10 results on this page
The article examines whether Unitree can become the defining company in embodied AI amid a surge of 93.5 billion yuan in funding. It argues that embodied-AI financing is entering a more rational phase, with investor enthusiasm facing closer scrutiny.
Ant Group’s quarterly profit barely grew after the company increased spending on artificial intelligence. The investments target healthcare, large language models, and payment services, while expansion costs also pressured results.
This survey frames self-evolving LLM agents as dynamic graphs whose memories, tools, skills, workflows, and relationships change over time. It presents four evolution taxonomies, connects nine dynamic-graph-learning fields to agent capabilities, and proposes graph-aware evaluation and governance protocols.
The Association for Advancing Automation says non-automotive industries accounted for 54.8% of North American robot orders in the first half of 2026. The shift is being driven by applications in semiconductors, life sciences, pharmaceuticals, food, and consumer products, while AI, collaborative robots, and mobile robots expand the market.
Alipay is positioning its Abao agent as a new commerce gateway where services are selected and executed through natural-language intent rather than ads, search rankings, or app navigation. Abao has connected more than 10,000 AI-enabled services across eight scenarios, including transportation, dining, travel, logistics, and government services.
A study found that LLM recommendations could persuade evaluators to reject promising innovations or approve weak ones. Providing narrative explanations made participants more likely to follow incorrect AI decisions, while unexplained recommendations led to better independent judgment.
Apple Machine Learning presents a lexical-intervention approach for improving cross-lingual knowledge transfer when target-language data is scarce. The work targets downstream capabilities such as scientific reasoning, commonsense inference, and world knowledge without relying heavily on parallel data, translation systems, or auxiliary models.
GitLab 19.3 introduces the Flow Creator agent, allowing users to describe custom automations in plain language and receive complete, runnable flow definitions. The agent clarifies missing details before generation, while existing permissions and approval controls remain in place.
Gemini Live can now start Deep Research reports through voice commands. It can work on reports in the background and let users discuss the findings when they are ready.
A study of roughly 1.8 million fitted SIRENs finds that applying exact function-preserving symmetries reproduces 79.1 of the 80.4 accuracy points separating shared- from random-initialization models. The results show symmetry is sufficient to explain the degradation, while cautioning that this does not prove naturally occurring symmetry causes the entire gap.