Yesterday here in California, we just saw some new legislation. Governor Gavin Newsome signed one of the first bills to regulate AI companies in this state. It is called Transparency and Frontier Artificial Intelligence Act or SB53.It is the first to focus on the safety of these powerful AI models by forcing companies to publish documents that lay out how they are going to uh work to try to create safe AI systems. And it also adds some new protections for whistleblowers who raise concerns about AI health an ...

SB 53 requires large AI labs – including OpenAI, Anthropic, Meta, and Google DeepMind – to be transparent about safety protocols. It also ensures whistleblower protections for employees at those companies. https://t.co/l8J472uW4M ...

Core Viewpoint - NVIDIA has launched the open-source Newton Physics Engine and the open NVIDIA Isaac GR00T N1.6 reasoning vision language action model, enhancing robotics development by providing an accelerated platform for simulation and real-world application [2][4][8] Group 1: Technology Advancements - The Newton Physics Engine is GPU-accelerated and co-developed with Google DeepMind and Disney Research, enabling complex robot simulations [4][5] - The Isaac GR00T N1.6 model integrates NVIDIA Cosmos Reason, allowing robots to interpret vague instructions and perform tasks using prior knowledge and common sense [7][8] - The new dexterous grasping workflow in Isaac Lab 2.3 trains robots in a virtual environment, gradually increasing task complexity [13][14] Group 2: Adoption and Impact - Over a quarter-million robotics developers globally require accurate physics for safe real-world execution of robot skills [3] - Leading research institutions and companies, including ETH Zurich and Boston Dynamics, are adopting NVIDIA's technologies for robotics research and development [5][10][19] - The Cosmos Reason model has been downloaded over 1 million times and is leading in the Physical Reasoning Leaderboard on Hugging Face [8][9] Group 3: Infrastructure and Evaluation - NVIDIA is developing AI infrastructure to support demanding robotics workloads, including the upcoming Cosmos Predict 2.5 and Cosmos Transfer 2.5 models [18] - The Isaac Lab - Arena framework, co-developed with Lightwheel, will allow for scalable experimentation and standardized testing of robot skills [17][19] - The open-source NVIDIA Physical AI Dataset has been downloaded over 4.8 million times, providing extensive data for post-training of Isaac GR00T N models [9]

Collaboration & Technology - Apptronik is collaborating with Google DeepMind on Gemini Robotics 1.5 to deploy Gemini-powered Apollo humanoid robots [2] - The collaboration aims to bring AGI (Artificial General Intelligence) into the physical world [1] - Gemini Robotics VLA (Visual Language Adaptation) is reaching a milestone in embodied intelligence [1] - The project is moving beyond reactive models to a new era of robotic autonomy through agentic capabilities like reasoning, planning, and tool use [1] Deployment & Application - Apptronik is preparing to deploy Gemini-powered Apollo humanoid robots in additional customer facilities [2] - The goal is to transform a powerful model into a field-ready system with consistency, reliability, and purpose [2]

Core Insights - Figure has raised $1 billion in Series C funding, achieving a post-money valuation of $39 billion, with significant participation from top investment firms [1][3] - The company aims to address labor shortages and enhance human welfare through its humanoid robots, following a master plan similar to Tesla's [3][6] - Figure's vertical integration strategy combines hardware and software, creating a competitive advantage and high barriers to entry in the embodied intelligence sector [6][10] Funding and Valuation - Figure's Series C funding was led by Parkway Venture Capital, with participation from firms like NVIDIA, Intel Capital, and Salesforce [1] - The company previously secured $675 million in Series B funding from investors including Microsoft and OpenAI Startup Fund [1][3] Company Strategy and Vision - Founded by Brett Adcock, Figure's ultimate goal is to reduce labor costs and improve quality of life by integrating robots into the workforce [3][6] - The company's master plan includes developing a fully functional humanoid robot, achieving human-like control, and integrating these robots into the labor market [3][5] Competitive Advantages - Figure's vertical integration allows it to control all aspects of production, from proprietary chips to a closed operating system, similar to Apple and Tesla [6][10] - The company has developed its own humanoid robots and models, along with in-house manufacturing capabilities, creating a self-reinforcing cycle of efficiency and cost reduction [8][10] Technological Innovations - Figure's robots utilize a new battery system with a capacity of 2.3 kWh, supporting peak performance for up to 5 hours, and reducing costs by 78% compared to previous models [9][10] - The Helix intelligent system enables the robots to perform various tasks without task-specific tuning, facilitating large-scale deployment and coordination [10][11] Manufacturing Capabilities - The BotQ factory is designed to produce up to 12,000 humanoid robots annually, integrating automation into the assembly process [10][11] - The use of robots in the production line allows for data accumulation that can enhance the Helix system, creating a large-scale dataset for further optimization [11][12] Industry Context - The embodied intelligence sector is highly competitive, with many companies focusing on specific parts of the value chain, while Figure's comprehensive approach sets it apart [6][8] - The integration of AI and hardware is seen as a transformative force in the industry, with potential implications for the future of manufacturing and labor [12]

