作者 | 谭保罗 编辑 | 宝珠 视觉 | 诺言 理解机器人产业，有两条脉络。 第一条是感官的，故事发生在文艺作品之中。从上世纪20年代开始，机器人就开始出现在影视作品之 中。在100年时间内，电影里活蹦乱跳的机器人基本都具有人形特征和超级智商。 它们代表了人类最朴 素的终极愿望——让原本脆弱的肉身如同钢铁般强大，让人类大脑不断突破生理算力的极限。 2025年，穿着大花褂子的宇树机器人在春晚亮相/图源：宇树官网 2020-2024年深圳市机器人产业总产值与增速/图源：深圳机器人协会 面对全国乃至全球范围内的机器人浪潮，作为产业科创前沿的深圳，来自这里的观点必然值得一听。 为此，盐财经日前采访了深圳市人工智能与机器人研究院常务副院长丁宁。该研究院（简称AIRS）成立 于2019年，是深圳市政府依托香港中文大学（深圳）建立的市级十大基础研究机构之一，重点开展人工 智能与机器人领域的基础理论与原创技术研究，推动智能技术在医疗康复、智能制造、城市服务和可持 续发展等方面的应用转化。 建立常识，先懂得"机器人"分类 另一条脉络是经济性的，是真正发生在现实世界的故事。机器人在现实中的大规模出现，主要是工业机 器人，在相当 ...

Core Insights - OceanBase has launched PowerRAG, an AI-focused application product that enables ready-to-use RAG application development, marking its commitment to the AI era [1] - The company aims to evolve from an integrated database to an integrated data foundation, focusing on a comprehensive layout across computing power, infrastructure, platform, application, and delivery forms [1] - Ant Group's CTO emphasized the importance of data in the development of AI and large models, highlighting four major challenges: increased data acquisition costs, scarcity of rigorous industry data, the need for enhanced multi-modal data processing capabilities, and difficulties in data quality assessment [1][7] Company Strategy - Ant Group will support OceanBase in achieving breakthroughs in key AI scenarios across finance, healthcare, and daily life, while promoting the Data×AI concept and architectural innovation [2][10] - OceanBase is positioned as a representative of Ant Group's continuous innovation and technical breakthroughs, particularly in handling massive transaction data [9] Industry Challenges - The cost of data acquisition has significantly increased, with readily available and inexpensive data resources nearing exhaustion, leading to a focus on generating high-quality data as a key success factor for digital enterprises [7] - High rigor industries, such as legal and healthcare, face challenges in data circulation due to stringent data quality requirements and a lack of digital knowledge, which hampers the effective application of generative AI [8] - The processing of multi-modal data remains a significant challenge, as future data will encompass not only text but also visual and tactile information, necessitating advanced handling capabilities [8] - Quality assessment of data is crucial, as it directly impacts the performance of large models, with the need for extensive evaluation data posing a significant challenge [9]

Group 1 - The core viewpoint is that in the next two to three decades, it will not be AI replacing humans, but rather those who use AI replacing those who do not [1] - According to McKinsey Global Institute, within the next five years, 20% to 30% of jobs will be replaced by AI, and by 2030 to 2060, 50% of existing jobs may be affected, with a midpoint around 2045 [3] - The International Monetary Fund (IMF) estimates that by 2050, 60% of jobs in developed economies could be impacted by AI [3] Group 2 - The emergence of general artificial intelligence (AGI) could lead to the restructuring of over 90% of jobs by 2050, although the exact timeline remains debated [3] - There is a need to consider changes in educational content and models, with AI being integrated as a fundamental subject alongside traditional subjects like language and mathematics [3] - The goal of science education and popular science in the AI era is to cultivate future scientists and scientifically literate citizens who can engage with AI and contribute to its governance [4]

