Core Viewpoint - The article discusses the transformation of the tour guide profession in China, highlighting how individuals from various backgrounds are transitioning into this role due to its flexibility, potential income, and the growing demand for quality cultural storytelling [2][5][23]. Group 1: Personal Transformation Stories - A case study of a former tech employee, Zhu Liang, illustrates the shift from a corporate job to becoming a tour guide, emphasizing the newfound freedom and enjoyment in storytelling [2][4][22]. - Zhu's journey reflects a broader trend where many individuals are exploring tour guiding as a viable career option after leaving traditional jobs [4][14]. Group 2: Industry Dynamics - The tour guide profession has evolved from a basic service role to one that requires storytelling skills and cultural knowledge, making it a more respectable career choice [5][10]. - The entry barrier is relatively low, requiring only a tour guide certification, but excelling in the role demands significant skill in engaging and educating tourists [7][8]. Group 3: Income Potential - Current earnings for tour guides vary, with average daily wages for English-speaking guides ranging from 600 to 800 RMB, while guides for less common languages can earn up to 200 USD per day [9][10]. - The flexibility of the role attracts many individuals who prefer not to work traditional hours, allowing them to start part-time and transition to full-time if desired [10][12]. Group 4: Market Competition and Expectations - The market for tour guides is becoming increasingly competitive, with many new entrants, including students and recent graduates, who are willing to work for lower wages [16][17]. - Tourists, especially international visitors, have higher expectations for their experiences, demanding more engaging and accurate storytelling from guides [17][19]. Group 5: Evolving Role of Tour Guides - The profession is diversifying into different service models, from low-cost, high-volume guides to content-driven "IP-type" guides who leverage social media for marketing [18][19]. - Successful guides are now focusing on tailoring their content to meet the specific interests of their clients, enhancing the overall experience [20][21].

HUAWEI X HUXIU 没有人不在期待大模型能够成为下一个电动车，作为代表中国的新兴产业，在世界范围内掀 起狂澜。 而作为另一加速推理的技术，量化技术是DeepSeek模型采用FP8进行训练的核心，而推理时 采用Int8部署，亟需突破FP8到int8的无损量化，为此华为团队也提出了OptiQuant量化框架和 算法，让INT8性能可以打平FP8。 下面就分别介绍华为团队的这两个创新技术。 然而主流的MoE架构大模型，却苦于其结构上的"先天不足"：巨大的硬件成本与多重拖累效 率的环节，使得中国企业在这场芯片堆砌与效率挖掘的苦径上难以提速。 作为智能基础设施提供商，华为在这场战役中另辟蹊径，利用其在数学算法和工程领域的深 厚积累，为DeepSeek显著提升了效率及用户体验。 山就在那里，但中国企业找到了不一样的登顶之路。 近期，虎嗅将打造《华为技术披露集》系列内容，全面揭秘超大规模MoE模型推理部署技 术，通过一连串的技术报告，首次全面披露技术细节。 希望本系列内容能为业界起到参考价值，也希望更多人能与华为一起，共同打造长期持续的 开放协作生态环境，让昇腾生态在中国茁壮成长。 《华为技术披露集》系列 VOL ...

Core Insights - Google I/O showcased significant advancements in AI, particularly with the Gemini model, which is set to take over various Google services and enhance user interaction through AI-driven features [4][9][30]. Group 1: Google Glass Revival - The revival of Google Glass was a highlight of the event, demonstrating its capabilities through live demos that showcased real-time interaction and AI integration [6][7]. - The Android XR glasses featured Gemini's visual memory, allowing users to interact with their environment and receive contextual information seamlessly [8][9]. Group 2: Gemini's Dominance - Gemini has established itself as a leader in AI capabilities, with over 400 million monthly active users and a 50-fold increase in token processing [13][15]. - The model's performance has significantly improved, with a 300-point increase in Elo scores and a tenfold enhancement in TPU performance [15][16]. - Google’s search business remains robust, with AI integration driving user engagement and query complexity [12][32]. Group 3: Search Transformation - Google is transforming its search functionality by integrating Gemini, allowing for deeper exploration of queries and enhanced user experience through AI Mode [30][32]. - The introduction of features like virtual try-ons in Google Shopping demonstrates the potential for AI to revolutionize e-commerce interactions [33]. Group 4: New AI Tools and Features - The launch of Flow, a new app for video creation, highlights Google's commitment to empowering creators with advanced AI tools [37][39]. - Gemini's capabilities are being expanded with new features, including real-time interaction and enhanced audio-visual outputs, making it a versatile tool for users [20][26]. Group 5: Future Directions - Google aims to make Gemini a proactive assistant, capable of anticipating user needs and providing timely suggestions [25][24]. - The integration of Gemini across Google’s ecosystem is expected to enhance the functionality of various products, positioning them as universal agents [27][28].

