- **Nous Research** secured $75M in funding at a $1.5B valuation, led by Robot Ventures and USV, with strong investor interest. - **Hermes**, its open-source AI agent, offers built-in skills (coding, image analysis) and auto-learning, competing with OpenClaw. - Hermes gained 214k GitHub stars, with cloud-hosted tiers ($20–$200/month) for easier access. - New funds aim to expand Hermes’ products and business model, leveraging its popularity and open-source adoption.

- TOI Science Desk is a team of journalists dedicated to exploring science, offering news, features, and articles on topics like genetics, space, and AI. - They aim to demystify complex scientific concepts, making them accessible and engaging for readers of all backgrounds. - The desk focuses on connecting readers to the latest scientific discoveries and innovations through storytelling. - Their mission is to deliver daily insights, blending precision and passion to illuminate the wonders of the scientific world.

- Tesla is developing a purpose-built, wheelchair-accessible autonomous vehicle in Texas, aiming to address mobility challenges for disabled users. - Current Tesla fleets use non-accessible Model Ys and Cybercabs, with limited accessibility features like braille controls and higher seating. - No U.S. robotaxi company offers fully driverless, wheelchair-accessible services, though some use human operators or partner fleets. - Cruise’s 2023 wheelchair-accessible prototype faced setbacks after a pedestrian collision, halting its self-driving service and funding.

- Ground robots in Ukraine now evacuate the wounded, hold trenches, and conduct combat missions, replacing earlier supply roles. - Ukrainian forces captured a Russian-held position using drones and robots, avoiding direct soldier risk. - Aerial drones evolved faster with IT expertise, while ground robots rely on practical, hands-on engineering by infantry teams. - Ukraine’s tech innovations, like 3D-printed drones and naval systems, disrupt Russian military operations globally.

Here are the article's main points in four clear bullet points: • General Intuition CEO Pim de Witte believes embodied AI will follow a similar pattern to foundation models, focusing on better quality datasets that can produce general-purpose models. • The company built its own foundation model by training on millions of hours of video game data and has demonstrated it can play a video game for hours and power a quadrupedal robot with just eight minutes of real-world robotics data. • General Intuition's goal is to become the foundation model of physical AI, providing a base model for other robotics companies to build upon, rather than building robots itself. • The startup has raised $320 million at a $2.3 billion valuation and aims to make it 10 times easier for others to build self-driving car or other robotic machines by developing its general-purpose foundation model.

Here are four clear bullet points summarizing the news article: • Agility Robotics, a humanoid robotics startup, plans to go public through a merger with Churchill Capital Corp XI, valuing the company at $2.5 billion. • The deal will generate over $620 million in proceeds, including $200 million from new and existing investors, which Agility will use to expand production capacity, fulfill orders, and target new customers. • The combined company will trade under the ticker symbol AGLT on a North American stock exchange, with plans to increase production of its next-generation Digit v5 robot, which has secured over $300 million in multi-year orders. • Agility CEO Peggy Johnson says humanoid robots are poised to drive productivity, supply chain resilience, and American technology leadership, helping enterprises address labor shortages and improve efficiency.

Here are four concise bullet points summarizing the article: • Alibaba launches Qwen Robot Suite to push AI beyond chatbots and into physical environments through "embodied intelligence". • The suite consists of three interconnected layers: Qwen-RobotNav (vision-language navigation), Qwen-RobotWorld (video world model), and Qwen-RobotManip (VLA model for action). • The robot intelligence is designed to perceive, reason, and interact with physical environments, enabling machines to understand and move through spaces. • The suite has entered pilot testing with selected Alibaba Cloud enterprise clients, marking Alibaba's latest foray into embodied AI.

Here are four clear bullet points summarizing the article: • **AI-powered robots will revolutionize industries**: In the next few decades, billions of autonomous AI-powered robots will work alongside humans in factories, warehouses, healthcare, and more, performing tasks that were previously tedious or difficult. • **Single breakthrough unlikely**: The authors, experts in robotics and AI, emphasize that a single magical moment won't solve robotics challenges; instead, it'll take years of hard work and the development of coordinated systems using various AI tools. • **AI enables robots to learn and adapt**: Thanks to AI, robots can now learn to operate in real-world environments, perceive and understand their surroundings, reason about them, and perform tasks that are useful, reliable, and safe. • **Robotics advancements will be incremental**: The authors predict that robotics advancements will be driven by the well-engineered application of coordinated systems using different AI tools, rather than a single ChatGPT-style breakthrough.

