production. Several Chinese manufacturers have signaled plans to ramp output this year, suggesting that software capable of generalizing across tasks and environments will be a key differentiator.

Open Source as an Adoption Strategy

Like Alibaba’s other recent AI releases, RynnBrain is being offered under an open source model, allowing developers to use and modify it freely. Open sourcing has been central to Alibaba’s strategy for expanding the reach of its AI ecosystem, particularly outside China, and contrasts with more closed approaches taken by some Western rivals.

The decision also reflects a broader industry debate over how foundational physical AI systems should be distributed. By lowering barriers to experimentation, Alibaba is positioning RynnBrain as infrastructure rather than a proprietary product, betting that widespread adoption will accelerate progress in robotics while reinforcing its role in the global AI landscape.

China has taken another step in turning humanoid robotics from laboratory experiments into public spectacle and industrial validation. On February 9, the world’s first humanoid robot free combat league officially opened in Shenzhen, positioning competitive robot fighting as both entertainment and a stress test for embodied artificial intelligence.

The league, known as the Ultimate Robot Knockout Legend (URKL), will run through the end of 2026 and award a championship belt valued at 10 million yuan, or roughly $1.4 million. Organized by EngineAI, the competition brings together development teams from across China and provides each participant with a standardized humanoid robot platform, removing hardware barriers and shifting the focus toward software, control systems, and real-world performance.

Combat as a Testbed for Embodied AI

At the center of the league is EngineAI’s T800 humanoid robot, a full-size platform capable of executing complex martial movements such as aerial rotations, kicks, and rapid directional changes. While the visual appeal is undeniable, organizers and analysts emphasize that the event is designed less as a stunt and more as a proving ground for robotics technology under extreme conditions.

Combat scenarios place unusual demands on humanoid systems, including dynamic balance, motion planning under impact, actuator durability, and rapid decision-making. According to industry observers, these high-stress environments expose weaknesses that may remain hidden in controlled factory or laboratory settings, accelerating iteration cycles and hardware-software integration.

Experts note that robot combat also serves an important cultural function. By framing humanoid robots within a familiar and dramatic format inspired by martial arts, the league challenges public perceptions of robots as purely industrial tools and introduces them as expressive, adaptable machines. This visibility, particularly among younger audiences, may help build long-term talent pipelines into robotics and AI research.

Opportunity and Limits of Robot Fighting

Analysts caution, however, that combat performance does not automatically translate into commercial readiness. Optimizing robots for short bursts of high-impact activity can divert attention from the requirements of industrial, service, or household deployment, where endurance, safety, and cost efficiency matter far more than spectacle.

Even so, supporters argue that entertainment-focused trials can play a meaningful role in early-stage technology development. Real-world combat provides data that simulations struggle to replicate, especially around mechanical stress, failure modes, and system resilience. Some estimates suggest that such environments can shorten development cycles by more than 30 percent.

The league arrives as China’s humanoid robotics sector accelerates rapidly. Industry projections estimate the domestic humanoid robot market could approach 870 billion yuan by 2030, driven by advances in embodied intelligence and growing interest in both industrial automation and consumer-facing robots.

For now, the combat league is less about harvesting immediate commercial products and more about planting ideas. In Shenzhen, humanoid robots are no longer just walking, lifting, or performing demos. They are fighting, failing, adapting, and learning in public – offering a glimpse into how embodied AI might evolve beyond the lab and into everyday life.

Indian automation company Addverb has officially entered the humanoid robotics race with the launch of Elixis-W, a wheeled humanoid robot designed specifically for industrial environments. The robot was unveiled at LogiMAT India 2026, one of the country’s largest logistics and supply chain exhibitions, and positions India as an active contributor to the global shift toward physical AI.

Wheeled humanoid robots working in the warehouse🏭

Indian automation company @Addverb_Tech unveiled its first wheeled humanoid robot:Elixis-W, at LogiMAT India 2026.

Built on a physical AI architecture, Elixis-W uses sensors to perceive its environment and can perform path… pic.twitter.com/dD6MkAQOjA

— CyberRobo (@CyberRobooo) February 7, 2026

Unlike consumer-focused humanoids or lab prototypes, Elixis-W was introduced as a practical system built for warehouses, factories, and intralogistics operations. Industry reaction at the event reflected a broader inflection point: humanoid robotics in India is moving beyond demonstrations and into controlled, real-world deployment.

Addverb executives framed the launch not as a spectacle, but as an extension of the company’s long-standing focus on automation systems that solve concrete operational problems.