以下文章来源于数字生命卡兹克 ，作者数字生命卡兹克 数字生命卡兹克 . 努力分享一些很新、很酷的AI干货，愿我们永远对世界保持好奇。 本文来自微信公众号： 数字生命卡兹克 ，作者：数字生命卡兹克，题图来源：AI生成 在今年，DeepSeek R1火了之后，几乎快形成了一个共识，就是： AI推理能力越强，执行任务时就应该越聪明。 从2022年Chain-of-Thought横空出世，到今天Gemini 2.5 Pro、OpenAI o3、DeepSeek-R1、 Qwen3，这些旗舰模型的统治性表现，我们一直相信，让模型先想一想，是一个几乎不会出错的策 略。 不过，这种聪明，也会带来一些副作用。 就是提示词遵循能力，变得越来越差。 换句话说，就是越来越不听你的话了。 我在过年期间写DeepSeek的攻略文：《DeepSeek的提示词技巧，就是没有技巧》的时候，也提到了 这一点。 不过，这只是我自己使用中的感觉，它变得越来越聪明，但是感觉却越来越不听话了，以至于我现 在，最常用的模型，开始变成了GPT4o，所有的推理模型，反而会用的越来越少了。 不过，确实没有经历过验证，所以也不是特别敢说。 直到昨晚回来，在扒拉 ...

Core Viewpoint - Alibaba's financial performance for the fiscal year 2025 shows a revenue of 996.3 billion, a year-on-year growth of 5.9%, with a significant increase in net profit by 76.6% to 126 billion, driven by strong core business resilience and investor confidence in its AI capabilities [3][4]. Revenue Structure - The revenue structure of Alibaba's six major groups for fiscal year 2025 is as follows: Taotian Group (449.8 billion), International Digital Commerce Group (132.3 billion), Cloud Intelligence Group (118 billion), Cainiao Group (101.3 billion), Local Life Group (67.1 billion), and Entertainment Group (22.3 billion) [7][9]. - Taotian Group dominates with 45.1% of total revenue, while International Digital Commerce and Cloud Intelligence contribute 13.3% and 11.8% respectively, indicating a slight shift in rankings compared to the previous fiscal year [9]. Revenue Growth Drivers - The revenue growth of 81.5 billion in fiscal year 2025 was contributed by all six major groups, with International Digital Commerce leading with a growth of 29.4 billion, followed by Taotian Group with 14.9 billion [10][11]. - The international digital commerce segment is expected to remain a key driver for revenue growth, despite external uncertainties [20]. Taotian Group Performance - Taotian Group's revenue is primarily from "China Retail" and "China Wholesale," with retail accounting for over 90%. In fiscal year 2025, retail revenue was 332.35 billion, showing a growth of 11.8% [13][15]. - The restructuring of Taotian Group has led to improved profitability, with EBIT increasing to 196.2 billion and a profit margin of 43.5% in fiscal year 2025 [18]. Cloud Intelligence Insights - Alibaba Cloud, as the largest public cloud service provider in China, reported a revenue of 77.2 billion in fiscal year 2023, with a growth of 4%. The introduction of AI capabilities has significantly boosted service uptake [22][24]. - The cloud segment's profitability has improved steadily, with EBIT reaching 10.6 billion and a profit margin of 8.9% in fiscal year 2025 [24]. Strategic Outlook - The cross-border e-commerce segment is viewed as a "second growth curve," with significant potential for expansion, leveraging lower costs and a wider range of products compared to traditional trade [27]. - The integration of AI and cloud services is expected to enhance business operations and valuation, with Alibaba's current P/E ratio at 16 times, indicating potential for significant valuation uplift as AI capabilities are recognized [27][28].

