证 券 研 究 报 告 国内AI纵深发展，悦己消费全球化 ——2026年互联网传媒投资策略 证券分析师：林起贤 A0230519060002 袁伟嘉 A0230519080013 任梦妮 A0230521100005 夏嘉励 A0230522090001 赵航 A0230522100002 黄俊儒 A0230525070008 研究支持：张淇元 A0230124080001 陈悦 A0230524100003 联系人：林起贤 A0230519060002 2025.11.28 26年投资展望：AI重估更深入，悦己消费时间换空间 www.swsresearch.com 证券研究报告 2 ◼ 2025年互联网和传媒的机会集中在AI重估（主要是云计算）和悦己消费（潮玩、音乐、演唱会等）全球化、 年轻化的机遇，我们认为产业机遇将延续数年，但在2025年涨幅较大情况下，2026年需要关注边际变化： • AI云：AI capex扩张第二年，伴随着国产化，仍未结束，但参考海外，交易将更重AI投入的ROI，capex/经营现金流 成为投资更关注的指标。重点：阿里巴巴、百度、金山云。 • AI应用：从概念到更加关注商业化落地， ...

Core Viewpoint - The article discusses the rapid rise of the AI application "Lingguang," which has achieved significant user engagement and showcases a shift towards practical AI solutions that cater to everyday needs [2][3][35]. Group 1: Lingguang's Launch and User Engagement - Lingguang was officially launched on November 18, and within six days, it reached 2 million downloads, surpassing the initial user acquisition rates of ChatGPT and Claude [3]. - The application quickly became popular, reaching the sixth position on the Apple App Store's free chart in China and first in the free tools category [3]. Group 2: Features and User Demographics - The success of Lingguang is attributed to its "flash application" feature, allowing users to create small applications through simple dialogue without coding skills, with some applications being developed in as little as 30 seconds [5][27]. - Initially popular among internet workers, Lingguang has expanded its user base to students and parents, with various applications emerging for tasks like homework assistance and time management [7][10]. Group 3: Addressing Information Overload - Lingguang aims to simplify complex information, addressing the issue of information overload by providing concise and effective summaries [15][16]. - The application generates not only text but also images and 3D animations, enhancing the user experience and making information more accessible [21][22]. Group 4: Evolution of AI Applications - The article highlights a shift in AI applications from mere conversational capabilities to practical problem-solving abilities, marking a transition to a new phase in AI development [24][27]. - Lingguang enables users to fulfill fragmented needs quickly, effectively breaking down barriers for non-technical users to create applications [27]. Group 5: Future Trends in AI - The article notes that major players in the AI industry are increasingly focusing on the integration of multi-modal capabilities and practical applications [29][30]. - Lingguang's development is seen as part of a broader trend towards practical AI solutions, with expectations for enhanced coding capabilities and deeper integration with existing ecosystems in the future [35].

本周一，阿里巴巴宣布，旗下"千问"App 公测一周，下载量即突破1000万。而这一增长速度已经超越ChatGPT、Sora、DeepSeek，成为截至目前 增长最快的AI应用。 此外，谷歌盘前大涨超4.2%，若开盘保持该涨幅，总市值将首次突破4万亿美元，成为第四家总市值突破过4万亿美元的上市公司。英伟达美股盘 前跌超4%，消息面上，据智通财经报道，谷歌人工智能芯片TPU获Meta洽购，价值数十亿美元。 （声明：文章内容仅供参考，不构成投资建议。投资者据此操作，风险自担。） 11月25日傍晚，阿里美股盘前一度大涨超4%。 消息面上，阿里巴巴公布了2026财年第二季度财报。财报显示，阿里巴巴第二季度营收2,478.0亿元人民币，预估2,452亿元人民币；第二季度调整 后每ADS收益4.36元人民币，预估6.34元人民币；第二季度中国电商事业集团收入1,325.8亿元人民币；调整后净利润103.5亿元人民币，预估168亿 元人民币。财报还显示，阿里即时零售业务收入达到229.06亿元人民币，较上年同期的143.21亿元增长60%。 | | | 截至9月30日正三個月 | | | | --- | --- | --- ...

以下文章来源于节点财经 ，作者节点财经 节点财经 . 来源 | 节点财经 硅谷的深秋通常是科技巨头们最为焦虑的季节，但对于谷歌而言，这种焦虑在昨天戛然而止。 在OpenAI发布其备受瞩目的GPT-5.1模型仅仅数日后，谷歌没有选择沉默，而是以一种近乎激进的姿态，发布了其史上最强大的 AI 系统 Gemini 3.0。 专注价值，聚焦增长，超 100 位上市公司老板关注的商业内参 这不仅仅是一次例行的版本更新，更像是一场精心策划的"诺曼底登陆"。 过去两年，谷歌在AI战场上一直扮演着"追赶者"的角色，被外界诟病为动作迟缓、由于大公司病而畏首畏尾。然而，Gemini 3.0的亮相彻底 改变了这一叙事。 凭借一项被称为"Deep Think"（深度思考） 的核心突破，以及一个旨在重塑软件开发行业的 Antigravity（反重力） 平台，谷歌试图向世 界证明：人工智能的下一个时代，不再是关于"谁能陪你聊天"，而是关于"谁能帮你思考"和"谁能帮你行动"。 从 快 思 考 到 慢 思 考 ， A I 的 逻 辑 觉 醒 如果你使用过之前的AI模型，可能会发现它们像一个反应极快但喜欢不懂装懂的实习生，你问一个问题，它立刻 ...