Why Addverb Chose a Wheeled Humanoid

Elixis-W combines wheeled mobility with dual-arm manipulation, perception systems, and learning capabilities. The decision to adopt wheels rather than bipedal locomotion was deliberate. Most industrial facilities are built around flat, structured floors, where wheeled platforms offer greater stability, efficiency, and faster integration with existing infrastructure.

According to Addverb leadership, this approach allows teams to validate use cases such as material handling, repetitive tasks, and higher-risk operations without requiring costly changes to factory layouts. Legged humanoids remain part of the company’s longer-term roadmap, but the wheeled form factor enables earlier commercialization and safer deployment today.

The robot is intended to assist with physically demanding and repetitive work while operating alongside human workers. Addverb has emphasized that Elixis-W is designed to augment human labor, not replace it, with people retaining responsibility for supervision, decision-making, and exception handling.

From Automation Systems to Physical AI Platforms

The launch of Elixis-W reflects a broader evolution in Addverb’s strategy. The company initially built its reputation by solving structured automation challenges in warehouses and manufacturing. With Elixis-W, it is moving toward collaborative robotics capable of functioning in environments designed for humans.

Deployment will begin through limited, closely supervised proof-of-concept programs in controlled industrial settings. This phased rollout is intended to validate safety, reliability, and consistency before any wider adoption.

Alongside the robot, Addverb also introduced Addverb.ai, a platform aimed at advancing general-purpose robotics and physical AI. By opening the platform to developers and researchers, the company hopes to accelerate progress toward robots that can adapt across tasks rather than being locked into single-purpose roles.

For Addverb, Elixis-W is not positioned as a final product, but as a foundation. It represents a pragmatic step toward human-robot collaboration at scale, rooted in industrial realities rather than futuristic promises.

Li Auto is increasingly redefining what it means to be an automaker. As competition in China’s electric vehicle market intensifies and margins tighten, the company is framing artificial intelligence and embodied robotics as the foundation of its next decade of growth.

Rather than treating AI as a feature layered onto vehicles, Li Auto’s leadership is presenting intelligence as the product itself, with cars, and eventually robots, serving as physical embodiments of that capability. The shift reflects a broader reassessment underway across the auto industry, where autonomy, perception, and decision-making are emerging as the primary sources of long-term differentiation.

From Smart Vehicles to Embodied AI

Li Auto has spent years building in-house AI systems to support advanced driver assistance and autonomous driving. These systems handle perception, motion planning, and real-time decision-making in complex environments. Company executives now argue that the same technology stack can extend beyond vehicles into general-purpose physical AI.

Humanoid robotics has emerged as a focal point of that vision. Like autonomous cars, humanoid robots must operate safely in spaces designed for people, interpret human intent, and adapt continuously to changing conditions. Engineers see deep overlap between the two domains, particularly in vision systems, control software, and large-scale data training.

Li Auto has not announced a standalone humanoid product or timeline. Instead, the company has described its robotics ambitions as a long-term effort focused on foundational intelligence rather than near-term commercialization.

Li Xiang Frames the Car as a Robot

That philosophy was made explicit on Feb. 5, when Li Auto CEO Li Xiang published a lengthy post on Weibo outlining the company’s view of the automobile’s ultimate form. In the essay, Li revisited a question posed when the company was founded in 2015: what should a car ultimately become?

While the industry largely viewed the answer as faster or smarter transportation, Li Auto’s internal conclusion was different. The company saw the car evolving into a robot, a vision Li described as guiding its strategy for the past decade. He emphasized that the company’s long-standing slogan was not symbolic, but a literal roadmap that has shaped product and technology decisions over time.

Li pushed back on speculation that Li Auto’s focus on AI signals a retreat from vehicles. Instead, he argued that embodied intelligence must be grounded in a strong automotive foundation to create real value. He noted that the majority of his time remains devoted to the car itself, ensuring the organization understands how to evolve a vehicle into an intelligent system step by step.

According to Li, the newly released Li Auto L9 represents a turning point in that process. He described the vehicle as an intelligent agent equipped with perception, cognition, and actuation, transforming the car from a passive tool into an active partner that recognizes users, anticipates needs, and responds proactively. In his framing, the L9 is not just a new model, but the first concrete expression of Li Auto’s embodied AI vision as the company enters its second decade.