大家早上好！这里是今天的早报，每天早上，我都会在这里跟你聊聊昨夜今晨发生了哪些大事儿。 本 栏目由虎嗅出品。 热点追踪 【调查显示：00后超4成认为结婚不是人生必选项】 据长安街知事，中国人民大学社会与人口学院教授李明认为，单纯的经济激励难以解决深层次的社会问题。现 代年轻人对婚姻的态度已发生根本性转变，他们更看重婚姻质量而非形式。调查显示，认为"结婚不是人生必 选项"的比例，90后群体中为31%，00后中高达43%，而在80后中仅为19%，代际差异昭示，传统人口政策亟待 系统性革新。 【多家银行下调大额存单产品利率，降至"1字头"】 据界面新闻，新一轮存款降息来袭，包括工行、中行、建行等多家银行纷纷下调了活期存款、定期存款以及通 知存款等多个产品利率。随着银行存款"降息潮"的延续，曾被视为"揽储利器"的大额存单产品，正在逐渐"失 宠"。近期，有多家银行在调降定期存款利率的同时，也同步下调了大额存单产品利率。目前，大额存单产品 即将全面迈入"1时代"。 【黄杨钿甜艺考成绩被网友质疑有水分】 据上游新闻，近日，因"天价耳环"深陷舆论漩涡的年轻女演员黄杨钿甜，再因一段高中毕业典礼献唱视频引发 关注。5月19日，某 ...

以下文章来源于诗翰 ，作者卢诗翰的赞赏小号 首先是壁装电吹风壁装化妆镜的普及，很多物品连在墙上无法带走，剩下的毛巾牙刷烧水壶等价值不 高，其次是大众生活水平提高，偷东西的人大大减少，最后是酒店服务员的人工成本也变高了。 最终酒店经过计算发现，派一个服务员去检查房间是否丢失物品这个行为的成本，可能会高于丢失物 品本身。 一个月下来丢不了几个物品，但为此却需要多招一个服务员。所以干脆大家退房时不查房了，交了房 卡直接走人。 诗翰 . 世界是我们的游乐场 本文来自微信公众号： 诗翰 （ID：lixingdeqinggan1） ，作者：卢诗翰，头图来自：AI生成 对于当下的消费行业和创业者来说，我认为有一个知识点可能是必修课，我将其命名为——《白嫖经 济学》。 近期的一系列社会现象，比如女司机投诉女乘客，女装商家大战女装用户，淘宝屏蔽高退款人群等， 都可以用这个经济模型来解释。 白嫖经济学的核心原理是，对于一些行业来说，监督成本会大于监督收益，因此会出现白嫖空间。 比如酒店行业，前些年酒店行业退房的时候，酒店是会派服务员去检查房间是否丢失物品的。 但这两年这个操作基本绝迹，为什么？ 你说丢东西怎么办，哎呀没几个钱，给 ...