Here are four clear bullet points summarizing the article: • Genesis AI has unveiled its first model, GENE-26.5, which comes with surprise hands that can perform advanced tasks like cooking and playing piano. • The startup's robotic hand is designed to mimic a human hand, reducing the "embodiment gap" in robotics research and allowing for more data collection. • Genesis' CEO Zhou Xian and President Théophile Gervet hope that building their own robotic hands will give them an upper hand in the industry, with the goal of creating a full-body robot shortly. • The startup is also developing a sensor-loaded glove to collect data, which could be used to train models for various applications like lab work, and plans to reveal its first general-purpose robot soon.

Here are four clear bullet points summarizing the article: • Eka, a startup in Cambridge, Massachusetts, has developed a robot arm that can perform tasks with surprising naturalness and dexterity, including screwing in a light bulb. • The company's approach involves training robots using reinforcement learning, which combines experimentation with positive and negative feedback, to develop real-world problem-solving skills. • Eka's founders, Pulkit Agrawal and Tuomas Haarnoja, believe that their approach can enable robots to achieve human-like dexterity, revolutionizing the way they are used in industries such as manufacturing, logistics, and service sectors. • The company is focused on scaling up its approach to solve the problem of robot dexterity, which could have significant implications for various aspects of life, from factory floors to household chores.

Here are four clear bullet points summarizing the article: • Lightning, an android robot, won a half-marathon race in Beijing by finishing in 50 minutes and 26 seconds. • The robot's time was faster than any human record holder, including Jacob Kiplimo of Uganda, who finished in 57 minutes and 20 seconds. • Despite falling during the final stretch, Lightning picked itself up with human help and continued to run across the finish line. • The android was designed by Chinese smartphone brand Honor and features legs about three feet long, as well as a liquid-cooling system adapted from consumer electronics.

Here are four clear bullet points summarizing the article: • A humanoid robot from Chinese smartphone maker Honor won a Beijing half-marathon for robots in 50:26 minutes, significantly faster than the human world record. • The winning robot was autonomous, while some other participating robots were remote-controlled; about 40% of robots competed autonomously, with varying levels of success. • One Honor robot finished even faster at 48:19 minutes, but it was remotely controlled; the winning robot's achievement is notable given its autonomy. • This year's results represent a significant improvement over last year's event, when the fastest robot took around 2 hours and 40 minutes to complete the course.

Here are four clear bullet points summarizing the news article: • Chinese-made humanoid robots showed off their athletic and autonomous navigation skills at the Beijing E-Town Half Marathon and Humanoid Robot Half Marathon, with several robots beating human winners by over 10 minutes. • The winning robot, developed by Honor, finished in 50:26 seconds, setting a world record, while teams from Honor took the top three podium spots with self-navigated robots posting world-record-beating times. • The event highlighted the rapid technical advances in the robotics sector, with experts saying that humanoid robots have the potential to reshape industries such as manufacturing and battlefield combat. • While the robots' physical prowess is impressive, experts note that developing AI software that enables humanoids to match human factory workers' efficiency remains a challenge for Chinese robotics firms.

Here are the article's main points in four concise bullet points: • Chef, a robotics company, has reached 100 million servings using AI-powered robot arms to automate large-scale food production, primarily serving enterprise customers like Amy's Kitchen and school lunch providers. • The company has pivoted from fast casual restaurants to focus on larger, institutional-scale customers, such as airline catering companies, ghost kitchens, stadiums, and prisons, with plans to expand into these markets. • Chef is using the data generated from its 100 million servings to improve its AI models for food handling and packaging, aiming to make the robots more efficient and scalable. • The company's CEO, Rajat Bhageria, believes that by leveraging AI and robotics, his industry can overcome previous challenges in automating food production and achieve long-term growth.

Here are four clear bullet points summarizing the article: • Robotic "workers" have reduced the workload of human staff in at least three restaurants in eastern China by 60%. • The robots handle tasks such as ordering, serving, cleaning, and cooking, with eight robots used at one restaurant. • Before taking orders, the robots analyze customers' faces, tongues, and questionnaires to create reports on their lifestyle, emotions, and digestion status. • The robots can cook over 100 dishes, including traditional Chinese recipes like Three Cup Chicken and Braised Pork Trotters.

Here are four clear bullet points summarizing the article: • Shanghai Gigafactory could be a "golden key" to mass-producing humanoid robots, according to Tesla China head Allan Wang Hao. • The factory can manufacture new products, including robots, and contribute to Tesla's company goals. • This marks the first public mention of using the Shanghai factory for building humanoid robots, a strategic focus set by Elon Musk. • Tesla aims to transform into a physical AI firm from a hardware-centric business, leveraging its edge in artificial intelligence and autonomous driving.