Core Insights - The dissolution of Tesla's Dojo supercomputer project marks a significant shift in the company's AI strategy, which was once valued at $500 billion by Wall Street analysts but has now been deemed a failure [1][4][14] - The end of Dojo raises uncertainties and challenges for the next phase of AI development in the automotive industry, prompting companies to reconsider their strategic paths [3][12] Group 1: Dojo Project Overview - Dojo was initiated with the ambition to create a supercomputer specifically optimized for video data, aimed at enhancing Tesla's Full Self-Driving (FSD) capabilities [6][8] - The project faced numerous challenges, including internal team turmoil and technological complexities, leading to delays and ultimately its termination [8][12] - Key personnel, including Peter Bannon and Ganesh Venkataramanan, have left the project, indicating a significant shift in Tesla's AI development team [4][8] Group 2: Industry Implications - The automotive industry is increasingly recognizing AI as a core component of its future, with companies like Xpeng and Li Auto positioning themselves as AI-centric firms [9][11] - The definition of automotive products is evolving from mere transportation tools to intelligent mobile spaces, emphasizing the importance of computational power and AI models over traditional mechanical performance [12][15] - The shift towards AI-driven business models is evident, with a move from one-time hardware sales to sustainable revenue streams such as FSD subscriptions and software services [12][14] Group 3: Future Directions - Tesla plans to rely more on technology partners like NVIDIA and AMD for computational capabilities, indicating a strategic pivot from in-house development to collaboration [14] - The integration of AI in vehicles is expected to deepen, with advancements towards true Level 4/5 autonomous driving and the emergence of generative AI for enhanced in-car interactions [15]

根据得州9月1日即将生效的新法律，自动驾驶汽车网约车运营将被纳入监管，且监管方式与人类驾驶服 务相同，得州监管部门8日表示，这一许可证允许特斯拉在该州的任何地方运营网约车业务，包括"自动 驾驶汽车"，而且不要求特斯拉在车上配备一名安全员。 （央视财经《经济信息联播》）据CNBC报道，美国得克萨斯州监管部门官网信息显示，特斯拉已经正 式获得了共享出行许可，这为特斯拉自主驾驶出租车在得州开展网约车服务打通了关键通道，也标志着 其在自动驾驶商业化运营道路上迈出了重要一步。 △ 8月10日《经济信息联播》视频 根据美国得克萨斯州监管部门官网上的一份名单，特斯拉无人驾驶出租车有限责任公司已获许可成为持 牌"运输网络公司"，加入优步、来福车（Lyft）等网约车公司的行列。 自6月底以来，特斯拉无人驾驶出租车在得州首府奥斯汀开展试运营，为一些受邀的用户在限定区域提 供出行服务，车费统一为每趟4.2美元。试运营车型为特斯拉最新款Model Y。虽然驾驶座无人，但一名 特斯拉公司工作人员会在副驾座位担任安全员。特斯拉首席执行官马斯克在上个月的财报电话会议上表 示，他相信到2025年底，特斯拉可以为美国一半的人口提供自动驾驶网约 ...

Core Viewpoint - Tesla has halted its Dojo chip design project, reallocating resources to focus on AI5 and AI6 chips, which are expected to enhance AI and autonomous driving capabilities [1][3]. Group 1: Project Changes - The Dojo project leader, Peter Bannon, is leaving Tesla as the company shifts its strategy [1]. - Elon Musk confirmed that the decision to stop the Dojo project was due to the inefficiency of spreading resources across two different AI chip designs [6]. Group 2: Chip Production and Partnerships - The AI6 chip will be manufactured by Samsung Electronics, with a contract value of at least $16.5 billion [1]. - The AI5 chip will be produced by TSMC, initially in Taiwan, before moving to Arizona, USA [2]. - TSMC was considered a potential partner for the Dojo project but was ultimately not included, indicating a preference for Korean suppliers [2]. Group 3: Technical Insights - Musk stated that integrating multiple AI5 or AI6 chips on a single motherboard will significantly reduce complexity and costs [3]. - The AI5 chip is expected to provide substantial performance improvements, meeting the needs for Tesla's Full Self-Driving (FSD) architecture [3][7]. - Future efforts will focus on the AI6 chip, which will incorporate features from the Dojo project, such as a 5x5 grid chip arrangement for high-speed data transfer [5][6].