Core Viewpoint - Alibaba's Q4 FY2025 financial results show a revenue increase of 7% year-on-year, driven by a user-first and AI-driven strategy, despite a stock price drop due to lower-than-expected revenue [2][3][4]. Financial Performance - Alibaba reported Q4 revenue of 2364.54 billion yuan, up from 2218.74 billion yuan in the same quarter last year, marking a 7% increase [3]. - Non-GAAP net profit for Q4 was 298.47 billion yuan, a 22% increase from 244.18 billion yuan year-on-year [2]. - Taotian Group's customer management revenue grew by 12% to 710.77 billion yuan in Q4 [6]. Business Segment Performance - Taotian Group revenue reached 1013.69 billion yuan, a 9% increase from 932.16 billion yuan year-on-year [3]. - Alibaba International Digital Commerce Group revenue increased by 22% to 335.79 billion yuan [3]. - Alibaba Cloud revenue grew by 18% to 301.27 billion yuan, driven by faster public cloud business growth and increased adoption of AI-related products [3][4]. AI and Cloud Strategy - Alibaba plans to invest over 380 billion yuan in cloud and AI hardware infrastructure over the next three years to meet growing AI demand [5]. - AI-related product revenue has seen triple-digit year-on-year growth for seven consecutive quarters, contributing to Alibaba Cloud's double-digit annual growth [4][5]. Monetization and User Engagement - The number of Taotian 88VIP members exceeded 50 million, maintaining a double-digit year-on-year growth [6]. - The implementation of the "All-Station Promotion" tool has improved monetization efficiency, allowing merchants to achieve a more predictable ROI [6][7]. - Alibaba's e-commerce strategy includes a focus on "instant retail," aiming to convert more users into instant retail customers [8].

Core Viewpoint - Alibaba Group's financial results for the fourth quarter and fiscal year 2025 demonstrate strong growth driven by AI initiatives, with significant increases in revenue and shareholder returns [1][21]. Financial Performance - For the fourth quarter ending March 31, 2025, Alibaba reported revenue of 236.45 billion RMB, a year-on-year increase of 7%, and a non-GAAP net profit of 29.85 billion RMB, up 22% from the previous year [4][20]. - In fiscal year 2025, Alibaba achieved total revenue of 996.35 billion RMB, a 6% increase year-on-year, with adjusted EBITA growing 5% to 173.07 billion RMB [6][20]. - Alibaba Cloud's revenue grew 18% year-on-year to 30.13 billion RMB, marking the fastest growth in three years, driven by strong AI demand [10][12]. Shareholder Returns - Alibaba repurchased 1.197 billion shares for a total of 11.9 billion USD in fiscal year 2025, becoming one of the largest repurchasers among Chinese concept stocks [1][21]. - The company announced a total dividend of 4.6 billion USD for fiscal year 2025, reinforcing its commitment to enhancing shareholder returns [1][21]. AI and Cloud Business Growth - AI-related product revenue has seen triple-digit growth for seven consecutive quarters, indicating robust demand across various industries [1][12]. - Alibaba Cloud's market share has been recovering for three consecutive quarters, reflecting the increasing adoption of AI solutions [10][12]. Strategic Focus and Trends - CEO Wu Yongming highlighted two key trends in AI: the shift of AI applications from internal systems to user-facing scenarios in medium and large enterprises, and the expansion of AI usage among small and medium-sized enterprises [2][13]. - The company plans to invest over 380 billion RMB in cloud and AI hardware infrastructure over the next three years to meet growing AI demand [21]. Business Segment Performance - The International Digital Commerce Group's revenue grew 22% year-on-year to 33.58 billion RMB, driven by strong performance in cross-border business [19][20]. - The Local Services Group's revenue increased by 10% to 16.13 billion RMB, supported by order growth from Gaode and Ele.me [19][20]. - The Digital Entertainment Group's revenue rose 12% to 5.55 billion RMB, primarily due to strong performance in film and entertainment [19][20].

