Industry Briefing
A single destination for timely, editor-curated robotics news from around the world.
General Intuition raises $320M to use video game data to train robots
General Intuition has secured $320 million in funding to enhance its innovative approach to artificial intelligence training for robotics. The company is leveraging video game clips that feature embedded action labels to accelerate the training process for AI systems. This funding will enable General Intuition to further develop its technology, which aims to improve the efficiency and effectiveness of robotic learning. By utilizing the rich data from video games, the company seeks to provide robots with a more nuanced understanding of actions and environments, ultimately advancing the field of robotics. The investment comes at a time when the demand for sophisticated AI solutions in various industries is on the rise, highlighting the potential impact of General Intuition's work in shaping the future of robotics.
Artificial Intelligence Artificial Intelligence / Cognition Design / Development Financial Investments News
Why this CEO thinks video games make better training data than the internet
Recent discussions in the field of artificial intelligence have highlighted the limitations of large language models, such as ChatGPT and Claude, in achieving artificial general intelligence (AGI). While these models excel in text generation, they struggle with understanding the dynamics of movement through space and time, a critical component for developing generalized intelligence. To address this gap, researchers are exploring the potential of gaming data as a solution. This innovative approach, known as General Intuition, aims to leverage the rich, interactive environments found in video games to enhance AI's understanding of real-world physics and dynamics. By integrating insights from gaming, experts believe they can create more sophisticated models capable of reasoning and adapting in complex scenarios. The exploration of this method is ongoing, with the hope of advancing the field of AGI significantly.
AI Startups AI Funding general intuition physical ai Pim DeWit
This startup thinks robotics is about to have its ChatGPT moment
General Intuition is investing significantly in the potential of video game data to enhance the development of physical artificial intelligence. The company believes that millions of hours of gameplay data can be utilized to train foundational models, which could streamline the process of creating more intelligent robots that require less real-world data for training. This initiative comes as the demand for advanced robotics continues to grow, with applications spanning various industries. By leveraging the rich datasets generated from video games, General Intuition aims to accelerate the evolution of AI technologies, making them more efficient and capable in real-world scenarios. The project is set to unfold over the coming months, with the goal of transforming how robots learn and adapt to their environments.
AI Robotics Equity general intuition pim de witte
Why Are Large Language Models So Terrible at Video Games?
Recent advancements in large language models (LLMs) have led to significant improvements in various domains, particularly in coding. However, a notable limitation remains: LLMs struggle to play video games effectively. Despite some successes, such as Gemini 2.5 Pro defeating Pokémon Blue in May 2025, these models often perform poorly compared to human players, making frequent mistakes and requiring specialized software to assist them. Julian Togelius, director of New York University’s Game Innovation Lab and co-founder of AI game-testing firm Modl.ai, discussed these challenges in a recent interview with IEEE Spectrum. He highlighted that while coding resembles a well-structured game with clear tasks and immediate feedback, video games present a more complex landscape that LLMs have yet to navigate successfully. Unlike games like chess or Go, which have been mastered by AI through retraining, video games vary significantly in mechanics and input requirements, complicating the development of a general game AI. Togelius pointed out that the lack of comprehensive benchmarks for video games further hinders LLMs' performance. While benchmarks have driven improvements in coding, the diverse nature of video games makes it difficult to establish similar metrics. He noted that current LLMs perform poorly even compared to basic algorithms in gaming contexts, primarily due to insufficient training data and challenges in spatial reasoning. Despite their coding capabilities, LLMs cannot engage in the iterative process of game development, which involves testing and refining gameplay. This disparity raises questions about the future of AI in mastering video games and its implications for broader AI applications.
Llms Artificial-intelligence Video-games
General Intuition’s $2.3B bet that video games can train AI agents for the real world
General Intuition has successfully secured $320 million in funding to enhance its artificial intelligence capabilities, leveraging extensive data derived from millions of hours of gameplay and betting action. This significant investment aims to advance the development of AI that mimics human intuition more closely. The funding round, which took place recently, underscores the growing interest in AI technologies that can interpret complex patterns and make decisions similar to human behavior. By utilizing this vast dataset, General Intuition seeks to refine its algorithms and improve the performance of its AI systems, positioning itself at the forefront of innovation in the gaming and betting industries.
AI Fundraising Robotics Startups AI training data general intuition
From Fortnite to robots: General Intuition raises $2.3B on bet that video games can train AI agents for the real world
General Intuition has successfully secured $320 million in funding to enhance its artificial intelligence capabilities, which are designed to mimic human intuition by leveraging extensive data from millions of hours of gameplay and betting actions. This significant investment aims to propel the development of AI systems that can better understand and predict human behavior in gaming contexts. The funding round, which reflects growing interest in AI applications within the gaming and betting industries, will enable General Intuition to expand its research and development efforts. The company plans to utilize this capital to refine its algorithms and improve the accuracy of its AI models, ultimately striving to create technology that resonates more closely with human decision-making processes.
Startups AI Robotics Fundraising Khosla Ventures AI training data
Figure Launches 'Project Go-Big' to Train Humanoid Robots on Human Video Data
Figure, a humanoid robotics company, has initiated an extensive data collection effort named 'Project Go-Big,' which seeks to develop a comprehensive pretraining dataset by recording human interactions in everyday home environments. This ambitious project is bolstered by a collaboration with Brookfield, a leading real estate firm. As a result of this partnership, Figure's robot has successfully learned navigation skills solely from the human video footage captured during the initiative. The project represents a significant step forward in the field of robotics, aiming to enhance the capabilities of humanoid robots by leveraging real-world data.