11月20日美沙投资论坛上，马斯克一开口就戳破了AI圈的"皇帝新衣"："AI瓶颈根本不是钱和算法—— 全球不缺投资人砸钱，缺的是能喂饱AI的电，还有装算力的机房。" 这话可不是吹牛：华盛顿州深夜数据中心，H100 GPU为GPT-6训练闪得比蹦迪灯还密；德州AI产业 园，Sora生成1秒视频的电，够1.7万个美国家庭用一天。 而巨头早用行动投票：微软签了20年核电协议，谷歌直接买核反应堆。明眼人都懂：AI要不停跑，得 靠核电当"永动机充电宝"。 01 大摩测算，2023-2027年全球数据中心耗电430-748太瓦时，占全球电力2%-4%，生成式AI电力需求年增 105%。 最扎心的是建设脱节：建一个数据中心要2年，建发电站3-5年，，而一条长距离高容量传输线居然要8- 10年。AI按年迭代，电网按十年建，根本追不上。 AI不是"电老虎"，是"24小时吞电怪" 谁能想到，靠代码起家的AI，现在成了电力系统的"噩梦"？ 它的硬件就透着"能吃"：普通服务器机架撑死14千瓦，AI机架直接飙到40-60千瓦——相当于把整栋楼 的电塞进一个铁皮柜。 大模型训练更夸张：GPT-5训练一次耗10万兆瓦时，够中型城市嗨一周— ...

Core Insights - The article emphasizes that the bottleneck for AI development is not funding or algorithms, but rather the availability of electricity and data centers to support AI operations [2] - Major tech companies are increasingly turning to nuclear power as a reliable energy source to meet the growing demands of AI [20][21] Group 1: AI's Energy Consumption - AI systems are consuming vast amounts of electricity, with a single training session for models like GPT-5 requiring 100,000 MWh, enough to power a medium-sized city for a week [3][6] - Daily operations of AI applications like ChatGPT consume over 500,000 kWh, which is 17,000 times the average daily electricity usage of a U.S. household [4] - The energy consumption for inference operations can exceed that of training, leading to a long-term energy demand that is unsustainable without reliable power sources [5] Group 2: Current Energy Landscape - Data centers in the U.S. currently account for 2.5% of total electricity consumption, projected to rise to 15% by 2028, with global data center energy demand expected to grow by 105% annually due to AI [6] - The existing energy infrastructure is struggling to keep pace with AI's rapid growth, with significant delays in building new power plants and transmission lines [6][11] - Renewable energy sources like wind and solar are not sufficient to meet AI's continuous power needs, as their utilization rates are low [7][9] Group 3: Nuclear Power's Resurgence - Nuclear power is gaining traction as a stable energy source for AI, with a utilization rate of 92%, making it a reliable option for continuous operation [14][18] - Major companies like Microsoft and Google are investing in nuclear energy, signing long-term agreements for nuclear power to support their AI data centers [20][21] - The integration of AI into nuclear operations can enhance efficiency and reduce operational costs, making nuclear power more attractive [23][24] Group 4: Future Outlook - The demand for nuclear energy is expected to increase significantly as AI capabilities expand, with projections indicating a 3-5 times rise in nuclear power needs by 2030 [29][30] - The collaboration between AI and nuclear power is seen as a mutually beneficial relationship, where AI can optimize nuclear operations while nuclear power provides the necessary energy for AI [33][37] - The article concludes that the future of AI is closely tied to nuclear energy, positioning it as a critical component for sustaining AI's growth [38]