General-Level团队 投稿 量子位 | 公众号 QbitAI 正因如此，在AI竞赛进入"下半场"之际（由最近的OpenAI研究员姚顺雨所引发的共识观点）， 设计科学的评估机制俨然成为决定胜负的核心 关键 。 近期录用于ICML'25 (Spotlight)的论文《On Path to Multimodal Generalist: General-Level and General-Bench》提出了 一套全新的评 测框架General-Level 和 配套的数据集General-Bench ，为这一议题带来了奠基性的解答和突破。 该评测框架已落地于社区：上述论文的项目团队构建了涵盖700多个任务、覆盖5大常见模态、29个领域、多达32万+测试数据的超大规模评 测基准和业界最完善的多模态通才模型排行榜 Leaderboard ，为公平、公正、全面地比较不同多模态通才大模型提供了基础设施。 General-Level评估算法：五级段位体系与协同效应 General-Level评测框架引入了一个 五级段位体系 ，类似"段位晋级"的方式来衡量多模态模型的通才能力。 General-Level评估的核心在于 ...

Core Insights - The core viewpoint of the articles highlights the significant role of AI in driving growth for major tech companies like Alibaba and Tencent, with both companies intensifying their investments in AI to capture future growth opportunities [1][5][8]. Group 1: Alibaba's Performance and Strategy - Alibaba's revenue for Q4 of FY2025 reached 2364.54 billion yuan, marking a 7% year-on-year increase, with AI-driven initiatives contributing to this growth [2]. - Alibaba Cloud's revenue for Q4 grew from 3% to 18% year-on-year, with AI-related product revenue achieving triple-digit growth for seven consecutive quarters [2]. - For FY2025, Alibaba Cloud's revenue reached 1180 billion yuan, reflecting an 11% year-on-year increase, and its market share has been recovering for three consecutive quarters [2][7]. Group 2: Tencent's Performance and Strategy - Tencent's Q1 FY2025 revenue was 1800.2 billion yuan, showing a 13% year-on-year growth, with AI strategies revitalizing its marketing services, value-added services, and fintech sectors [2][3]. - Tencent has increased its R&D investment by 21% year-on-year in Q1 FY2025, focusing on enhancing AI capabilities across its product lines [6]. - The company has adopted an "AI in All" strategy, accelerating AI integration across all business lines to create new synergies and value [6]. Group 3: Industry Trends and Future Outlook - The articles indicate two major trends in AI application: the shift from internal systems to user-side scenarios in medium and large enterprises, and the expansion of AI product usage from large enterprises to a significant number of small and medium enterprises [1][6]. - Both Alibaba and Tencent are committed to investing heavily in AI infrastructure, with Alibaba planning to invest over 380 billion yuan in cloud and AI hardware over the next three years [5]. - The demand for AI is rapidly growing across various industries, including traditional sectors like agriculture and manufacturing, indicating a broadening application of AI technologies [7][8].

Core Viewpoint - The prediction of the "iPhone moment" for general-purpose robots will arrive in 5 to 7 years, driven by technological advancements, cost reductions, and evolving market demands [31][34]. Company Overview - Zhihui Square, founded by Guo Yandong, aims to achieve a production capacity of one million units by 2033, covering diverse scenarios such as industrial, logistics, and home services [2][34]. - Guo Yandong has a strong background in AI, having worked with Microsoft, Xiaopeng Motors, and OPPO, and is focused on expanding AGI from the digital to the physical world [2][3]. Technological Challenges - The company identifies three major challenges: ensuring the generality and robustness of technology, balancing cost and value, and deepening the exploration and standardization of application scenarios [7][8]. - The goal is to develop robots that can adapt across various industries and tasks without extensive reprogramming, which requires breakthroughs in multi-modal perception and autonomous decision-making [7][31]. Business Model - Zhihui Square adopts a "soft and hard integration" model, emphasizing the delivery of complete solutions rather than standalone technology components [10][12]. - The company plans to self-build production lines to maintain control over product quality, iteration speed, and long-term cost management [10][34]. Product Development - The GOVLA model is claimed to be the world's first all-domain, all-body VLA model, enabling robots to understand and adapt to dynamic environments [13][14]. - The model's capabilities allow robots to perform complex tasks autonomously, such as preparing breakfast by navigating to different locations [14][15]. Market Strategy - The initial focus on high-end industrial scenarios like automotive, semiconductor, and biotechnology is due to their urgent need for automation and the high willingness to pay from clients [25][26]. - The company aims to gradually expand into public services and home services, creating a closed loop of technology, scenarios, and data [28]. Future Outlook - The expectation is that the hardware costs for high-performance robots will significantly decrease within the next 5 to 7 years, making them accessible to ordinary consumers [32][34]. - The emergence of "killer applications" in key scenarios will be crucial for stimulating large-scale demand for general-purpose robots [32][31]. Competitive Advantage - China is seen as having a unique advantage in developing the embodied intelligence industry due to its robust hardware supply chain, diverse application scenarios, and strong support from investors and government [36][37]. - The company emphasizes the importance of core technology self-research and innovation to compete globally, particularly in the development of intelligent models and algorithms [37].