Brett Adcock Figure AI brookfield helix robotics
From R&D to Data Analysis, AI Is Rewriting How Game Companies Operate
YOOZOO Interactive deploys AI agents across 75% of its workforce, using Tencent Cloud WorkBuddy to transform game development from coding to global marketing analytics.
Industry
Video: Atlas humanoid robot learns soccer skills from World Cup game footage
Boston Dynamics has unveiled a new video featuring its Atlas humanoid robot, showcasing its advanced capabilities in preparation for the 2026 World Cup. Released recently, the video highlights the robot's agility and precision as it performs a series of complex movements, including running, jumping, and navigating obstacles. This demonstration aims to illustrate the potential applications of robotics in sports and entertainment, particularly as the world gears up for the highly anticipated tournament. The release comes as part of Boston Dynamics' ongoing efforts to push the boundaries of robotic technology and its integration into various fields. By showcasing Atlas's skills, the company hopes to inspire innovation and spark discussions about the future role of robots in athletic events and beyond.
NVIDIA and ComfyUI Streamline Local AI Video Generation for Game Developers and Creators at GDC
At the Game Developers Conference (GDC) held this week in San Francisco, NVIDIA unveiled a series of updates aimed at enhancing the capabilities of game developers and artists. The company showcased how its RTX AI PCs are enabling the creation of cinematic worlds and iconic characters, significantly elevating the standard for immersive gaming experiences. This initiative reflects NVIDIA's commitment to pushing the boundaries of technology in the gaming industry, providing developers with advanced tools to craft more engaging and visually stunning content. The updates are expected to empower creators to leverage artificial intelligence in innovative ways, ultimately transforming the gaming landscape.
ByteDance Seedance: How China's Video-Gen Model Turned the Tide
ByteDance has unveiled Seedance 2.0, marking a significant advancement in the realm of video generation technology in China. This innovative model has quickly established itself as the leading force in the market, boasting impressive gross margins ranging from 70% to 90%. The launch, which took place in October 2023, is poised to redefine the artificial intelligence business landscape by setting new standards for profitability and efficiency in video content creation. By leveraging cutting-edge technology and extensive data analysis, ByteDance aims to capitalize on the growing demand for high-quality video content, positioning itself at the forefront of the industry.
News
Alibaba's Qwen3.7-Plus supports text, video and imagery inputs at low cost of $0.4/$1.6 per 1M token — but it's proprietary
Alibaba has unveiled its latest AI large language model, Qwen3.7-Plus, this week, enhancing its Qwen family with advanced multimodal capabilities and a cost reduction of 60% compared to the previous text-only model, Qwen3.7-Max. Unlike earlier versions, Qwen3.7-Plus is available solely under a closed commercial license through proprietary APIs and Qwen Chat, marking a significant shift from Alibaba's previous strategy of offering open-source models. This change may disappoint users and enterprises, including major U.S. companies like Airbnb, that relied on open-source versions. The new model excels in multimodal tasks, such as generating enterprise-grade visuals and analyzing videos and images, which its predecessor could not perform. With a competitive pricing structure, Qwen3.7-Plus is positioned just above its Chinese competitor's discounted model. It features a 1-million token context window and a unique parameter called 'preserve_thinking,' which helps maintain continuity during complex tasks, a critical need for developers. Despite its advantages, benchmarks indicate that Qwen3.7-Plus still falls short of leading U.S. models in raw capability. However, it is designed to replace high-cost models in developer workflows and robotic process automation, offering a cost-effective solution for enterprises. The model's cloud-based deployment raises compliance concerns for organizations with strict data residency requirements, as it cannot be downloaded or hosted locally. Overall, Qwen3.7-Plus presents a compelling option for enterprises seeking efficient, multimodal AI solutions without incurring high operational costs.
Technology
Google’s Gemini Omni turns images, audio, and text into video — and that’s just the start
Google has unveiled its latest innovation, Gemini Omni, a multimodal model designed to enhance user interaction by reasoning across various formats, including text, images, audio, and video. This cutting-edge technology allows users to generate and edit videos through straightforward conversational prompts, with the initial feature being Omni Flash. The launch of Gemini Omni marks a significant advancement in artificial intelligence, aiming to simplify content creation and editing processes for users. The model is trained on data available up to October 2023, ensuring it incorporates the most recent developments in AI capabilities. This initiative reflects Google's commitment to pushing the boundaries of technology and improving user experiences in digital content creation.
Media & Entertainment AI Google Veo google io 2026 google gemini omni
Video Friday: Heavy Robotic Machinery Operates Itself
IEEE Spectrum robotics has released its latest edition of "Video Friday," showcasing a variety of innovative robotics videos and announcing upcoming robotics events. Notable conferences include ICRA 2026 in Vienna from June 1-5, and the Summer School on Multi-Robot Systems in Prague from July 29 to August 4, 2026. Among the featured content is a groundbreaking autonomous material-handling solution developed by ETH Zurich, capable of operating a 40-ton material handler, traditionally a labor-intensive task. Other highlights include the introduction of the world's first production-ready manned mecha by Unitree, and the launch of NIX, an embodied AI exploring dance and movement, which will be made available for free to select partners. NTNU has unveiled the Unified Autonomy Stack, a robust system designed for aerial and ground robots, enhancing their autonomy through advanced perception and navigation capabilities. This system has been validated in challenging environments, showcasing its effectiveness in exploration and object discovery. Robotics expert Rodney Brooks, cofounder of Robust AI, shared insights on the challenges of innovation in robotics and the implications of the current AI surge during a Q&A session ahead of his panel at the Forbes America Innovates event in San Francisco. As the field continues to evolve, the integration of data collection methods remains a significant challenge, with the Koala platform exemplifying innovative approaches to enhance robotic manipulation tasks.