Core Viewpoint - The article emphasizes the importance of individual approaches and methodologies in the field of large language models (LLMs), particularly in the context of fine-tuning and data quality, suggesting that the technical depth of work in this area is highly dependent on personal engagement and practices [5][16]. Data Work - Method 1 involves inheriting training data from colleagues without checking data quality, which may lead to suboptimal results [7]. - Method 2 suggests downloading open-source data to create a "system + query + answer" dataset [8]. - Method 3 focuses on generating data using GPT-4, emphasizing the diversity of prompts and the importance of data quality checks [8]. - Method 4 advocates using user interaction logs to drive data construction, analyzing user feedback to improve answer quality [9]. - Method 5 recommends breaking down complex tasks at the data level to enhance model performance [9]. Training Code - Method 1 involves inheriting training code and making minimal modifications [11]. - Method 2 encourages a thorough understanding of training code parameters and their implications [11]. - Method 3 promotes questioning and improving training code, such as optimizing speed and framework choices [12]. Experimental Analysis - Method 1 suggests running prepared evaluation sets and addressing data quality issues based on results [14]. - Method 2 involves analyzing bad cases from models to identify underlying issues and designing experiments to validate findings [14]. - Method 3 emphasizes the relationship between model results, data quality, and training methods, advocating for a comprehensive analysis of training logs and evaluation results [15]. Community and Collaboration - The article highlights the establishment of a large community focused on various aspects of autonomous driving technology, including large models and multi-sensor fusion, with nearly 4,000 members and over 300 companies and research institutions involved [18].

大家都感觉到自动驾驶技术栈开始趋同，以前大大小小几十个方向都需要算法工程师，现在one model、 VLM、VLA，统一方案的背后其实是更高的技术壁垒。博主一直在鼓励大家坚持、多多交流，但归根结底个 人的力量是有限的。我们希望共建一个大的社群和大家一起成长，真正能够帮助到一些有需要的小伙伴，成为 一个汇集全行业人才的综合型平台。所以我们也开始正式运营求职与行业相关的社群。社群内部主要讨论相关 产业、公司、产品研发、求职与跳槽相关内容。如果您想结交更多同行业的朋友，第一时间了解产业。欢迎加 入我们！ 微信扫码添加小助理邀请进群，备注自驾+昵称+求职； ...

"企业合作一直是中德经贸关系的基石。"中国驻慕尼黑总领事邱学军近日在出席中国汽车品牌蔚来在德 国慕尼黑举办的在欧洲运营10周年庆典时表示，目前德资企业在华设立企业已超过1.2万家，同时，德 国正成为中国企业投资欧洲的热门地区，约2000家中资企业在德投资，其中500家落户巴伐利亚州。 （文章来源：经济日报） 邱学军指出，越来越多像蔚来这样的中国高科技企业将巴伐利亚州视为创新发展的重要桥头堡，与当地 工商界在科技、研发和创新等领域的合作日益紧密，形成了"你中有我、我中有你"的良好局面，充分表 明只有坚持开放合作，才能实现互利共赢。 艾旺格进一步表示，在全球经济不确定性上升的背景下，德国需坚持开放合作，应与包括中国在内的伙 伴维持稳定关系。他指出，中国在原材料、零部件、能源价格和生产效率等方面具备优势，中国的数字 化和创新能力与德国的工业技术形成互补，双方可通过合作，开发更具全球竞争力的产品。他还提到， 中德消费者对汽车需求不同，中国车企在巴伐利亚设立全球设计中心是明智之举，通过中德设计团队对 市场偏好的深入理解，可打造更贴合用户需求的定制化产品。 在当天举行的庆典上，蔚来汽车欧洲副总裁张晖表示："欧洲为蔚来提供 ...

New Car Launches - Geely Galaxy A7 launched with a limited price range of 81,800 to 117,800 yuan, featuring seven models and a focus on energy efficiency with a fuel consumption of 2.67L per 100km and a comprehensive range of over 2100km [3][20][26] - Audi Q6 e-tron family launched with prices ranging from 348,800 to 398,800 yuan, featuring a new design language and advanced technology including a 107kWh battery and a range of 709-765km [27][30][36][37] - Li Auto i8 adjusted its price to 339,800 yuan, offering a unified configuration with additional features such as a platinum sound system and rear entertainment screen [37][49] - New models from Zhijie R7 and S7 opened for pre-sale with starting prices of 268,000 and 258,000 yuan respectively, achieving over 20,000 reservations in just 8 hours [50][58] - The all-new XPeng P7 opened for pre-sale, achieving over 10,000 reservations in just 6 minutes and 37 seconds, showcasing strong market interest [59][74] - New MG4 launched with a pre-sale price range of 73,800 to 105,800 yuan, featuring advanced safety configurations and a unique battery integration technology [75][84] Company Developments - Buick's new Zhi Jing L7 features the True Dragon range extender system, offering a pure electric range of 302km and a comprehensive range exceeding 1400km [146][150] - Shenzhou Car Rental plans to introduce 100,000 units of CATL's chocolate battery swap models, aiming to enhance the rental experience with efficient battery management [155][160] - Xiaoma Zhixing launched its fully autonomous Robotaxi service in Shanghai, marking a significant step in commercializing autonomous driving technology [161] - Luobo Kuai Pao announced a strategic partnership with Lyft to provide autonomous driving services in Europe by 2026, indicating a push into international markets [164][165] - Tesla's board approved a stock grant worth approximately $29 billion to CEO Elon Musk, reflecting the company's reliance on his leadership amid competitive pressures [167][169]