An Industry Looking Beyond Cars

Li Auto’s messaging aligns with a broader shift across China’s technology and manufacturing sectors. As electric vehicles become increasingly commoditized, companies are looking to intelligence, not hardware alone, as the defining advantage. Humanoid robotics and embodied AI offer a path to reuse autonomy expertise while opening new markets in logistics, services, and industrial automation.

For Li Auto, cars remain the primary product and revenue driver. But the company is increasingly clear about its long-term ambition. It does not see itself solely as an automaker, but as a builder of intelligent machines whose form may evolve over time.

Whether humanoid robots become a commercial reality or remain a research frontier, Li Auto’s strategy underscores a growing belief across the industry: the future of mobility companies may be determined less by what they manufacture, and more by how intelligent those machines can become.

Faraday Future is no longer framing intelligence as something confined to cars. The company has officially launched its embodied AI robotics division and introduced three commercial robotic products, marking a decisive expansion beyond electric vehicles and into physical artificial intelligence.

The move positions Faraday Future alongside a growing list of automakers and tech firms redefining mobility as an intelligent, autonomous system rather than a traditional product category. Company executives argue that embodied AI represents a natural evolution of the vehicle-as-robot concept Faraday Future has promoted for nearly a decade.

The company has launched a new internal initiative centered on embodied artificial intelligence and revealed plans to develop a lineup of humanoid robots designed for real-world interaction. Executives described the effort as an expansion of the company’s mission rather than a departure from mobility, arguing that modern vehicles are increasingly becoming intelligent robotic systems.

The announcement arrives amid a broader industry shift. Just days earlier, Tesla announced it would wind down production of its flagship Model S and Model X vehicles to redirect resources toward Optimus, its humanoid robot platform. Together, the moves highlight how legacy automotive players increasingly see physical AI – not just transportation – as the next frontier.

Three Robots, One Ecosystem Strategy

Faraday Future’s first robotics lineup includes three distinct platforms. FF Futurist is positioned as a full-size professional humanoid robot designed for public-facing and enterprise roles, from hospitality and retail to education and exhibitions. FF Master targets consumer and institutional environments as a more athletic, interaction-focused humanoid companion. FX Aegis rounds out the lineup as a quadruped robot aimed at security, patrol, and outdoor assistance tasks.

All three products are being offered with a new ecosystem-based pricing model that separates hardware from software capabilities. Base prices range from $2,499 for FX Aegis to $34,990 for FF Futurist, with optional ecosystem skill packages enabling advanced AI functions, customization, and continuous upgrades. The company describes the model as demand-driven, designed to mirror how software platforms evolve over time rather than traditional one-time hardware sales.

Faraday Future said more than 1,200 units are already covered by paid B2B deposits, signaling early commercial interest ahead of first deliveries expected by the end of February. Production preparation is underway, while customization, testing, and data training continue in parallel to accelerate deployment.

Why Faraday Future Believes Robotics Can Scale Faster Than Cars

Company leadership argues that embodied AI robotics offer structural advantages over vehicles: lighter capital requirements, faster iteration cycles, and earlier paths to positive cash flow. Robotics and vehicles are positioned as dual engines, sharing AI models, manufacturing expertise, and distribution channels.

At the technical level, Faraday Future is betting on a modular AI architecture built around what it calls a three-part system: physical devices, an open AI brain platform, and a decentralized data factory designed to continuously improve robot performance through real-world usage. The company believes this approach allows robots to adapt across industries without being locked into single-purpose designs.

The company also sees automotive dealerships as a natural sales and service channel for robotics. Executives argue that future dealers will evolve into intelligent terminal operators, offering vehicles and robots through shared infrastructure, financing models, and service networks.

Faraday Future’s leadership frames the launch not as a speculative experiment, but as the opening move in a long-term transition. As embodied AI advances and robots move from labs into everyday environments, the company believes ownership could eventually surpass today’s global vehicle fleet.

For Faraday Future, the future of mobility is no longer defined by wheels alone – but by intelligent machines capable of working, learning, and interacting alongside humans.

At some Starbucks drive-throughs, customers may still hear a friendly greeting, but the voice taking their order is no longer always human. Artificial intelligence is increasingly stepping in, quietly reshaping how the coffee giant operates as it works to regain momentum after years of uneven sales.

The shift is part of a broader push by Starbucks to integrate AI and automation across its business. The company has invested hundreds of millions of dollars in technology designed to speed up service, reduce operational friction, and support baristas with behind-the-scenes tasks. Early results suggest the approach is beginning to pay off. Starbucks recently posted its first U.S. same-store sales increase in two years, a notable turnaround in a market that generates roughly 70 percent of its revenue.

