大消费活跃 一是当地时间周一，美股因马丁·路德·金纪念日休市，周二恢复正常交易。美股股指期货今天大跌，市场对周二美股表现产生担忧情绪，今天上午亚太股 市多数表现不佳。 消息面上，新华社报道，美国总统特朗普17日在社交媒体上发文，宣布美国将从2月1日起对丹麦、挪威、瑞典、法国、德国、英国、荷兰和芬兰的输美商 品加征10%关税。 二是市场资金面的变化。首先是从1月15日以来，大盘成交额快速萎缩。其次是融资余额结束10连增，数据显示，截至1月19日，A股市场融资余额报 27059亿元，较前一日减少85亿元。2026年以来，截至1月16日，A股市场融资余额已连续10个交易日增长。此外，近期多只宽基ETF持续放量，包括沪深 300ETF华泰柏瑞（510300）、上证50ETF（510050）、科创50ETF易方达（588080）、沪深300ETF易方达（510310）等。 三是商业航天概念降温。 截至上午收盘，上证指数下跌0.3%，深证成指下跌1.22%，创业板指下跌1.83%。 今天上午，指数明显波动。上证指数盘中一度跌破4100点，最大跌超0.8%，之后迎来震荡修复，跌幅收窄。盘面上，科技股回调，算力产业链大跌。大 ...

复旦大学张江校区计算机楼内，隐藏着一间实验室。多屏电脑上跳跃的代码，只有键盘敲击的轻响…… 这里的核心资产是人与思想。 这间特殊的"实验室"，就在楼内的办公室和研究生工位里。它的官方名称是"复旦大学人工智能创新与 产业研究院"，聚焦"AI for Science"，师生们更喜欢它的"昵称"——AI3院。三个"I"，源自其英文 名"Artificial Intelligence Innovation and Incubation"，寓意"创新"与"孵化"。更深层的含义，是"连点成 线、连线成面、连面成体"的体系化AI愿景，也是"AI三生万物"的无限可能。 AI3院与多家企业共建校 企联合实验室，推动前沿技术协同攻关。其中，与企业无限光年共同成立"人工智能大模型校企联合研 究中心"，在人工智能与金融、科学智能等前沿交叉领域开展探索。复旦大学人工智能创新与产业研究 院副院长韩丽妹解释道："传统校企合作，企业提需求，学校团队攻关交付。但在AI前沿，许多时候 连'问题是什么'都无人知晓。"这正是AI3与无限光年公司共建的校企联合实验室尝试突破的范式——它 不是成果交易市场，而是一个共同探索未知的"创新容器"。 1200 ...

SES AI (NYSE:SES) FY Conference Summary Company Overview - **Company**: SES AI - **Industry**: Battery Technology - **Focus**: Accelerating energy transition through AI, specializing in lithium metal battery chemistry and energy storage solutions Key Points Industry and Market Insights - The energy storage system (ESS) market is projected to be over **10 times** the size of the electric vehicle (EV) market [2][3] - The battery markets for ESS and drones are expected to experience significant growth [3] - The company is uniquely positioned due to over a decade of experience in battery technology and AI integration [2] Financial Performance - In the first three quarters of 2025, SES AI reported a revenue of **$16.4 million** and provided guidance of **$22.5 million** for the full year [3] - The company anticipates doubling its revenue in 2026 compared to 2025 [34] Product Development and Innovations - SES AI has developed a platform called **Molecular Universe**, which focuses on material discovery and battery health management [3][6] - The company has made **six breakthroughs** in battery materials, enhancing performance for various applications including EVs, drones, and consumer electronics [8][9] - The company has established three business units focusing on ESS, drones, and materials to drive revenue generation [6][12] Strategic Partnerships and Acquisitions - SES AI acquired **UZ**, a company with strong hardware capabilities, to enhance its ESS offerings [7][18] - The partnership with **HySun** aims to commercialize materials discovered through Molecular Universe [9][27] Market Opportunities - The company sees military drones and data centers as the most exciting market opportunities [13][14] - SES AI is focusing on **NDAA-compliant** cells, which are priced **2 to 3 times** higher than non-compliant cells [15][16] - The company is targeting about **100 key accounts** for NDAA-compliant cells, focusing on larger clients rather than numerous small accounts [16] Technology and Competitive Advantage - SES AI's **Molecular Universe** is the first AI for science platform that has made actual breakthroughs in material discovery, unlike other platforms that are still in development [9][11] - The company emphasizes the importance of having a combination of model, data, and domain expertise to succeed in the AI for science space [11] Future Outlook - SES AI aims to create a closed-loop system integrating hardware and software solutions for data centers and commercial applications [17][18] - The company plans to leverage its software for recurring revenue through updates and maintenance subscriptions [23] - SES AI is confident in its ability to achieve break-even cash flow within the next **one to two years** [34] Conclusion - The battery industry is undergoing a significant transition, with SES AI focusing on ESS, drones, and materials as key growth areas [34] - The integration of AI with battery technology presents a unique opportunity for SES AI to capitalize on emerging market trends and drive innovation in energy storage solutions [34]

