12月18日，在2025光合组织人工智能创新大会（HAIC 2025）现场，广州国家实验室、天津大学、湖南 应用数学中心、中科院高能所、国家天文台、中科院大气所、中石油东方物探、中科曙光、合肥大数据 公司等22家高校、科研机构及企业共同发起 "科学智能联合攻关行动" 。该行动将重点围绕科学大模型 开发、超智融合算力平台建设、模型训练推理优化、科学数据开放共享等方面开展协同工作。光合组织 透露，在未来三年内预计将推动超过20个行业级AI4S软硬协同解决方案落地。 ...

Core Insights - The report outlines the top ten core trends in the AI field for 2025, emphasizing the transformation from computational infrastructure to industrial application, highlighting China's rise in open-source ecology and self-controlled routes [1][3]. Group 1: Infrastructure - The core pillars of AI infrastructure are the establishment of computational power and the AI-native architecture of chips. Major global tech companies are investing heavily in large-scale data center construction, with projects like Google's "Stargate" and Microsoft's AI super park exceeding $10 billion [1][3]. - The shift in the chip sector is moving from general computing to AI-native architectures, with GPUs remaining central to training while NPUs become standard for edge devices. Domestic chips have achieved self-sufficiency in training models with hundreds of billions of parameters, breaking foreign technology monopolies [1][3]. Group 2: Model Evolution - The evolution of models focuses on breakthroughs in efficiency and capability. Innovations in pre-training architectures, such as the MoE (Mixture of Experts) model, balance performance and cost, with domestic models like GLM-4.6 and Qwen3 adopting this architecture [1][3]. - Upgrades in inference capabilities are driving the development of adaptive inference and heterogeneous computing technologies, with embodied intelligence becoming a popular area, as humanoid robots begin to enter industrial and household scenarios [1][3]. Group 3: Application Landscape - The application landscape shows a characteristic of "full-scene penetration," with the Agentic internet reshaping traffic entry points from "people finding services" to "services finding people." Multi-Agent collaboration frameworks lower development barriers and promote the execution of complex tasks [2][3]. - The rapid proliferation of AI hardware, including AI PCs, smart wearables, and AI toys, is reshaping human-computer interaction methods, with edge AI gaining popularity due to its low latency and high privacy advantages [2][3]. Group 4: China's Route - China's approach highlights a dual drive of open-source ecology and independent innovation. Open-source AI is entering a "China time," with models like DeepSeek and Qwen achieving high download rates in global open-source communities, establishing international influence [2][3]. - The national strategy incorporates AGI into top-level design, with tech giants and startups shifting focus from applications to core technology development, creating a full-stack ecosystem of "domestic chips + self-developed models + independent SDKs" [2][3].

"药物研发的第一性原理应是治病。" 文 ｜ 海若镜 封面来源 ｜ Unsplash 36氪获悉， AI4S（AI For Science）企业 哲源科技近日完成亿元A1轮融资，由国科投资（中国科技产业投资管理有限公司）领投，泽源基金、睿智医药 跟投。 国内AI制药赛道历经起伏，当前正逐步回归理性。伴随着分子虚筛、自由能预测、抗体结构优化等AIDD工具走向成熟，药物分子的设计难度正在下降， 但首尾两端（靶点发现和临床试验）的挑战，尚未因技术发展而产生本质改变。 成熟靶点的分子开发进入红海，新靶点的发现面临枯竭，药企在漫长的临床试验环节需要投入巨额经费，但仍面临"九死一生"的失败风险。因此，借助AI 技术，挖掘新机制、评估新靶点、提高临床试验效率和成功率，成为当下产业界的重要关注点。 在与北京肿瘤医院合作的虚拟临床平行试验项目中，哲源科技利用其计算医学平台，对8位入组患者进行了药物响应的预测，其介绍称"就像是通过AI给患 者'算命'，预测用药结果（如疾病缓解程度等）。揭盲结果显示：AI预测的药物响应结果与真实临床试验结果完全一致"。 面对复杂疾病，利用AI手段预判药物在临床试验中的结果，就意味着，可以在药物正式开 ...