Technology is now embedded throughout the Starbucks workflow. Inside stores, baristas can rely on AI-powered virtual assistants to recall drink recipes, manage shift schedules, and answer operational questions in real time.

In back rooms, automated scanning systems handle inventory checks, replacing one of retail’s most time-consuming manual tasks and helping address out-of-stock issues that have frustrated customers. At select drive-through locations, AI systems are being tested to process orders, freeing employees to focus on drink preparation and customer interaction.

The company frames these tools not as replacements for workers, but as support systems. In a recent statement, Starbucks described AI as a way to “support the moments that matter,” arguing that technology should reduce distractions and allow staff to focus on human connection. New features under development include AI chat tools that help customers discover drinks based on preferences or mood, as well as scheduling options that allow orders to be timed in advance to minimize waiting.

The investment has not come without trade-offs. Starbucks’ spending spree – which also includes about $500 million to increase staffing levels – has weighed on margins, contributing to investor concerns and recent share price volatility. Management has pledged to find $2 billion in cost savings over the next three years, a goal that makes automation central to its long-term profitability.

Leading the effort is Brian Niccol, who took over in 2024 amid rising prices, labor tensions, and intensifying competition. Since then, he has paused price hikes, simplified the menu, set a four-minute service target, and cut thousands of corporate roles. Underperforming stores have been closed, and Starbucks has reduced its exposure to China by selling a large stake in the business there.

Despite the focus on technology, Niccol has repeatedly emphasized that Starbucks lost its way by prioritizing efficiency over experience. His turnaround plan includes a renewed emphasis on handwritten names on cups, more comfortable seating, ceramic mugs, and store redesigns aimed at restoring a neighborhood coffeehouse feel. Each store refresh can cost up to $150,000 and will take several years to complete.

To Niccol, the combination is not contradictory. Automation handles repetitive and invisible work, while baristas are meant to deliver warmth and speed at the counter. In that sense, Starbucks’ AI push mirrors a wider trend across retail and food service, where technology is increasingly used to stabilize operations and enhance consistency rather than replace frontline workers.

Whether robots and algorithms can help Starbucks fully recapture its cultural relevance remains uncertain. But for now, the company is betting that a carefully calibrated blend of automation and human touch can turn improved sales into lasting recovery.

Mechanical engineers at Duke University have developed a new class of programmable materials that blur the line between passive structures and living systems. The work, which was published in Science Advances, demonstrates solid building blocks whose mechanical behavior can be rewritten on demand – allowing the same structure to behave like soft rubber, rigid plastic, or something in between without being rebuilt or reshaped.

At the center of the research are Lego-like cubes, each composed of 27 internal cells. Every cell contains a gallium-iron composite that can switch between solid and liquid states at room temperature. By selectively heating individual cells with small electrical currents, researchers can liquify precise regions inside the block, effectively encoding stiffness, damping, and movement into an otherwise rigid object.

The geometry never changes. Only the internal state does.

In early demonstrations, the team assembled multiple cubes into beams and columns. Simply altering which internal cells were liquified caused the same structure to bend, vibrate, or resist motion in dramatically different ways. Mechanical behavior was no longer fixed at the time of manufacture but became a variable that could be adjusted repeatedly after assembly.

One of the most striking experiments took place underwater. Researchers connected ten cubes into a straight column and attached it to a motor, forming a programmable tail for a robotic fish. With the same motor input, different internal configurations caused the fish to swim along sharply different paths. Motion was altered not by changing motors or control software, but by reprogramming the material itself.

“We want to make materials that are alive,” said Yun Bai, the study’s first author and a PhD student at Duke. “Traditional manufacturing lets you print a material with a certain stiffness, but to change it you have to start over. We wanted something closer to human muscle – a material that can adjust its mechanical response in real time.”

Unlike shape-shifting systems, the Duke approach does not rely on changing form. Instead, it rewrites how forces propagate through a structure. In two-dimensional tests, thin sheets made from the same composite demonstrated a wide range of stiffness and damping behaviors while maintaining identical shapes. In performance tests, the sheets rivaled or exceeded commercially available materials across multiple mechanical metrics.

The modular design adds another layer of flexibility. Each cube can be attached or removed like a building block, allowing engineers to assemble larger systems with highly customized mechanical behavior. Once a configuration has been tested, freezing the structure at zero degrees Celsius returns all internal cells to a solid state, effectively resetting the system for reprogramming.