Group 1 - The stock of Zhite New Materials (300986.SZ) experienced a 20% drop limit on January 16, closing at 26.70 CNY per share, with a total market value of 11 billion CNY, following a significant increase of 198.57% over six consecutive trading days [2] - The company announced that its business does not involve AI applications, quantum technology, robotics, or commercial aerospace, and has not generated any related revenue since its establishment [2] - Zhite New Materials specializes in the research, production, and sales of aluminum molds, protective platforms, and prefabricated components, with no changes in its main business operations [2] Group 2 - The company confirmed that while AI technology has been utilized to shorten the R&D cycle for new materials, the AI-related technology is provided by partners, and the new materials are still in the optimization phase and have not yet reached mass production [3] - Zhite New Materials was established in 2011 and focuses on the research, design, production, rental, and technical services of formwork and prefabricated construction PC products [3] - For the first three quarters of 2025, the company reported revenue of 2.023 billion CNY, a year-on-year increase of 12.09%, and a net profit attributable to shareholders of 118 million CNY, reflecting a year-on-year growth of 98.65% [3]

在产品与技术层面，该公司持续深化"AIforScience"（AI赋能科技）战略，其自主研发的TKAG大模型已成功应用于灯检设 备，独创AI粒子轨迹跟踪算法可精准识别微小异物，成为制药智能化的标杆应用。此外，楚天科技生物制药耗材业务加速突 破，一次性反应袋、配液袋等产品已落地数千万元订单，楚天微球产品通过多家国际客户供应商审计，实验室及小试需求逐步 放量，有望成长为新的亿元级增长点。楚天科技子公司德国ROMACO集团经营状况也持续改善，核心业务指标稳步回升，协 同效应逐步释放。 楚天科技2025年全年业绩大幅扭亏有迹可循。2025年前三季度，公司海外业务收入同比快速增长，带动整体毛利率从2024 年的25.94%提升至31.17%。该公司管理层在与投资者互动中也明确表示，国际业务正面临"蓬勃发展的历史机遇"，未来将进一 步提升其在整体营收中的比重。 在国内市场，楚天科技主动优化订单结构，战略性收缩低毛利项目，聚焦高价值客户与高附加值产品。2025年前三季度， 公司综合毛利率维持在31%以上，单季净利润在第三季度达9676.81万元，实现单季扭亏。与此同时，公司通过研发端优化方 案、供应链降本、制造端提效等系统 ...

Group 1 - The core viewpoint emphasizes that the recent adjustment of Jingtai Technology is not a warning signal but rather a strategic entry point for investors in the AI sector, particularly in the first half of 2026 [1] - The market is transitioning from volatile fluctuations to a healthier correction, which is seen as a foundation for sustainable growth, aligning with the national "slow bull" market strategy [1] Group 2 - AI applications are accelerating from "concept validation" to "commercialization," with Jingtai Technology maintaining its leading position in the domestic AI application sector [2] - The recent market correction does not alter Jingtai Technology's role as a core engine in the industry, and as market sentiment recovers, funds will refocus on companies with technological barriers and commercial value [2] Group 3 - The "AI for Science" sector, where Jingtai Technology operates, is highly valued by the capital market and is supported by national policies aimed at achieving breakthroughs in basic research and high-end manufacturing [3] - The dual support from policy and capital creates a unique strategic value for Jingtai Technology in the industry landscape [3] Group 4 - Jingtai Technology possesses over 10 billion in cash reserves, providing strong risk resistance and the ability to expand during market downturns [4] - This financial strength allows the company to invest in cutting-edge research, attract top talent, and pursue strategic mergers and acquisitions [4] Group 5 - The recent entry of two AI large model companies into the capital market has boosted the overall AI sector, increasing market attention [5] - Jingtai Technology, as a leading enterprise in the high-barrier AI for Science field, has a clear business model and strong financial health, suggesting that its current valuation is relatively low [5] Group 6 - Historical trends indicate that top domestic AI research institutions often announce significant technological breakthroughs around the Spring Festival, which could act as a catalyst for a new wave of technological advancements in the AI field [6] - Jingtai Technology, as a key player in AI frontier technology, is expected to benefit from these breakthroughs and market dynamics [6] Group 7 - The current market correction is viewed as a valuable opportunity for investors to enter at relatively low levels, potentially marking the last chance to invest in a core leader in AI science before significant value appreciation [7] - Jingtai Technology's adjustment is seen as a natural correction within a healthy "slow bull" market, providing a "golden pit" for discerning investors [7]