Group 1 - Tesla has officially obtained a ride-hailing permit for its Robotaxi service in Texas, allowing it to operate without a safety driver in the vehicle, although a staff member will monitor remotely [2] - The fare for Tesla's autonomous taxi service is set at approximately $4.2 per trip, equivalent to about 30.2 RMB [2] - Elon Musk believes that by the end of 2025, Tesla could provide autonomous ride-hailing services to half of the U.S. population [2] Group 2 - Following the passage of the "Big and Beautiful" Act, which ends electric vehicle tax credits on September 30, American consumers are rushing to purchase electric vehicles to benefit from a $7,500 tax credit [2] - In July, electric vehicle sales accounted for 9.1% of total passenger car sales, marking a historical high, with nearly 36,700 used electric vehicles sold in the same month [2] - It is anticipated that the third quarter may set a record for electric vehicle sales, while a significant drop in sales is expected in the fourth quarter [2] Group 3 - Ledo Automotive announced that its L90 model will not have a buyout option for a 60 kWh battery, but there is a demand for leasing such a battery, which the company is considering [2] - The L90 model is standardly equipped with an 85 kWh battery, aligning with the company's philosophy of "chargeable, swappable, and upgradeable" energy solutions [2] Group 4 - On August 9, the remote ethanol-hydrogen electric tractor and intercity passenger bus were officially launched in Hohhot, with 270 orders signed on the same day [2] - This event marks a significant step in promoting the ethanol-hydrogen ecosystem strategy in Hohhot [2]

Core Viewpoint - The furniture and autonomous driving industries are experiencing significant growth in production, financing, and recruitment, with a strong emphasis on practical technology and skilled talent acquisition [1][2]. Group 1: Industry Trends - The autonomous driving sector is seeing a surge in companies scaling up production and hiring, indicating a competitive job market where securing positions is challenging due to high skill requirements [1]. - The emergence of high-level autonomous driving demonstration zones, such as in Beijing, is fostering innovation in policy, technology, and commercialization [1]. Group 2: Learning and Community Resources - Several influential communities focused on embodied intelligence, autonomous driving, computer vision, and AI are recommended for systematic learning and skill enhancement [1]. - The "Automatic Driving Heart" community is the largest developer community in China, focusing on various technical aspects of autonomous driving, attracting significant attention from industry professionals [2]. - The "Computer Vision Research Institute" shares the latest research and practical applications in AI, emphasizing technology research and implementation [5]. - The "Embodied Intelligence Heart" community is the first full-stack technical exchange platform in China, covering a wide range of topics related to embodied intelligence [8].

目前自动驾驶和具身智能两个方向我们已经和业内主流的公司及相关高校建立起深度的合作，大模型方向 也正在快速搭建。我们不止聚焦在技术本身，更愿意和大家一起共创整个AI领域，分享认知成长的喜悦。 对于热门事件，我同样希望我们提供全网独一份的内容价值。 点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近30个 方向 学习 路线 大家好，我们是自动驾驶之心/具身智能/大模型之心Tech团队。非常高兴在这里和你相遇，如果你也认同技 术内容可以改变世界，那你可能就是我们在找的人！ 我们在做什么？ 我们希望通过技术内容连接学术界和工业界，成为企业和学校沟通的桥梁，更乃至数十万的AI开发者和创 业者。我们致力于为大家带来全网最新最权威的技术信息，团队聚焦在自动驾驶、具身智能、大模型等AI 最前沿的技术领域，涵盖学术论文解读、业内量产方案分析、大模型评测、商业动态、行业招聘、开源项 目等，并通过公众号、社群、视频号、知乎、小红书、B站等平台进行内容分享、粉丝交流及企业联系。 工作时间： 不积跬步无以至千里，我们深知一个人的力量是有限的，所以我们期待更多优秀的小伙伴与我们一起同行~ 内容运营 - 实习生 ...