风险提示：提及个股仅用于行业事件分析，不构成任何个股推荐或投资建议。指数等短期涨跌仅供 参考，不代表其未来表现，亦不构成对基金业绩的承诺或保证。观点可能随市场环境变化而调整，不构 成投资建议或承诺。提及基金风险收益特征各不相同，敬请投资者仔细阅读基金法律文件，充分了解产 品要素、风险等级及收益分配原则，选择与自身风险承受能力匹配的产品，谨慎投资。 每日经济新闻 AI加速材料发现，技术进步催化，2月9日，新材料ETF国泰（159761）收涨超2%。 华泰证券指出，AI for Science（AI4S）的研究范式正通过赋能量子、原子与连续介质系统中的高级 建模、仿真与预测，引领科研革命。AI构建了新材料研发新范式，驱动产业跨越"创新鸿沟"。AI新材料 预计将成为AI4S的重点应用与投资方向，以"AI预测+自动化实验"为核心的研发闭环已进入规模化验证 阶段，能十倍级缩短传统研发周期。未来AI有望从光伏钙钛矿向固态电池电解质、高温超导材料、半 导体光刻胶等更高附加值的领域复制。AI不仅加速材料发现，更通过数字化工艺优化直接推动产业 化，是实现制造产业升级的核心引擎。 新材料ETF国泰（159761）跟踪的是新材料指 ...

AI不只是应用工具，已经开始帮助人类攻克基础科学的"卡脖子"问题。 文|《中国企业家》见习记者 孙欣 记者 王怡洁 见习编辑|李原 编辑|何伊凡 图片来源|受访者 "如果公司是艘在深海中探索的船，我是最不能跳船那个人。"深度原理总部位于杭州，其创始人贾皓钧 将自己的办公室命名为"哥伦布"。在他看来，在AI for Science这个全新的风口创业，无异于"哥伦布探 索新大陆"。 成立一年多时间以来，深度原理正在研发、融资和商业化上一路狂奔。贾皓钧每天早上会保持五到十分 钟的深度思考，盘点目前公司有哪些风险，以及下一个目标在哪里。这个习惯，从2023年贾皓钧创业时 开始。那时他正在麻省理工学院（以下简称MIT）攻读博士学位。 AI for Science（科学智能，行业简称"AI4S"），意指用AI来做新的科学发现。2023年，美国贝克团队与 谷歌DeepMind开发的深度学习模型"RFdiffusion"问世，该模型预测了约2亿个蛋白质结构，并可一键设 计和生成蛋白质。2024年，诺贝尔化学奖被授予了贝克团队和DeepMind团队。 同一年，贾皓钧正式创立深度原理，团队基于生成式AI和第一性原理的融合，将AI ...

在模型层面，上海交通大学人工智能学院助理教授张林峰发布了Innovator基座模型。该科学基座模 型实现了科学多模态感知、科学推理、科学工具调用的三个目标。感知方面，面向化学、材料、物理 等学科多模态科学数据建立理解能力，支持20多种科学模态，且同时具备顶尖的通用视觉理解能 力。科学推理方面，在科学编程任务上超越30倍参数量的模型。 研讨会现场还举行战略合作签约仪式。上海赛兰德智能科技有限公司分别与上海埃迪希科技服务有限 公司、上海库帕思科技有限公司签署战略合作协议，围绕科研算力供给与数据价值挖掘开展合作。 2026.01. 29 本文字数：1021，阅读时长大约2分钟 来源 | 一财区域 当前，面向AI for Science（AI4S）的关键基础设施已逐步成形，规模化、智能体驱动的科学研究从 概念走向现实的时机趋于成熟。 1月29日，由上海交通大学人工智能学院与上海算法创新研究院联合主办的"Agentic Science at Scale ——AI4S科学基座模型和通用科研智能体研讨会"在上海模速空间举行。会议发布科学基座模型 Innovator、科研智能体SciMaster等核心成果，并通过产学研战略签 ...