Home-robots Humanoid-robots Video-friday Material-handling-technology Robot-videos Robot-grippers
Video Friday: These Robots Were Born to Run
IEEE Spectrum's weekly feature, Video Friday, showcases a collection of innovative robotics videos and highlights upcoming robotics events, including the International Conference on Robotics and Automation (ICRA) scheduled for June 1-5, 2026, in Vienna. This week’s videos introduce advanced modular building blocks that allow for the automatic design and rapid assembly of agile robots capable of navigating unstructured outdoor environments, a significant advancement from traditional legged robots that rely on predefined designs. Additionally, the RIVR project presents a concept for an ideal urban delivery robot, while KIMLAB introduces TRIP-Bag, a portable teleoperation system designed for high-fidelity data collection in various settings. TUM has developed an open-vocabulary semantic exploration system that enhances robots' ability to map and locate objects in dynamic environments. In a practical demonstration, a reconfigurable robot assists a team responding to a simulated radiation leak at a nuclear facility, showcasing its adaptability in challenging conditions. Meanwhile, Carnegie Mellon University's Robomechanics Lab reflects on a devastating flood that destroyed many of its robots, honoring them with a tribute video. Lastly, the 2025 Silicon Valley Humanoids Summit features a keynote by Disney Research's Moritz Baecher, discussing insights into robotic character development.
Robotics Humanoid-robots Video-friday Modular-robots Robot-videos Quadruped-robots
Video Friday: A Robot Hand With Artificial Muscles and Tendons
IEEE Spectrum robotics has released its weekly roundup of notable robotics videos and a calendar of upcoming events, including the International Conference on Robotics and Automation (ICRA) scheduled for June 1-5, 2026, in Vienna. This week's highlights feature advancements in biomimetic design, showcasing a printed hand that integrates soft and rigid components with artificial muscles, enhancing our understanding of natural kinematic structures. Boston Dynamics product managers reflect on classic robots, including LittleDog, which contributed to legged locomotion research over a decade ago. Additionally, DRAGON Lab has introduced a new trajectory planning method for floating-based articulated robots, facilitating exploration in complex environments. Their OmniPlanner tool has been tested extensively across various terrains, including underground mines and forests. The FZI Research Center, in collaboration with ETH Zurich and other institutions, has made strides in preparing for lunar missions by testing cooperative autonomous multirobot teams outdoors. Meanwhile, advancements in humanoid robotics are being discussed, with Kamel Saidi from NIST emphasizing the importance of performance standards for broader adoption. In academia, Junyao Shi from UPenn's GRASP lab will address the challenges of building general-purpose robots, focusing on how human data and foundation models can bridge existing gaps in robotics. This ongoing exploration of robotics technology underscores the field's rapid evolution and its potential impact on various sectors.
Humanoid-robots Video-friday Underwater-robots Bipedal-robots Robot-videos
5 Best Audio to Video AI Generators for Modern Content Workflows
The audio to video AI generator sector has become an essential component of contemporary content production, facilitating the conversion of spoken audio, voiceovers, and scripts into cohesive visual narratives. This innovative technology streamlines the content creation process by automating scene generation, timing alignment, and visual selection, thereby eliminating the need for traditional editing methods. As a result, users can produce high-quality content at scale, significantly enhancing efficiency and creativity in various media projects. The advancements in this field reflect the growing demand for automated solutions in content creation, driven by the need for faster turnaround times and increased output in an ever-evolving digital landscape.
Business Design Software ai avatars AI content tools AI storytelling
Peking University team develops new generation data acquisition device using EMG wristband, backed by Gong Hongjia, Lu Qi, and overseas
The SnowOrigin team, composed of researchers from Peking University, has secured investments from notable figures including Gong Hongjia and Lu Qi, as well as overseas institutions. This innovative team focuses on surface electromyography (sEMG) technology to develop a new generation of human control data collection solutions, utilizing wearable devices like neural wristbands and panoramic headsets, along with their proprietary Neural Math Hybrid (NMH) AI decoding model. As the fields of embodied intelligence and Physical AI rapidly evolve, there is an increasing demand for high-quality human control data. Current mainstream data collection methods, such as first-person video and motion capture, often fail to capture critical information about the intent and nuances of human actions. SnowOrigin's wearable devices aim to bridge this gap by integrating muscle and neural signal decoding technologies to create structured data that includes posture, force, and micro-control, thereby supporting the training of robots and world models. Founder Qin Xu emphasized that unlike traditional lab-based motion capture systems, their wearable solutions are cost-effective, lightweight, and suitable for long-term use without disrupting daily activities. The team is advancing two commercialization pathways: enhancing human-robot interaction for AI devices and building a foundational data infrastructure for Physical AI applications. With a strong academic background and a commitment to innovation, SnowOrigin is positioned to lead in the emerging market for embodied data collection, having already made significant strides in real-time decoding of sEMG signals into actionable insights. As the demand for comprehensive interaction data grows, the team is poised to capitalize on this shift in paradigm.