就在刚刚，清华大学的一项AI for Science研究不仅登上Nature，而且还被Science深度报道了。 这项来自清华大学李勇团队的研究通过分析全球2.5亿篇科学文献，揭示了AI for Science领域存在的一个典型矛盾—— 第一步：寻觅AI for Science的演化踪迹 回到起点，团队之所以进行这项研究，主要是发现AI for Science领域存在一个明显矛盾—— 在AI持续赋能科研的背景下，为何各学科的整体科学进展未见明显加速？ 一方面，AI for Science研究已经产生了AlphaFold这样的荣获诺贝尔奖的成果；但另一方面，统计表明各学科领域的颠覆性研究成果在逐年下降，似乎未 能获得AI助力。 AI在助力科学家"个体加速"的同时，却导致科学界的集体注意力窄化和趋同优化的"群体登山"现象。 就是说，虽然AI帮助科学家发表了更多论文、更早成为项目负责人，但却导致人们集体涌入少量适合AI研究的"热门山峰"，从而无形中削弱了科学探索的 广度。 而且进一步分析表明，这一矛盾绝非偶然，而是由当前科学智能AI模型缺乏通用性导致的系统性影响。 下面详细来看这到底是一项怎样的研究。 这背后的原因 ...

证券研究报告：计算机|点评报告 行业投资评级 强于大市|维持 行业基本情况 ⚫ 复盘 3 年 25 倍的 PLTR，我们认为好的 AI 应用赛道应当 具备以下特征 | 收盘点位 | 5947.69 | | --- | --- | | 52 周最高 | 6063.95 | | 52 周最低 | 4080.58 | 近期研究报告 《国产 AI 或将全面超越》 - 2026.01.12 AI for Science，重在 AI+制药赛道 行业相对指数表现（相对值） -6% -1% 4% 9% 14% 19% 24% 29% 34% 39% 2025-01 2025-03 2025-06 2025-08 2025-10 2026-01 计算机 沪深300 资料来源：聚源，中邮证券研究所 研究所 分析师:陈涵泊 SAC 登记编号:S1340525080001 Email:chenhanbo@cnpsec.com 分析师:李佩京 SAC 登记编号:S1340525080003 Email:lipeijing@cnpsec.com 分析师:王思 SAC 登记编号:S1340525080002 Email:wangsi1@c ...

这篇中国团队领衔发布的论文，已经在外网刷屏了，仅一夜阅读就达到了200万！这位MIT博士回国创业后组建的团队，拉来全球24所顶级机 构，给AI如何助力科学发现来了一剂猛药。 最近，一篇由中国团队领衔全球24所TOP高校机构发布，用于评测LLMs for Science能力高低的论文，在外网炸了！ 当晚，Keras （最高效易用的深度学习框架之一）缔造者François Chollet转发论文链接，并喊出：「我们迫切需要新思路来推动人工智能走向科学创 新。」 AI领域KOL Alex Prompter分享论文核心摘要后，NBA独行侠队老板Mark Cuban跟帖转发，硅谷投资人、欧洲家族办公室、体育媒体同时涌进评论区。 仅一夜，累计阅读量逼近200万。 值得一提的是，同一时间窗里，OpenAI也发布了对于AI在科学发现领域能力评测的论文《FrontierScience: Evaluating Al's Ability to Perform Scientific Research Tasks》概述，指出现有评测标准在AI for Science领域失灵。 神同步OpenAI、海外讨论出圈，究竟是什么样的一份工作成 ...

一水 发自 凹非寺 量子位 | 公众号 QbitAI 就在刚刚，清华大学的一项 AI for Science研究 不仅登上Nature，而且还被Science深度报道了。 | Explore content | | --- | 这项来自清华大学李勇团队的研究通过分析全球 2.5亿篇 科学文献，揭示了 AI for Science领域存在的一个典型矛盾 —— AI在助力科学家"个体加速"的同时，却导致科学界的集体注意力窄化和趋同优化的"群体登山"现象。 就是说，虽然AI帮助科学家发表了更多论文、更早成为项目负责人，但却导致人们集体涌入少量适合AI研究的"热门山峰"，从而无形中削弱了 科学探索的广度。 而且进一步分析表明，这一矛盾绝非偶然，而是由当前科学智能AI模型 缺乏通用性 导致的系统性影响。 下面详细来看这到底是一项怎样的研究。 第一步： 寻觅 AI for Science的演化踪迹 回到起点，团队之所以进行这项研究，主要是发现AI for Science领域存在一个明显矛盾—— 在AI持续赋能科研的背景下，为何各学科的整体科学进展未见明显加速？ 在论文中，团队进行的首项工作是： 从浩如烟海的文献中找出那些 ...