Group 1 - The key infrastructure for AI for Science (AI4S) is gradually taking shape, marking the maturity of large-scale, agent-driven scientific research [1][3] - The "Agentic Science at Scale" era has officially begun, as stated by the chief advisor of the Shanghai Jiao Tong University AI Institute during the opening report [3] - The conference introduced core achievements such as the Innovator scientific base model and the SciMaster research agent, aimed at bridging the last mile of intelligent and scalable scientific research through industry-academia-research strategic agreements [1][3] Group 2 - The SciMaster research agent is designed to achieve a closed-loop process in scientific research across all disciplines, providing an "autonomous driving" experience that can match the output of a senior theoretical physics PhD in just 6 hours of operation [3][4] - The Innovator base model achieves three goals: multi-modal scientific perception, scientific reasoning, and scientific tool invocation, supporting over 20 scientific modalities and demonstrating superior general visual understanding capabilities [4] - Strategic cooperation agreements were signed between Shanghai Sailande Intelligent Technology Co., Ltd. and other companies to collaborate on research computing power supply and data value mining [4]

Core Viewpoint - The article emphasizes that while AI has made significant advancements, particularly in language models, it still lacks a true understanding of the physical world, which limits its potential applications in scientific fields [1][20]. Group 1: AI's Limitations and Future Directions - Current mainstream AI excels in language and statistical associations but struggles to grasp fundamental concepts like distance, scale, and causality [1]. - The concept of "AI for Science" (AI4S) is introduced as a critical pathway that aims to integrate AI into scientific research, focusing on understanding the physical world governed by chemistry and physics [2][20]. - AI4S is not merely an enhancement of computational power but a targeted approach to solving complex scientific problems [2]. Group 2: Industry Applications and Capital Market Interest - AI4S is transitioning from concept to practical applications, with SES AI's "Molecular Universe" platform demonstrating real economic value through the development of new electrolyte materials [3]. - The capital market is increasingly interested in AI4S, with several companies in this space achieving billion-dollar valuations, indicating a growing recognition of its commercial potential [3][4]. - SES AI has developed six breakthrough electrolyte materials, showcasing the practical applications of AI4S in industries like battery manufacturing [3][7]. Group 3: Case Studies and Success Stories - The success of companies like Jingtai Technology, which became the first "specialized technology stock" in Hong Kong, illustrates the potential of AI4S in the pharmaceutical sector [4]. - The growth of AI4S companies is often rooted in long-term, practical experience in specific scientific fields rather than merely competing in model capabilities [4][6]. Group 4: Technological Innovations and Breakthroughs - SES's MU platform has produced innovative solutions across various applications, including electric vehicles and drones, with significant performance improvements over industry benchmarks [7][8][10]. - The introduction of the "Flavor" system in MU-1.5 allows AI to leverage both known scientific knowledge and hidden data correlations, enhancing its predictive capabilities [14][15]. Group 5: Efficiency and Future Prospects - The MU platform aims to transform research efficiency by integrating a comprehensive workflow that reduces costs and accelerates development cycles [16][17]. - The "MU in a Box" initiative allows for localized deployment of the MU platform, enabling companies to utilize their proprietary data for tailored AI applications [17][18]. - The article concludes that the true value of AI4S lies in its ability to enhance scientific understanding and drive efficient research, positioning it as a critical component of future innovations in battery technology and beyond [20][22].