MIT researchers channel AI to turn hand gestures into robot training data
Researchers have developed an innovative method to enhance the capabilities of humanoid robots, particularly in tasks such as grasping objects. This advancement involves the use of a specialized ultrasound wristband worn by a human instructor, which captures the intricate movements of muscles, tendons, and ligaments beneath the skin. By analyzing this data, the robots can learn to replicate these movements more effectively. The initiative, which began in late 2023, aims to improve the dexterity and functionality of robots in various applications, from manufacturing to personal assistance. The ultrasound technology provides real-time feedback, allowing the robots to adjust their movements based on the instructor's actions. This approach not only enhances the robots' ability to perform complex tasks but also opens new avenues for human-robot interaction. The research is being conducted at a leading robotics lab, where experts are focused on bridging the gap between human-like movement and robotic precision. By mimicking the natural motion of human hands, the robots are expected to achieve greater efficiency and adaptability in their operations. This breakthrough could significantly impact industries that rely on automation, making robots more versatile and capable of handling delicate tasks that require a human touch.
Robotics
NetEase Games’ Eggy Party PC version goes live today
NetEase Games has officially launched its party game, Eggy Party, on Windows, expanding access to PC players. This new version allows for cross-platform account and data synchronization, enabling players to switch seamlessly between mobile and PC while maintaining their progress. Available today, the PC edition retains popular game modes such as Peak Party, Casual Party, and Eggy Farm, ensuring that fans of the game can enjoy their favorite features on a new platform. The introduction of the PC version aims to enhance the gaming experience and attract a broader audience.
News Feed
AIRoA collects 80,000 hours of robot operation data through industry-academia collaboration for physical AI infrastructure.
The AI Robot Association (AIRoA) released a YouTube video on May 29, 2026, showcasing a groundbreaking initiative titled "A Massive Collaborative Physical AI Data Initiative." The video highlights the ongoing operations of robots across various universities and research institutions, illustrating the accumulation of a global data infrastructure. This initiative aims to enhance the development of artificial intelligence by creating a comprehensive database that supports collaborative research and innovation in robotics. Through this project, AIRoA seeks to foster advancements in AI technology and improve its applications in real-world scenarios.
How One Million Hours of Human Video Became a 'Textbook' for Robot Learning
A research team at Peking University has unveiled the HumanNet dataset, a comprehensive collection of one million hours of human-centered videos aimed at advancing robot training in physical tasks. Released in October 2023, this extensive dataset offers a wealth of diverse perspectives and detailed annotations, enhancing the learning capabilities of robots. The initiative seeks to improve the interaction between robots and humans by providing a rich resource that reflects real-world scenarios, ultimately fostering more effective and adaptable robotic systems.
Robot Learning Human-Centered Data AI Training Computer Vision
Why traditional robotics data collection is obsolete and what replaces it
Eric Chan has highlighted Rhoda AI's groundbreaking strategy in the field of robotics, emphasizing the company's use of video data to enhance the scalability and efficiency of machine learning processes. In a recent discussion, Chan pointed out that traditional methods of data collection in robotics have become outdated, necessitating a shift towards more advanced techniques. This innovative approach not only streamlines the learning process for robots but also addresses the limitations of conventional data gathering methods. The insights shared by Chan underscore the importance of adapting to new technologies in order to improve robotic capabilities and performance.
Artificial Intelligence Business Resources News Opinion Podcast Robots / Platforms
Deepfake Detection Dataset Aims to Keep Up With Generative AI
A collaborative effort involving researchers from Microsoft, Northwestern University, and the non-profit organization Witness has led to the development of a new dataset aimed at enhancing the detection of AI-generated media. Announced in a study published on April 10 in IEEE Intelligent Systems, the Microsoft-Northwestern-Witness (MNW) deepfake detection benchmark is designed to address the growing challenge of distinguishing real from fake content in an era where generative AI technology is rapidly advancing. The dataset includes a diverse array of AI-generated images, audio, and videos, reflecting the current landscape of generative AI. Thomas Roca, a principal research scientist at Microsoft, emphasized the increasing sophistication of AI-generated media, which can easily be produced by anyone using accessible applications. This proliferation raises significant concerns, including identity fraud and the creation of harmful content. The MNW benchmark aims to improve the effectiveness of detection systems by providing a wider variety of AI-generated materials, including those that have undergone post-processing manipulations. Researchers acknowledge that while this dataset could potentially be misused to develop new evasion techniques, it is crucial for enhancing the ability to assess the authenticity of media as generative AI continues to evolve. The team plans to update the dataset biannually to incorporate the latest developments in generative AI and detection challenges, with the goal of fostering transparency and raising standards in the fight against deepfake content.
Deepfakes Generative-ai Artificial-intelligence Microsoft Journal-watch
Video Friday: This Floor Lamp Will Do Your Chores
IEEE Spectrum robotics has released its weekly roundup of notable robotics videos, along with a calendar of upcoming events in the field. Key events include the International Conference on Robotics and Automation (ICRA) scheduled for June 1-5, 2026, in Vienna, and the Robotics Science and Systems (RSS) conference from July 13-17, 2026, in Sydney. Among the highlights is Lume, a sculptural floor lamp that reportedly can perform household chores such as making beds and folding laundry, though skepticism surrounds its capabilities due to the presentation of its functionality in promotional materials. Additionally, researchers from MIT Media Lab and Politecnico di Bari have introduced Electrofluidic Fiber Muscles, a new type of artificial muscle that is soft and flexible, utilizing electric fields to operate silently without external pumps. Another innovation is GEN-1, a general-purpose AI model that significantly improves success rates in physical tasks and enhances speed, marking a step towards achieving generalist intelligence in robotics. The report also discusses advancements in legged manipulators, which face challenges in interacting with diverse articulated objects, and highlights the development of Tether, a system for autonomous interactive play that enhances policy learning through efficient data generation. As the robotics community continues to innovate, these developments signal exciting possibilities for the future of robotics applications.