Core Viewpoint - The establishment of the "Fusion and New Energy Joint Laboratory" between Yanchao Energy and Peking University marks the beginning of deep collaborative research in the fields of fusion and new energy [1] Group 1: Collaboration and Research Focus - The unveiling ceremony of the joint laboratory is part of the 24th anniversary celebration of the Peking University Shenzhen Graduate School [1] - The joint laboratory will focus on research related to stellarator fusion device physics and engineering, AI4S, and materials research [1] - The laboratory aims to develop applications of superconducting technology in the energy sector, including wind power, photovoltaics, energy storage, and energy conservation [1] Group 2: Objectives and Impact - The initiative seeks to address fundamental scientific issues and key technological challenges in the fusion and new energy fields [1] - The laboratory aims to become a global leader in fusion and new energy technology innovation, promoting the development of related disciplines and scientific exploration [1]

在积极利用AI提升知识积累效率的同时，我们需审慎思考，如何确保科学成果的质量与深度不被速度 稀释，并将最终的解释权、决策权等"认知主动权"牢牢掌握在人类自己手中 文 |《瞭望》新闻周刊记者 钱沛杉 11月27日晚，DeepSeek发布其最新数学模型DeepSeekMath-V2，以自验证方式突破了目前AI在深度推理 方面的局限，展现了强大的定理推理能力。DeepSeek方面表示，Math-V2让大模型在数学领域不再只 是"做题家"，而有可能依靠自身全面、严谨的数学推理能力对科学研究产生深远影响。 将AI引入科学研究并非全新事物。在国内，2018年，中国科学院院士、北京大学国际机器学习研究中 心主任鄂维南率先引入AI4S（人工智能驱动科学研究）概念，认为传统科研模式效率较低、学科间壁 垒较高，AI可以实现跨学科融合，加速突破。 特别是，当传统物理模型在极端复杂的系统前束手无策时，数据驱动范式可以跳过对微观机制的繁琐推 演，利用算法直接从海量数据中寻找规律、构建"隐性"模型，为应对气候变化等紧迫挑战，提供可行路 径。 这种模式已影响渗透到科学研究的多个环节，在多个领域催生重大突破，实现科研效率的指数级提升。 在生 ...

当胡启朝拿着一个盒子走上台，很少有人意识到，电池的创新范式发生了怎样的巨变。 "我们擅长从第一原理出发做研发，但AI让我们可以从第零原理出发。" 11月，SES AI创始人兼CEO胡启朝博士，在2025高工锂电年会闭幕式上如是说。 而当他带着一个盒子走上台时，很少有人意识到，电池的创新范式发生了怎样的巨变。 01 重新定义创新的起点 传统电池研发的模式已为行业所熟知。实验室、中试线、生产线的紧密配合，依靠科研人员的经验积累和物理化学规律的探索，在无数次试错中寻找 突破。 "我们擅长这个，"胡启朝说，"但代价是巨大的设备、人才和场地投入，以及超过90%的失败率。" 然而，AI带来了一种截然不同的思考方式。 "AI捕捉的不是物理化学规律，而是实验数据背后的数学规律。"胡启朝这样描述其中的差异。传统方法或许能发现十几个物理化学规律，而AI系统 却能揭示上千个数学规律。 "第零原理"的核心在于，它比"第一原理"更为基础。如果"第一原理"是从已知的科学、物理化学规律出 发进行推演 ，那么"第零原理"则是从数据 中直接发现那些尚未被人类归纳的底层模式，这是一种更为本质的创新起点。 "听起来很抽象，但其实他也可以有物理形态 ...

摘要 欲弯道超车 11月24日，美国总统特朗普签署行政令，启动Genesis Mission，由能源部牵头、17家国家实验室参与，建设覆盖AI、量子计算、先进实验装置与 联邦科研数据的"国家级发现平台"，明确将AI4S提升至国家战略层级。 该平台旨在打破数据孤岛，建立一套由"数据、算力、实验"构成的闭环系统： 数据层： 汇聚联邦政府数十年来积累的涉密及专有科研数据，构建高质量科学基座模型，以解决AI缺乏高质量训练数据的痛点。 算力层 ：深度捆绑科技巨头：NVIDIA、AMD、Microsoft、Google、AWS等有望提供GPU、云平台和工程团队。 行政令设定了激进的时间表： 60天内， 能源部须提交不少于20个"国家级挑战"清单，覆盖先进核能、电网现代化、关键材料、半导体、高端制造等； 90天内： 完成全美联邦算力与数据资源的彻底清查。 9个月内 拿出完整实施方案和预算路径，定义平台架构、数据接入规则、对产业和高校的开放方式。 对能源与材料，官方已点出几条主线： 1，利用AI与高性能计算加速聚变与先进核能研究，包括堆型设计、燃料和结构材料开发等； 2，在"电网现代化"框架下，用AI优化电网运行与规划，提升 ...

