Robotics

Today's Stories

1. 1

   Tesla deploys robotaxi in Dallas and Houston

   Tesla launched its robotaxi service on the streets of Dallas and Houston, fulfilling a decade-old promise to operate autonomous vehicles as a ride-hailing service. The move validates Tesla's long-standing bet on fully autonomous driving as a business model, marking a tangible shift from announcement to real-world deployment after years of delay.

   The service's expansion beyond the two initial cities and whether it achieves meaningful adoption levels that justify the autonomous vehicle investment.

2. 2

   China's Orca world model matches specialized robotics systems without action labels

   Researchers at BAAI (Beijing Academy of Artificial Intelligence) developed Orca, a world model that learns how scenes change from unlabeled video and text descriptions, then uses frozen core weights with swappable output modules to generate text, images, and robot commands. On robot manipulation tasks—shelving books, stacking bowls, scooping sugar—Orca-4B matched π0.5, a system built specifically on robot data, despite never seeing action labels during pre-training. World models that build a shared internal understanding of cause-and-effect could reduce the need for large labeled action datasets, a chronic bottleneck in robotics. Orca also outperformed larger specialized models (Emu3.5 at 34B parameters, FLUX.2, OmniGen2) on text and image prediction benchmarks, suggesting that a well-trained state representation can serve multiple downstream tasks without retraining the core.

   Orca was trained on only one-tenth of available video data—125,000 hours of footage, 160 million event descriptions, and 11.5 million question-answer pairs—and the researchers note that a native world model trained from scratch on sound, force, and touch signals remains their end goal.

3. 3

   South Korean startup Mobilint pushes NPUs for robots, drones, autonomous vehicles

   Mobilint, a South Korean AI semiconductor startup, is building neural processing units (NPUs)—specialized chips for edge devices—to power physical AI applications like robots, autonomous vehicles, and drones, as AI work shifts from cloud computing to on-device processing. Physical AI (AI embedded in machines that interact with the physical world) represents a new frontier beyond cloud-based language models; edge processing means faster response times and lower latency, which are critical for real-time autonomous systems. For businesses deploying robots or autonomous vehicles, this shift may reduce their dependence on cloud infrastructure and improve performance.

   Mobilint CEO Shin Dong-joo is urging South Korea to accelerate development in this area, suggesting the company sees government support as key to competing in the emerging physical AI chip market.

4. 4

   Path Robotics deploys AI-guided welding robots with Boston Dynamics' Spot

   Path Robotics, a Columbus, Ohio-based company, has applied AI to optimize robotic welding by identifying torch paths and using real-time vision guidance to maintain optimal movement during welding operations. The company is also deploying Boston Dynamics' Spot quadruped robots into mobile welding applications in shipbuilding. The company's work centers on building adaptive, AI-driven robotic systems designed for real-world production environments, addressing longstanding difficulties in setting up and using robots for welding applications. This approach could improve how manufacturers deploy robots for complex, precision-dependent tasks.

   Path Robotics' strategy focuses on applying physical AI to manufacturing challenges, combining vision-guided torch control with mobile robot platforms to expand welding automation into new shipbuilding workflows.

5. 5

   AI Safety Community Underestimates Executive Power as Control Risk

   An AI safety researcher argues that discussions of AI loss-of-control scenarios typically focus on elaborate mechanisms like AI-developed nanotech or cyber attacks, but overlook a simpler pathway: the existing centralized power of national leaders like the US President and Chinese General Secretary. The observation suggests that the easiest means for any entity to seize permanent AI-enabled power runs through persuading or subverting these already-powerful executives, rather than through the complex technological schemes most safety analyses emphasize. This reframes where governance and safety efforts may need to focus.

   The researcher calls for detailed analysis of the specific mechanisms by which executive power could be leveraged in the US and China, noting that the high-level pattern likely applies across most states that rely on a single leader with control over the security apparatus.

6. 6

   AI² Robotics raises $735M, valuation tops $2.8B on wheeled humanoid push

   AI² Robotics, a Shenzhen-based robotics company, raised about $735 million（約1200億円） in funding that valued the firm past 50 billion RMB, or about $2.8 billion（約4500億円） U.S. Backers include government funds, industrial corporations like Moutai Group, and financial firms such as CICC Capital and GSR Ventures. The diverse investor base signals strategic importance placed on physical AI in China's robotics sector. AI² Robotics differs from most competitors by using a wheeled base instead of legged locomotion on its humanoid-style robots—a design choice that lowers production costs, improves durability, and reduces regulatory friction for public deployment, positioning the company as a leader among China's physical AI firms.

   AI² Robotics is deploying AlphaBot 2 into structured industrial and commercial environments—logistics, manufacturing, biotech, public service, and retail—rather than pursuing consumer or household applications in the near term. The robot operates on the company's proprietary Alpha Brain vision-language-action model for real-time spatial reasoning and multi-step task planning.