Home-robots Video-friday Artificial-muscle Agricultural-robots Robot-ai Quadruped-robots
Video Friday: Beep! Beep! Roadrunner Bipedal Bot Breaks the Mold
IEEE Spectrum robotics has released its weekly roundup of notable robotics videos and events. Among the highlights is the introduction of "Roadrunner," a new bipedal wheeled robot prototype that can switch between various locomotion modes, designed for enhanced navigation. Weighing approximately 15 kg, it features symmetric legs that can adapt for obstacle avoidance and movement management. NASA has announced two ambitious missions: SkyFall, which will deploy next-generation helicopters on Mars to scout landing sites and map subsurface water ice, and MoonFall, aimed at preparing for future Artemis missions by sending drones to explore the lunar South Pole. These drones will operate independently for 14 Earth days, surveying challenging terrains. In research advancements, a team from MIT has developed Electrofluidic Fiber Muscles, a new class of soft and flexible artificial muscles for robots and wearables, promising improved agility and integration into textiles. Additionally, the open-source quadruped robot MEVIUS2 has been unveiled, capable of climbing stairs and steep slopes. Other innovations include a wristband from MIT that allows users to control a robotic hand through their own movements, and a cooking robot from Zhejiang Lab that autonomously processes ingredients and performs cooking tasks with high precision. The CMU Robotics Institute is set to host a seminar by Hadas Kress-Gazit from Cornell, focusing on the role of formal methods in robotics amidst the rise of big data.
Video-friday Nasa Bipedal-robots Quadruped-robots Artificial-muscles Humanoid-robots
Video Friday: Humanoid Learns Tennis Skills Playing Humans
IEEE Spectrum robotics has released its latest edition of Video Friday, showcasing a variety of innovative robotics videos and announcing upcoming events in the field. Notable events include the International Conference on Robotics and Automation (ICRA) scheduled for June 1-5, 2026, in Vienna, and a Summer School on Multi-Robot Systems from July 29 to August 4, 2026, in Prague. Among the featured advancements, researchers have developed LATENT, a system designed to teach humanoid robots tennis skills by learning from imperfect human motion data. This innovation addresses the challenges of replicating human-like athleticism in robotics. Additionally, a breakthrough has been achieved in robotic manipulation, with a robot successfully peeling an apple using dual dexterous hands, showcasing significant progress in bimanual tasks. The development of MoDE-VLA, a control system that integrates vision, language, force, and touch data, further enhances the robot's ability to perform complex tasks with stability and precision. This shared-autonomy approach allows human operators to guide robots in executing intricate movements. In other highlights, collaborations between Tesollo and Hanyang University have led to advancements in robotic hand technology, while the Fluent Robotics Lab at the University of Michigan is set to present a paper on operational PR2 robots. The KAIST DRCD Lab has also demonstrated the capabilities of its humanoid robot, trained through deep reinforcement learning. As robotics continues to evolve, these innovations reflect the ongoing efforts to bridge the gap between human-like dexterity and robotic functionality.
Humanoid-robots Video-friday Robot-locomotion Nvidia Robot-manipulation Quadruped-robots
Video Friday: Robot Dogs Haul Produce From the Field
IEEE Spectrum's weekly feature, Video Friday, showcases a variety of innovative robotics videos and highlights upcoming robotics events, including the International Conference on Robotics and Automation (ICRA) scheduled for June 1-5, 2026, in Vienna. This week’s selection includes demonstrations of the Lynx M20 robots, which are designed to address the logistical challenges of transporting harvested crops in mountainous regions. Research from a collaboration between the Max Planck Institute for Intelligent Systems, the University of Michigan, and Cornell University reveals that magnetic microrobot swarms can manipulate larger objects without direct contact, showcasing their potential for complex tasks such as assembly and movement of small items. Meanwhile, Georgia Tech is investigating how bipedal robots can recover from balance loss in unpredictable environments, aiming to enhance their functionality in real-world applications. In a separate initiative, Carnegie Mellon University's TartanAUV team is refining their autonomous underwater vehicle, Osprey, in preparation for the annual RoboSub competition. Additionally, advancements in tilt-rotor aerial robots are being explored to improve control and maneuverability through reinforcement learning techniques. The feature also includes educational tools like the Astorino robot, designed for teaching robotics in schools, and discussions on the need for more realistic datasets for autonomous driving. Overall, the content reflects the ongoing evolution and application of robotics across various fields, emphasizing both technical advancements and educational initiatives.
Humanoid-robots Video-friday Swarm-robotics Quadruped-robots Farm-robots Bipedal-robots
NVIDIA Open-Sources DreamDojo: A 44,000-Hour "Dream" to Solve the Robotics Data Gap
NVIDIA GEAR Lab has unveiled DreamDojo, an innovative open-source world model that leverages a substantial dataset of 44,000 hours of human egocentric videos. This advanced model employs "latent actions" to effectively connect human movements with robotic actions, enabling it to achieve zero-shot generalization and real-time controllability for applications in teleoperation and planning. The release of DreamDojo marks a significant advancement in the field of robotics, enhancing the potential for seamless interaction between humans and machines.