Core Insights - The article discusses the challenges in deploying scientific software, emphasizing that most tools are published but not executable, which limits reproducibility and integration in scientific research [3][6][11] - The emergence of AI for Science (AI4S) highlights the need for tools that can interact seamlessly with AI systems, making the ability to run these tools a fundamental issue [8][9][10] - Deploy-Master is introduced as a solution to streamline the deployment process, focusing on creating a shared infrastructure that ensures tools are executable [12][35][37] Group 1: Challenges in Scientific Software - Scientific software often requires extensive manual effort to compile and run, leading to inefficiencies and reliance on individual expertise [4][5] - The deployment bottleneck persists despite advancements in containerization and cloud computing, affecting the usability of scientific software [7] - The lack of a systematic approach to convert tools into executable formats is identified as a structural barrier to the scalability of AI4S and Agentic Science [11][35] Group 2: Deploy-Master Overview - Deploy-Master is designed as a one-stop automated workflow centered on execution, addressing the entire deployment chain from discovery to execution [12] - The tool employs a multi-stage funnel process to filter and validate scientific tools, reducing an initial pool of 500,000 repositories to 52,550 candidates for automated deployment [15] - A dual model review mechanism is implemented to enhance the success rate of building specifications, achieving over 95% success in generating executable tools [22] Group 3: Deployment Insights - The deployment process is characterized by a long-tail distribution of build times, with most tools completing in around 7 minutes, but some requiring significantly longer due to complexity [25][26] - A diverse language distribution is observed among successfully deployed tools, with Python being the most prevalent, followed by C/C++, R, and Java [27][28] - Failure rates in builds are concentrated in specific areas, primarily due to inconsistencies in build processes and missing dependencies [31][32] Group 4: Future Implications - Deploy-Master's success in creating a large collection of executable tools provides a foundation for community agents and various master agents to operate effectively [35][36] - The methodology established by Deploy-Master can be applied beyond scientific computing to other software ecosystems, emphasizing the importance of a robust execution infrastructure [37]

Group 1 - The core viewpoint of the article highlights that 2023 may be a breakout year for AI for Science (AI4S), driven by significant advancements in capabilities that allow for more extensive scientific research [1] - AI4S is now recognized as a strategic battleground among global tech giants, moving beyond the "proof of concept" stage in laboratories [1] - AI4S is aiding scientists in efficiently identifying new research opportunities, such as predicting protein functions, designing new materials, and discovering new targets [1] Group 2 - Related A-share concept stocks mentioned include Health元 and Northeast Pharmaceutical [2]

Core Insights - The article discusses the challenges in deploying scientific software tools, emphasizing that many open-source tools are not readily executable, leading to inefficiencies in research practices [2][6][24] - Deploy-Master is introduced as a solution to automate the deployment process, enabling the execution of over 50,000 scientific tools and addressing the deployment bottleneck in AI for Science (AI4S) and Agentic Science [1][8][21] Group 1: Challenges in Scientific Software Deployment - The scientific computing field has seen an unprecedented accumulation of open-source software tools, yet most remain in a state of "published" rather than "executable" [2] - Researchers often spend days or weeks resolving issues related to compilation failures, dependency conflicts, and system incompatibilities, which hampers reproducibility and large-scale evaluation [2][6] - The deployment bottleneck persists despite advancements in containerization, cloud computing, and high-performance computing (HPC) platforms, limiting the usability of scientific software [2][7] Group 2: The Role of AI in Scientific Tools - The emergence of AI for Science (AI4S) has intensified the need for scientific tools to interact closely with AI systems, requiring them to execute simulations, run analysis pipelines, and process real data [3][4] - The ability of tools to be executed has become a fundamental issue, impacting the planning and execution capabilities of intelligent agents in scientific research [5][6] Group 3: Deploy-Master Overview - Deploy-Master is designed as a one-stop automated workflow centered on execution, addressing the entire deployment chain from tool discovery to execution [10][21] - The tool employs a multi-stage funnel process to filter and validate scientific tools, ultimately narrowing down from approximately 500,000 repositories to 52,550 candidates for automated deployment [12] Group 4: Building and Validating Tools - The Build Agent within Deploy-Master utilizes a dual-model debate mechanism to generate and validate build specifications, significantly increasing the success rate of deployments to over 95% [15] - The deployment process reveals a long-tail distribution in build times, with most tools completing in around 7 minutes, while some require significantly longer due to complex dependencies [17] Group 5: Observability and Continuous Improvement - Deploy-Master aims to transform the deployment of scientific software from an experiential judgment into a quantifiable and analyzable engineering object [20] - The system allows for systematic observation of deployment behaviors, identifying failure points and underlying assumptions that can be improved over time [18][19] Group 6: Implications for the Broader Software Ecosystem - The methodology developed through Deploy-Master is not limited to scientific computing but can be applied to various software tool ecosystems, including engineering tools and data processing systems [23][25] - The article concludes that establishing an execution-centered infrastructure is crucial for overcoming deployment challenges across different software domains [24][26]