“This gives us a way to build three-dimensional structures whose mechanical properties are not fixed,” Bai said. “You can test one configuration, reset it, and try another – again and again.”

The researchers see applications far beyond robotics. By adjusting the metal composition, the freezing and melting points could be tuned for environments such as the human body. Miniaturized versions could one day navigate blood vessels, form adaptive medical implants, or create electronics that physically respond to changing conditions.

“Our long-term goal is to construct larger systems using these composite materials,” said Xiaoyue Ni, an assistant professor of mechanical engineering and materials science at Duke. “We want to enable robots and machines to adapt mechanically to different tasks and environments without redesigning the entire system.”

The work suggests a future where materials are no longer passive components, but active participants – structures that can be programmed, reset, and adapted as easily as software.

New York Robotics has formally launched amid a surge in robotics investment, startup formation, and talent concentration across the New York Tri-State region. The announcement marks a turning point for New York’s robotics sector, which now counts more than 160 robotics startups across New York, New Jersey, and Connecticut, with nearly 100 based in New York City alone.

“Our vision is to leverage New York’s central role in the global economy to build a leading robotics hub,” said Jacob Hennessey-Rubin, founding board member and executive director of New York Robotics. “This is a place where founders, researchers, enterprises, and investors can collaborate to shape the future of robotics, while positioning the sector as a serious investment category on Wall Street.”

The milestone signals New York’s emergence as a serious global contender in robotics and embodied AI, joining established hubs such as Boston, Silicon Valley, Pittsburgh, Munich, and Zurich. Long viewed as strong in finance, media, and enterprise software, New York is increasingly becoming a center for robotics commercialization as the technology moves from research into real-world deployment.

A Dense and Expanding Robotics Network

New York Robotics was formed to address a long-standing gap in the region’s innovation landscape. While the city has deep pools of capital, customers, and technical talent, robotics activity historically developed in silos. NYR’s mission is to unify startups, investors, enterprises, researchers, and government organizations into a coordinated ecosystem capable of supporting companies from early research through large-scale deployment.

Over the past two years, New York Robotics has quietly assembled one of the most comprehensive robotics networks in the United States. The organization now engages more than 450 robotics startups worldwide, including over 160 in the Tri-State region, alongside more than 80 corporations, 20 academic institutions, 40 research labs, 300 venture capital firms, and dozens of domestic and international government organizations.

Founding members include organizations such as J.P. Morgan, New York University, AlleyCorp, EisnerAmper, and Cybernetix Ventures. The group has also hosted or co-hosted more than 20 ecosystem events, including the first dedicated robotics programming at NY TechWeek, helping elevate robotics as a visible and strategic sector within the city’s broader technology economy.

NYR’s official launch coincided with its sponsorship of AlleyCorp’s inaugural Deep Tech NY conference, underscoring the growing convergence of robotics, capital, and enterprise demand in the region.

Capital, Talent, and Commercial Pull

Unlike many robotics hubs driven primarily by academic research, New York’s ecosystem is shaped by proximity to enterprise customers in finance, healthcare, logistics, construction, and real estate. This commercial pull is attracting startups focused on deploying robots in real operating environments rather than remaining confined to labs.

“We see New York Robotics as a kind of exchange,” said Randy Howie, founding board member and managing partner of New York Robotics. “It’s a platform where startups, enterprises, investors, academia, and government can connect efficiently and translate robotics innovation into broad economic impact.”

Industry leaders echo that view. Founders point to New York’s ability to attract global engineering talent while offering access to industrial infrastructure in surrounding areas such as Long Island and New Jersey, where robotics companies can scale manufacturing and testing.

Mapping the Ecosystem

As part of its platform, New York Robotics recently released a private beta of the NYR Index, an ecosystem intelligence tool designed to map startups, investors, labs, and enterprise participants. The tool is intended to give partners visibility into deal flow, talent movement, and sector trends as robotics investment accelerates.

NYR has also partnered with organizations such as C10 Labs, an operator selected by NYCEDC for the NYC AI Nexus, to support applied AI ventures across robotics, advanced manufacturing, healthcare, and sustainability.

With robotics increasingly viewed as essential infrastructure for addressing labor shortages, productivity challenges, and demographic shifts, New York Robotics’ formal launch reflects a broader reality: New York is no longer just a consumer of robotics technology. It is becoming a place where the next generation of robots is built, funded, and deployed.