Dr Jim Fan NVIDIA World-Models open-source world-model
Video Friday: Multitasking Robots Smoothly Do the Things Together
IEEE Spectrum robotics has released its weekly roundup of notable robotics videos and upcoming events, including the ICRA 2026 conference scheduled for June 1-5, 2026, in Vienna. Among the highlights, Westwood Robotics unveiled THEMIS Gen2.5, the first commercial full-size humanoid robot capable of walking and manipulating objects simultaneously. This advancement builds on Helix's previous work, which demonstrated a single neural network controlling a humanoid's upper body, now expanded to encompass the entire robot's functions. In a demonstration of practical applications, Kimberly Elenberg from Carnegie Mellon University showcased how data from robotic responders can enhance life-saving efforts during mass casualty incidents. Meanwhile, Sphero continues to thrive in the competitive educational robotics market since its inception in 2011. Innovative flight testing methods were discussed by Zipline, emphasizing the importance of testing drones in extreme conditions. Additionally, researchers from the University of Tokyo introduced a concept of 3D-printing both skin and skeleton, while LimX presented small bipedal robots capable of skiing and resembling dinosaurs. The EPFL Reconfigurable Robotics Lab introduced a novel user-guided control system for modular robots, demonstrating its effectiveness through various tasks. Texas A&M University showcased its Quadrotor Biplane Tailsitter (QBiT) UAVs, which combine the agility of quadrotor drones with the efficiency of fixed-wing aircraft. Lastly, DARPA announced a new challenge aimed at developing drones capable of carrying payloads exceeding four times their weight, promising to transform drone usage across multiple sectors.
Humanoid-robots Video-friday Commercial-robots Drones Educational-robots Bipedal-robots
Video Friday: Bipedal Robot Stops Itself From Falling
IEEE Spectrum robotics has released its latest edition of Video Friday, showcasing a variety of innovative robotics videos and announcing upcoming events in the field. Among the highlights is the Robotic Autonomy in Complex Environments with Resiliency (RACER) program, which is nearing completion after extensive collaboration with the U.S. Army and Marine Corps. This program is expected to leave a lasting impact on military operations and stimulate private-sector investment in autonomous technologies. Notable advancements include the introduction of COSA, a cognitive operating system that enhances humanoid robots' capabilities for high-level cognition and motion control. Meanwhile, the 1X World Model has made significant strides in robot learning, allowing its NEO model to perform tasks autonomously based on voice or text prompts, even for unfamiliar objects. In assistive technology, the GuideData Dataset has been launched to improve interactions between guide dog trainers and visually impaired individuals, aiming to enhance mobility and safety. Additionally, Fourier's Care-Bot prototype is gaining attention for its interactive features at CES 2026. In environmental monitoring, ETH Zurich has developed an autonomous quadruped robot for volcanic gas measurements, successfully tested on Mount Etna. Humanoid robots have also made progress in industrial logistics, completing proof-of-concept testing at Siemens's factory in Erlangen. Columbia Engineers have created a robot capable of learning facial lip motions for speech and singing through observational learning, marking a significant milestone in robotics. Lastly, DEEP Robotics showcased its quadruped robots' capabilities in complex firefighting scenarios, while Synapticon introduced its POSITRON platform to enhance safety in humanoid robots for real-world applications.
Robotics Video-friday Bipedal-robots Humanoid-robots Quadruped-robots Industrial-robots
NVIDIA DLSS 4.5, Path Tracing and G-SYNC Pulsar Supercharge Gameplay With Enhanced Performance and Visuals
At the CES trade show, NVIDIA unveiled its latest advancement in gaming technology, DLSS 4.5, which features Dynamic Multi Frame Generation. This new capability includes a 6X Multi Frame Generation mode and a second-generation transformer model for DLSS Super Resolution. The enhancements aim to provide gamers with an improved visual experience and smoother gameplay. By leveraging cutting-edge AI technology, NVIDIA continues to push the boundaries of gaming performance, addressing the growing demand for high-quality graphics and responsiveness in modern video games. The announcement is part of NVIDIA's ongoing commitment to innovation in the gaming industry, showcasing their leadership in developing tools that enhance user experience.
Build AI Scales to 100,000 Hours as Data Scaling Becomes Robotics’ New Frontier
Build AI has secured $15 million in new funding to enhance its efforts in developing universal robot intelligence, focusing on the integration of human video data. This significant investment comes as the company aims to leverage a vast new dataset, which it believes is crucial for advancing the capabilities of robots. By utilizing human video, Build AI intends to improve the learning processes of robots, making them more adaptable and intelligent. The funding will support the company's ongoing research and development initiatives, positioning it at the forefront of innovation in artificial intelligence and robotics.
Data Collection Build AI embodied-ai
Physical Intelligence Finds 'Emergent' Bridge Between Human Video and Robot Action
A $5.6 billion startup has announced a breakthrough in artificial intelligence, revealing that sufficiently large robot models can autonomously learn to comprehend human video. This development has the potential to address a significant data bottleneck that has long plagued the industry. The startup's findings, which were shared recently, highlight the capabilities of advanced machine learning techniques in enhancing the understanding of visual content. By leveraging large-scale models, the company aims to improve the efficiency and effectiveness of AI systems in processing and interpreting video data, paving the way for more sophisticated applications in various fields. This innovation could transform how machines interact with visual information, ultimately leading to more intuitive and responsive AI technologies.