Core Insights - The company aims to expand its production capacity to one million units by 2033, targeting diverse applications in industrial, logistics, and home services [2][31] - The CEO, Guo Yandong, emphasizes the integration of software and hardware, with a focus on self-developed AI models to drive the capabilities of their robots [3][9] - The company is positioned in the emerging field of embodied intelligence, aiming to transition general artificial intelligence (AGI) from the digital realm to physical robots [4][5] Technology and Innovation - The GOVLA model is claimed to be the world's first all-encompassing VLA (Vision-Language-Action) model, enabling robots to understand and interact with dynamic environments [13][14] - The company focuses on creating robots that can autonomously navigate and perform tasks in real-world settings, moving beyond traditional automation tools [14][15] - The technology aims to achieve robustness and generalization across various industries, addressing the challenges of adapting to different environments without extensive reprogramming [7][8] Business Model and Strategy - The company adopts a "robot as a service" model, providing comprehensive solutions rather than just selling hardware [27][11] - It plans to validate its business model through early revenue generation in high-value industrial sectors such as semiconductors and automotive manufacturing [24][11] - The strategy includes self-building production lines to maintain control over product quality and cost, avoiding reliance on external suppliers [31][32] Market Position and Future Outlook - The CEO predicts that the "iPhone moment" for general-purpose robots will occur in 5 to 7 years, driven by technological advancements and decreasing costs [28][29] - The company aims to leverage China's robust supply chain and diverse application scenarios to establish a competitive edge in the global market [34][35] - The focus on core technology development and understanding market needs is seen as crucial for the company's success in the competitive landscape of embodied intelligence [36][35]

Group 1 - Zhongke Zhidong Taichu (Beijing) Technology Co., Ltd. has completed its first round of financing amounting to several hundred million yuan, led by Zhongke Chuangxing, with participation from multiple investment institutions [1] - The financing will be used for research and application of multimodal artificial intelligence technology, aiming to enhance the development of general artificial intelligence (AGI) and deepen the industrial layout in the "AI+" field [1][7] - The company is recognized as a key player in the national strategy for multimodal large models, with its self-controlled AGI model being pivotal in global technological competition [3] Group 2 - The multimodal large model developed by Zhongke Zhidong Taichu is listed as a key project in the 14th Five-Year Plan by the Chinese Academy of Sciences and has received the highest award at the 2022 World Artificial Intelligence Conference [5] - The model has evolved from version 1.0, the world's first trillion-parameter multimodal model, to version 3.0, which fully benchmarks against GPT-4, setting a new standard for multimodal AI development [5] - The chairman of the company emphasized the importance of AI as a new focus in international strategic competition and its potential to reshape various industries, committing to technological innovation and industrial empowerment [7] Group 3 - By May 2025, Zhongke Zhidong Taichu is expected to be listed among the top 50 AI technology companies in China by Forbes, with its chairman recognized as an influential figure in AI [9] - The completion of this financing is seen as a dual recognition of the company's technological accumulation and industrial practice, providing strong momentum for future development [9] - The company plans to leverage this opportunity to increase investment in core technology research and contribute to the high-quality development of China's AI industry [9]