Core Insights - The article discusses the challenges in deploying scientific software, emphasizing that most tools are published but not executable, leading to inefficiencies in research practices [3][5][21] - It introduces Deploy-Master as a solution to create a shared infrastructure that transforms scientific tools into executable entities, addressing the deployment bottleneck in AI for Science (AI4S) and Agentic Science [5][19][20] Group 1: Challenges in Scientific Software Deployment - A significant issue is that scientific software often requires extensive time to resolve compilation failures and dependency conflicts, resulting in a lack of reproducibility and integration [3][4] - The emergence of AI4S has intensified the need for tools that can interact seamlessly with scientific processes, making the ability to execute tools a fundamental concern [3][5] - The deployment process is not isolated but part of a continuous chain that includes discovery, understanding, environment construction, and execution [5][19] Group 2: Deploy-Master Overview - Deploy-Master is designed to automate the deployment workflow, focusing on execution readiness and addressing the challenges of discovering and verifying scientific tools [5][19] - The initial phase involved searching through 91 scientific and engineering domains, resulting in a refined list of 52,550 candidates for automated deployment from an initial pool of 500,000 repositories [8][9] - A dual-model debate mechanism was implemented to enhance the success rate of building specifications, increasing it to over 95% by iteratively refining the proposed build plans [12][13] Group 3: Deployment Insights and Observations - The deployment process exhibits a long-tail distribution in build times, with most tools completing in around 7 minutes, while some require significantly longer due to complex dependencies [15] - A diverse language distribution was observed among the successfully deployed tools, with Python being the most prevalent, followed by C/C++, R, and Java [16] - The primary reasons for build failures were identified as inconsistencies in the build process, missing dependencies, and mismatched compilers or system libraries, highlighting the need for improved deployment strategies [16][17] Group 4: Implications for the Future - Deploy-Master provides a foundational infrastructure for community agents, enabling them to share verified tools and ensuring a stable action space for planning and execution [19][20] - The methodology established through Deploy-Master can be applied to broader software ecosystems, indicating that deployment challenges are not limited to scientific tools but are prevalent across various software types [20] - The article concludes that in the era of Agentic Science, execution is a prerequisite for all capabilities, and establishing a robust execution infrastructure is essential for future advancements [20][21]

Core Insights - Jinqiu Fund participated in a strategic financing round for the AI for Science startup "Deep Principle," focusing on breakthrough technologies and innovative business models in the AI sector [3] - "Deep Principle" emerged as one of the top three winners in the "Foreseeing New Product Research" track at the 2025 BIG BANG Beauty Tech Co-Creation Program, which had over 700 participating startups [3] - The founder and CEO of "Deep Principle," Dr. Jia Haojun, highlighted the transformative impact of AI on material research during a roundtable forum, emphasizing the balance of depth, speed, and budget in beauty formulation development [6][8] Investment Highlights - Jinqiu Fund's long-term investment philosophy is centered on identifying promising AI startups with innovative approaches [3] - The AI-driven platform ReactiveAI developed by "Deep Principle" enhances research efficiency by predicting and explaining the effects of ingredients on formulation performance, leading to quantifiable benefits such as reduced R&D cycles and lower costs [6] Industry Trends - The 2025 BIG BANG event showcased the importance of open innovation in the beauty industry, as emphasized by L'Oréal's global R&D executive, who noted that collaboration is key to driving disruptive solutions [8][9] - The integration of AI in material research is seen as a significant shift from discovery to rational design, indicating a future where AI continuously improves through experimentation [6]