Data Collection Physical Intelligence embodied-ai
Research Spotlight: X-Humanoid ‘Robotizes’ Human Videos to Train the Next Generation of Androids
Researchers at the National University of Singapore have unveiled an innovative generative video pipeline designed to transform third-person footage of human activities into synthetic training data for humanoid robots. This groundbreaking development aims to address the embodiment gap in robotics, enabling more effective training of robots by providing them with diverse and realistic scenarios. The project, which leverages advanced video synthesis techniques, represents a significant advancement in the field of robotics and artificial intelligence. By creating a scalable solution for generating training data, the researchers hope to enhance the capabilities of humanoid robots, making them more adept at understanding and interacting with the world around them.
Data Collection embodied-ai
Build AI Open-Sources 10,000 Hours of Factory-Worker Video to Scale Robot Learning
A robotics startup has unveiled Egocentric-10K, which it claims to be the largest egocentric video dataset ever created. This extensive collection was gathered exclusively from real factory environments and aims to address the challenges associated with the "physical AI bottleneck" by utilizing human-generated data. The release of this dataset marks a significant advancement in the field of robotics and artificial intelligence, providing researchers and developers with valuable resources to enhance machine learning algorithms and improve AI performance in physical tasks.
Build AI
Unitree's New ''Embodied Avatar'' Is More Than a ''Real Steel'' Toy—It's a Data Engine
Unitree has introduced an innovative full-body teleoperation platform designed for its G1 humanoid robot. This unveiling comes alongside the release of new videos demonstrating the robot's capabilities in low-latency boxing and soccer. However, the primary objective of this advanced system appears to be its function as a "data acquisition platform," aimed at addressing the existing physical AI bottleneck within the industry. By enhancing data collection and interaction, Unitree seeks to propel advancements in artificial intelligence and robotics.
G1 Unitree Robotics
The Physical AI Bottleneck: Comparing the Data Strategies of 1X, Figure, Tesla, and Neura
A recent report by the LA Times reveals that leading robotics companies are engaged in a significant, low-tech initiative to collect real-world data, which has emerged as a critical challenge in the field. As of October 2023, these companies are exploring various strategies to address this data bottleneck, with approaches ranging from human-video capture and teleoperation to extensive simulation techniques. The diversity in methods reflects the industry's urgent need to enhance robotic capabilities and improve performance in real-world applications.
Data Collection 1X-technologies Tesla Figure Neura Robotics Neura Gym
Game Science reaches $1 billion in revenue, climbs to 10th place on Steam publishers list
Game Science, a Chinese video game developer, has achieved a significant milestone by becoming the 10th highest-grossing publisher on Steam, amassing one billion dollars in revenue. This accomplishment positions the company ahead of renowned industry giants such as Rockstar, known for the Grand Theft Auto series, 2K, the developer behind the NBA franchise, and Capcom, famous for the Monster Hunter games. The success is largely attributed to the anticipation surrounding their upcoming action role-playing game, Black Myth: Wukong, which has generated considerable excitement among gamers. This financial achievement highlights the growing influence of Chinese game developers in the global market, showcasing their ability to compete with established Western companies.
News Feed
TuSimple to produce animation and game based on Three-Body Problem, its first major project since delisting
TuSimple, a provider of self-driving truck technology, announced on Thursday its plans to produce the first animated feature-length film and video game inspired by the acclaimed novel "The Three-Body Problem." This project follows the novel's previous adaptations for drama by Netflix and Tencent Video. The collaboration marks a significant expansion for TuSimple, which recently delisted from the Nasdaq in February, reflecting its strategic shift in focus. The initiative aims to leverage the novel's popularity and complex themes to engage a broader audience through innovative storytelling in both film and gaming formats.
News Feed
Sync and Stream: GeForce NOW Connects to Members’ Game Libraries Across Devices
GeForce NOW, a cloud gaming service, allows users to play their favorite titles from various popular game libraries while keeping their progress synchronized across multiple devices. This innovative platform enables gamers to seamlessly transition between devices, enhancing their gaming experience by providing flexibility and convenience. With access to a wide range of games, GeForce NOW caters to the growing demand for cloud-based gaming solutions, making it easier for players to engage with their favorite titles anytime and anywhere. As the gaming industry continues to evolve, services like GeForce NOW are becoming increasingly relevant, offering solutions that meet the needs of modern gamers.
Huawei's Young Talent Wang Yuxin Launches New Venture, Secures Millions in First Month: Can Streaming Video Open New Horizons?
Wang Yuxin, a notable figure from Huawei and a key member of Yuanstone Technology, has launched a new venture named Xingjie Intelligent, dedicated to advancing 'streaming video generation.' Within just one month of its inception, the company has successfully raised millions in angel funding. Xingjie Intelligent aims to transform video interactivity by enabling real-time user engagement, thereby surpassing conventional video creation techniques. This innovative strategy is designed to tackle the evolving challenges associated with content creation in interactive environments.