About Us 我们是一个对 AI 和 foundation model 痴迷的团队。 2022 年秋，我们在硅谷看到了 AI 的火苗，从此只专注研究 AI。 专注研究和投资 AI，让我们取得了还不错的成绩。我们在管 AUM 超过 15 亿美金，有 5 亿美元在投的长 线基金，一二级联动，有足够的子弹抓住 AI 机会。 我们过去投资并见证了 6 家 portfolio 从数十亿，数 百亿美金，成长为千亿美金公司——这也是拾象的寓意，只研究全球最重要的技术变化，投资有大象级 潜力的公司。 我们有数位千亿美金业务的 CEO / leadership 提供洞察，帮 portfolio 做好 AI 和全球化。 我们通过海外独角兽和 AI 讨论社群，持续讨论重要问题，帮助和影响了中美两地的华人创业者，也在 AI 从业者中获取了一些宝贵的信任。 现在，我们希望邀请你加入，一起做 全球 AI 投资，一起捕捉大机会，成为 AI 领域的最佳投手 。我们是 一个年轻（平均年龄不到 30 岁）、扁平、高人才密度的团队，推崇 high-trust，low-ego，团队内信息极 度透明，讨论氛围热烈。 我们尤其喜欢的特质是：对 AI ...

Core Viewpoint - The article discusses how Yibin is positioning itself as a hub for advanced technologies such as all-solid-state batteries, AI for Science (AI4S), and embodied intelligence, aiming to create a sustainable ecosystem for innovation and industrial evolution [7][8][87]. Group 1: Technological Advancements - All-solid-state batteries signify a transition from liquid to solid energy systems, enhancing safety and energy density, and supporting comprehensive electrification [9][10]. - AI4S combines first-principles reasoning with deep learning to accelerate scientific discovery, particularly in complex fields like drug screening and new material generation [11][12][13]. - Embodied intelligence represents a critical transition for AI, moving from theoretical language models to practical applications in the physical world, addressing uncertainties and feedback [14][15][16]. Group 2: Yibin's Industrial Strategy - Yibin has established a complete industrial chain for power batteries, leveraging local resources and green electricity to enhance supply chain efficiency [21][22][29]. - The city is not creating a new industrial zone but is instead integrating existing resources to facilitate the next wave of technological evolution [23][24]. - Yibin's approach involves reusing materials and data from the 1.0 era to feed into the 2.0 technological advancements, creating a feedback loop for continuous improvement [25][28]. Group 3: Collaborative Ecosystem - The collaboration between local companies and research institutions is crucial for developing a self-learning system that integrates virtual and physical experimentation [42][43]. - Yibin's strategy includes a multi-route approach to technology development, allowing for parallel advancements in various materials and methods without betting on a single direction [72][75]. - The city is fostering a culture of innovation by allowing enterprises to define real needs and challenges, thus creating a dynamic and responsive industrial environment [68][70]. Group 4: Future Challenges and Opportunities - Yibin faces the challenge of maintaining a continuous cycle of high-quality innovation amidst rapid technological changes [88][92]. - The city is transitioning from being a mere industrial base to becoming an experimental production city, capable of adapting to new technological demands [108][110]. - The focus is on developing a platform that can iterate and adapt, ensuring that Yibin remains relevant in the fast-evolving technological landscape [94][102].

Core Insights - The "ZhiXing Plan - Overseas Recruitment" is a strategic talent recruitment initiative by the Beijing Academy of Artificial Intelligence aimed at attracting and nurturing high-level AI research talent from around the world [2]. Group 1: Program Overview - The "ZhiXing Plan" targets top graduates from global universities, inviting selected candidates for a research visit of six months or more at the Beijing Academy of Artificial Intelligence [2]. - Successful candidates will collaborate with leading international research teams on cutting-edge topics and gain entry into the ZhiYuan talent ecosystem [2]. Group 2: Recruitment Highlights - Exceptional performers during the visit may receive priority for full-time positions at the research institute, including roles such as Principal Investigator (PI), researcher, or postdoctoral fellow [5]. - The program offers access to advanced research topics and supercomputing resources, allowing collaboration with leading researchers in the field [5]. Group 3: Support and Resources - Participants will benefit from long-term academic development and resource support, including entry into the ZhiYuan academic network for ongoing research collaboration and career opportunities [7]. - Comprehensive support services will be provided, including academic mentorship and administrative assistance for both research and daily life [7][8]. Group 4: Target Audience and Research Areas - The recruitment is aimed at Chinese doctoral candidates, postdoctoral researchers, and formal research staff from top universities and laboratories [9]. - Research directions include brain-like models, intelligent systems, embodied intelligence, multimodal AI, and AI for Science, with candidates expected to have published at least three first-author papers in top conferences [9]. Group 5: Application Process - Applications are accepted year-round with rolling reviews until positions are filled [11]. - The process includes academic evaluation, interviews, and formal invitations for successful candidates [10]. Group 6: Vision and Goals - The "ZhiXing Plan - Overseas Recruitment" is positioned as a starting point for integrating into China's leading AI research platform, aiming to conduct globally impactful research and build long-term collaborative relationships [12].