点击下方 卡片 ，关注" 具身智能 之心 "公众号 编辑丨具身智能之心 所有内容出自国内首个具身智能全栈学习社区：具身智能之心知识星球。国庆优惠，欢迎和近2000名成员 一起交流具身产业与学术。 Real2Sim2Real近3年工作一览 论文题目： Incremental Few-Shot Adaptation for Non-Prehensile Object Manipulation using Parallelizable Physics Simulators 论文链接：https://arxiv.org/pdf/2409.13228? 论文时间： ICRA 2025 作者单位： 马克斯·普朗克智能系统研究所 论文题目： RL-GSBridge: 3D Gaussian Splatting Based Real2Sim2Real Method for Robotic Manipulation Learning 本文只做学术分享，如有侵权，联系删文 由于真实数据采集成本高，国内外具身领域有不少团队在研究real2sim、Real2Sim2Real 相关工作。和一些 具身公司坚定走真机采集路线不同，他们相信 ...

AI Engineering & Industry Leaders - Neo4j's Co-Founder and CEO discusses "The State of AI Engineering" [1] - Docker focuses on "Democratizing AI Agents: Building, Sharing, and Securing Made Simple" [1] - GitHub addresses "Building MCP's at GitHub Scale" [1] - H Company is assembling open source bricks for the next generation of AI [1] - Google DeepMind shares updates on generative AI [1] AI Infrastructure & Tools - Koyeb explores "Building for the Agentic Era: The Future of AI Infrastructure" [1] - Black Forest Labs presents "Inside FLUX, How It Really Works" [1] - LlamaIndex is building an open-source NotebookLM alternative [1] Open Source & Community - Hugging Face reports on the "State of Open LLMs in 2025" [1] AI Applications & Techniques - Arize AI studies "System Prompt Learning for Agents" [1] - ZML is working "Towards unlimited contexts: faster-than-GPU sparse logarithmic attention on CPU" [1] - Kyutai is scaling real-time voice AI [1]

Over at the UN, more warnings about potentially apocalyptic civilization ending scenarios if AI doesn't get some strong guard rails soon. This time we are talking more than scientists. We are talking Nobel Prize winners, even two of the three so-called godfathers of AI are raising the alarms, urging political leaders across the world to put some meaningful regulation up by the end of next year.And here's what one Nobel Peace Prize winner said earlier today on the floor. We urge your governments to establish ...

RT Pushmeet Kohli (@pushmeet)Alongside the US-UK tech deal, I'm thrilled to share the partnership between @GoogleDeepMind and @UKAEAofficial on developing physics simulation models (TORAX) to accelerate AI research on fusion energy reactors. #FusionEnergy #AIforGoodhttps://t.co/YVX7bTTEbt ...

Core Insights - Thinking Machines Lab has achieved a remarkable valuation of $12 billion (approximately 85 billion RMB) within just seven months of its establishment, despite not having launched any formal products or having actual users [1][3] - The company, founded by former OpenAI CTO Mira Murati, has successfully completed a $2 billion seed funding round, attracting investments from major industry players like AMD and NVIDIA, positioning itself as a potential competitor to leading firms such as OpenAI, Anthropic, and Google DeepMind [1][3][4] Company Overview - Thinking Machines Lab focuses on multimodal foundational models and next-generation human-machine collaboration, with a core team of around 30 members, two-thirds of whom are from OpenAI [3][4] - The company has established a partnership with Google Cloud for computing power and plans to release its first product, which will include open-source components, in the coming months [3][4] Investment Dynamics - The investment landscape has shifted towards a GPU arms race, with Thinking Machines Lab securing a significant allocation of NVIDIA and AMD GPUs, which are critical for training large models [4][6] - The valuation reflects not just potential revenue but also the strategic positioning within the AI ecosystem, as the company is seen as a last major opportunity for investors to back a team with OpenAI's core decision-makers [5][6] Research and Development Focus - Thinking Machines Lab has adopted a "technology-driven" approach, using research publications and blogs to showcase its advancements in the field, which serves as a new model for AI startups [2][7] - The company recently published a paper addressing the non-determinism in large language model (LLM) inference, highlighting the importance of output stability and predictability for user trust and system reliability [7][8][10] Industry Implications - The focus on output consistency and predictability is crucial for high-risk sectors such as healthcare and finance, where user trust is paramount [10][12] - The insights from Thinking Machines Lab's research may lead to a shift in industry standards, emphasizing the need for "deterministic AI" and potentially creating a certification system for trustworthy AI [12][14] Future Trends - The AI industry is expected to evolve towards more efficient and interpretable model architectures, moving away from merely increasing parameter counts [13][14] - There will be a growing emphasis on energy efficiency and sustainable practices in AI model deployment, with expectations for significant reductions in energy consumption by 2027 [14]