Core Insights - The article highlights a significant strategic move by Alibaba and Ant Group in the AI sector, with the launch of two AI applications, "Lingguang" and "Qianwen," aimed at enhancing personal productivity and user engagement in the AI market [1][2][3] Group 1: Launch of AI Applications - On November 17, Alibaba launched the public beta of the "Qianwen" app, aiming to create a personal AI assistant that can chat and perform tasks, targeting the future of personal AI life [1][8] - The following day, Ant Group introduced "Lingguang," an AI assistant that allows users to generate applications in 30 seconds using natural language, marking a significant breakthrough in AGI application [2][5] Group 2: Performance and Features of "Lingguang" - "Lingguang" achieved impressive download milestones, surpassing 200,000 downloads on its first day and reaching over 1 million downloads within four days, outperforming early metrics of global AI products like ChatGPT [5][6] - The app features three core modules: "Lingguang Dialogue," "Lingguang Open Eye," and "Lingguang Flash Applications," enabling users to create interactive tools quickly and providing a visual interaction experience [6][7] Group 3: Strategic Positioning of "Qianwen" - The "Qianwen" app is positioned as a comprehensive AI assistant, directly competing with ChatGPT by integrating various life scenarios such as maps, food delivery, and shopping [8][17] - Alibaba's strategy reflects a shift from focusing on B2B to targeting ordinary users (C-end), aiming to convert technical advantages into a seamless user experience [12][13] Group 4: Competitive Landscape and Market Dynamics - The competition in the AI application market is evolving from measuring daily active users to focusing on who can meet core user needs effectively [20][21] - The article notes that the AI market has shifted from a blue ocean to a red ocean, with Alibaba's dual strategy of "Qianwen + Lingguang" aiming to capture the C-end market and challenge established players like ByteDance and Tencent [25][26]

格隆汇11月24日｜午后，以传媒、游戏板块为首的AI应用板块强势拉升，规模最大的游戏ETF涨3.5%， 机器人ETF翻红上涨1%。 AI+游戏产能升级：游戏ETF(159869)，+3.57%，最新规模108.85亿，近20日合计净流入6.68亿元，权重 股涵盖囊括吉比特、恺英网络、巨人网络、三七互娱（维权）、神州泰岳等游戏头部玩家，以及光线传 媒（IP 优势突出）、昆仑万维（维权）（爆款游戏《梦幻星球》+AI 技术领先）等。 ③网易第三季度总营收284亿元，同比增长8.2%；净利润86亿元，同比增长32.3%，绩后获多家大行唱 好。 ④工信部公示人形机器人标准化技术委员会委员名单，宇树科技王兴兴、智元创新彭志辉等为副主任委 员。 开源证券最新研报认为，谷歌和阿里巴巴近期在模型和Agent方面的进一步发力，并将更多的垂直AI应 用整合进自身的流量入口，从而有望大幅加速AI应用商业化，其他同样具备优势的科技巨头（如微 软、Meta，腾讯、字节）相关举措也值得关注，AI应用或迎来黄金时代，继续布局商业化落地较快的 AI游戏/AI广告/AI电商等。 相关产品，及截至发稿涨跌幅： 市场规模最大的机器人主题ETF：机器 ...

Core Insights - Alibaba's "Qianwen" project has officially launched, marking its entry into the AI to C market, and has quickly become the fastest-growing AI application in history, surpassing competitors like ChatGPT and DeepSeek [4][5][9] Group 1: Market Performance - Following the announcement of Qianwen, Alibaba's stock surged by 4.13% by midday [3] - The Qianwen app reached the fourth position on the Apple App Store's free applications chart within a day of its public beta launch, causing server congestion due to high traffic [5][6] - By November 19, just two days after its launch, Qianwen climbed to the third position on the App Store [6] Group 2: Competitive Landscape - Qianwen's download speed has significantly outpaced other popular AI applications, achieving over 10 million downloads faster than ChatGPT and DeepSeek [7][8] - The Qwen model, which powers Qianwen, has become a leading open-source model globally, with over 600 million downloads, and is recognized for its superior performance compared to competitors like Llama and DeepSeek [9] Group 3: Strategic Vision - Alibaba views Qianwen as a critical component in the "AI era future battle," aiming to establish a consumer-facing AI entry point [10] - Analysts suggest that Qianwen's initial success is just the beginning, with potential for further growth through subscription models and integration with Alibaba's other services [10] - The app is positioned as an "Agentic AI" capable of understanding and executing complex tasks, indicating a shift from passive AI tools to proactive AI agents [11]

Group 1 - Alibaba's stock price surged over 5% on November 24, closing at HKD 153.70 per share, with a total market capitalization of HKD 2.93 trillion [1] - The AI assistant app "Qianwen" launched by Alibaba has surpassed 10 million downloads within a week of its public beta, making it the fastest-growing AI application historically, outpacing competitors like ChatGPT and DeepSeek [1] - The app's rapid rise led to it reaching the fourth position in the Apple App Store's free applications chart, causing server congestion and trending on social media [1] Group 2 - Huayuan Securities highlights the resilience of leading internet companies, including Alibaba, Tencent, Meituan, Kuaishou, Tencent Music, NetEase Cloud Music, and Bilibili, emphasizing the strategic value of their internal organizational adjustments [2] - The report stresses the importance of AI foundational technology development and the execution of AI application products as core elements for industry growth and market transactions [2] - Continuous attention is recommended on the strategic value of leading companies' internal adjustments and their progress in AI and other long-term narratives [2]