Streaming Video AI Technology Interactive Media Video Generation
Perceptron Mk1 shocks with highly performant video analysis AI model 80-90% cheaper than Anthropic, OpenAI & Google
Perceptron Inc., a Bellevue-based startup founded by former Meta researchers Armen Aghajanyan and Akshat Shrivastava, has launched its flagship video analysis model, Mk1, aimed at revolutionizing how enterprises utilize AI in real-time video processing. Announced today, this innovative model is priced significantly lower than competitors, at $0.15 per million input tokens and $1.50 per million output tokens, making it accessible for large-scale industrial applications. The Mk1 model, developed over 16 months, excels in understanding complex physical interactions and temporal reasoning, outperforming established models like OpenAI's GPT-5 and Google's Gemini 3.1 Pro in various benchmarks. Its unique architecture allows it to process video streams continuously, maintaining object identity and providing precise analysis of dynamic scenes, which is particularly beneficial for sectors such as security, robotics, and marketing. Perceptron aims to position Mk1 as a leader in the "Efficiency Frontier," balancing high performance with cost-effectiveness. The model's capabilities extend to auto-clipping highlights from live sports and enhancing quality control in manufacturing. A public demo site is available for potential users to explore its functionalities. This launch signifies a significant step towards integrating advanced AI into real-world applications, as the company seeks to make "physical AI" as prevalent as its digital counterpart.
Technology
It’s Gonna Be May: 16 Games Hit the Cloud This Month, With More NVIDIA GeForce RTX 5080 Power
As May approaches, GeForce NOW is set to enhance its gaming offerings by adding 16 new titles to its platform. This update includes several highly anticipated AAA games that will be available for streaming on their launch day. Members will have the opportunity to access their PC game libraries instantly from popular services such as Steam, Xbox, PC Game Pass, and GOG. This initiative aims to provide gamers with seamless access to a diverse range of games, enhancing the overall cloud gaming experience.
Tag, You’re It: GeForce NOW Levels Up Game Discovery With Xbox Game Pass and Ubisoft+ Labels
GeForce NOW has announced significant upgrades aimed at enhancing the gaming experience for users. This week, the cloud gaming service introduced smarter library features that simplify the transition of PC game collections into a cloud-based format. The initiative focuses on maximizing the potential of existing game libraries, allowing gamers to access their favorite titles more efficiently. By streamlining the process, GeForce NOW aims to provide a more flexible and accessible gaming environment, catering to the evolving needs of the gaming community. These enhancements reflect the company's commitment to prioritizing gamers and adapting to their preferences in the rapidly changing landscape of cloud gaming.
Game On: Five New Titles Now Streaming on GeForce NOW
GeForce NOW, the cloud gaming platform, has announced the addition of five new titles to its library, allowing gamers to stream the latest games instantly across various devices. This update comes as a timely solution for players looking to tackle their gaming backlog. With the service's ability to deliver high-quality gaming experiences directly from the cloud, users can enjoy seamless access to new content without the need for extensive downloads or installations. The new titles are available for play immediately, enhancing the platform's offerings and catering to the growing demand for flexible gaming options.
NVIDIA Virtualizes Game Development With RTX PRO Server
Game development teams are increasingly navigating larger virtual environments and more intricate production processes, often collaborating across geographically dispersed locations. Despite these advancements, many studios continue to depend on traditional, stationary GPU hardware for essential tasks in their production workflows. This reliance on fixed technology poses challenges in efficiency and adaptability, as the industry evolves to meet the demands of more expansive and complex gaming experiences. As developers gather at the Game Developers Conference, discussions are expected to focus on innovative solutions that could enhance productivity and streamline operations in an era where flexibility and scalability are paramount.
March Into the Cloud With 15 New Games Coming to GeForce NOW
As March progresses, GeForce NOW is expanding its library with the addition of 15 new games, enhancing the cloud gaming experience for users. Among the highlights is "Crimson Desert," an open-world action-adventure game developed by Pearl Abyss. This new influx of titles aims to provide gamers with a diverse selection of experiences, catering to various interests and preferences. The rollout of these games throughout the month is part of GeForce NOW's ongoing commitment to deliver fresh content and keep its platform engaging for its community.
Import AI 447: The AGI economy; testing AIs with generated games; and agent ecologies
A recent exploration into the concept of superintelligence arcologies has sparked interest among futurists and urban planners. This innovative idea envisions self-sustaining, high-density urban environments that integrate advanced artificial intelligence to enhance living conditions and resource management. The discussion gained momentum in October 2023, as experts gathered at a technology conference to examine the potential benefits and challenges of such developments. Proponents argue that superintelligence arcologies could address pressing global issues, including climate change, urban overcrowding, and resource scarcity. By leveraging AI, these structures could optimize energy use, waste management, and transportation systems, creating a more sustainable urban ecosystem. The integration of smart technology would allow for real-time data analysis, enabling residents to make informed decisions about their environment. Critics, however, caution against the risks associated with relying heavily on AI for urban living. Concerns about privacy, security, and the potential for systemic failures in such complex systems were highlighted during the discussions. The feasibility of constructing these arcologies also raises questions about economic viability and societal acceptance. As the conversation continues, researchers and developers are exploring various models and designs for superintelligence arcologies, aiming to create prototypes that could be tested in real-world scenarios. The outcome of these efforts could significantly shape the future of urban living, potentially transforming cities into interconnected, intelligent habitats that prioritize sustainability and quality of life.
All About the Games: Play Over 4,500 Titles With GeForce NOW
The GeForce NOW anniversary celebration continues, highlighting the extensive library of games that has contributed to its success. This week, the cloud gaming service announced the addition of 12 new titles, bringing the total number of supported games to over 4,500. The initiative aims to enhance user experience by providing a diverse selection of games for streaming and sharing, ensuring that players always have fresh content to enjoy. The ongoing celebration underscores the platform's commitment to expanding its offerings and engaging its community of gamers.
Robotics needs a service framework.
RSF defines a common language for robot service capability, lifecycle operations, certification pathways, and service-provider networks.