Core Viewpoint - The article discusses the conflict between a food influencer, Sui Bian, and the MCN (Multi-Channel Network) that previously managed his account, highlighting issues of revenue sharing, account ownership, and creator rights in the digital content industry [4][15][17]. Group 1: Conflict Overview - Sui Bian, a food influencer, gained popularity through his original account "Te Chu Sui Bian" but left the MCN after their contract expired, leading to a legal dispute over account ownership and competition agreements [9][11]. - The MCN claims to have invested hundreds of thousands in developing Sui Bian's brand and asserts that the account ownership should remain with them, citing a non-compete clause [15][22]. - Legal experts suggest that the enforceability of non-compete clauses in this context may be questionable, as they often require compensation and a formal employment relationship [15][17]. Group 2: Revenue Sharing and Labor Conditions - Reports indicate that Sui Bian received only 500 RMB per commercial deal, while the MCN charged clients between 48,000 to 85,000 RMB for promotional videos, raising concerns about fair compensation [16][33]. - The contract required Sui Bian to produce 16 to 30 videos per month, which some critics label as "exploitative" working conditions [16][35]. - The MCN claims Sui Bian failed to meet performance targets, while he argues that the compensation structure was inadequate for the workload [16][22]. Group 3: Public Perception and Content Authenticity - The public largely supports Sui Bian, viewing the MCN's tactics as capital exploitation, and many believe that the editing and scripting of content are standard practices in the industry [17][18]. - The MCN attempted to release unedited footage to demonstrate Sui Bian's reliance on their support, but this backfired, leading to accusations of "self-exposure" and further damaging their reputation [17][31]. - Sui Bian's new account focuses on traditional culinary culture and has quickly gained followers, indicating a successful rebranding despite the conflict [17][18]. Group 4: Industry Implications - The situation highlights the need for MCNs to evolve from a controlling model to a service-oriented approach that respects creators' brand value and rights [17][58]. - Legal recommendations suggest that creators should clarify account ownership, name usage, and content control in contracts to avoid future disputes [17][18]. - The case may influence future regulations regarding creator rights, particularly with the proposed "Account Personality Rights" concept in upcoming legislation [18].

Core Viewpoint - The article discusses the challenges and advancements in the development of large language models, particularly focusing on the MoE (Mixture of Experts) architecture and how Huawei has innovated to enhance its performance and efficiency in this domain [1][4]. Group 1: Challenges of MoE Models - The MoE architecture faces significant challenges, particularly the "cold and hot expert" phenomenon, which leads to uneven load distribution and affects system performance [4][3]. - The uneven load results in increased inference latency and limited throughput due to underutilization of resources [4][3]. Group 2: Huawei's Innovations - Huawei has introduced an efficient load balancing strategy called OmniPlacement, which significantly improves the inference performance of MoE models through expert reallocation, inter-layer redundancy deployment, and near-real-time dynamic scheduling [7][6]. - The OmniPlacement algorithm optimizes the deployment order based on expert activation data, reducing the load imbalance and enhancing system performance [7][6]. Group 3: Key Features of OmniPlacement - The framework supports dynamic priority adjustment and communication domain optimization, which reduces communication overhead compared to traditional static allocation methods [7][9]. - It includes a near-real-time scheduling and dynamic monitoring mechanism that allows for efficient expert allocation and minimizes inference delays [10][9]. Group 4: Experimental Results - Testing on the DeepSeek-V3 model showed that OmniPlacement reduced inference latency by approximately 10% and increased system throughput by about 10%, demonstrating significant improvements in resource utilization [14][14]. - The system maintained stability under dynamic input and high-load conditions, ensuring no performance fluctuations or service interruptions [14][14]. Group 5: Future Directions - Future research will focus on optimizing scheduling algorithms, developing adaptive expert selection mechanisms, and expanding the OmniPlacement framework to support more types of MoE models [15][15]. - The release of OmniPlacement marks a significant advancement in MoE model inference performance and highlights Huawei's competitive edge in AI computing [15][15].

Core Viewpoint - The article discusses the rapid rise and fall of low-cost food businesses, specifically 2 yuan bread and 7 yuan beef noodle shops, highlighting the allure of quick profits that ultimately leads to many entrepreneurs facing significant losses [3][4][5]. Group 1: Market Trends - In 2023, there has been a surge in 2 yuan bread and 7 yuan beef noodle shops, particularly in cities like Guangzhou, where multiple shops can be found within close proximity [3][7]. - The expansion of 2 yuan bread shops is projected to exceed 4,300 new locations annually from 2023 to 2025, with cities like Jinan and Changsha seeing significant growth [7]. - Despite the initial popularity, many of these businesses are closing quickly, with reports of shops shutting down within days or months of opening due to poor performance [4][12][13]. Group 2: Entrepreneurial Challenges - Many entrepreneurs, particularly those new to the food industry, are experiencing regret after investing in low-cost food ventures, with some reporting losses of up to 150,000 yuan within months [4][12]. - The allure of low investment and quick returns has attracted numerous inexperienced individuals, leading to intense competition and a "fast growth, fast death" scenario in the market [19][22]. - The lack of experience among many new entrants contributes to poor decision-making, particularly in site selection, which is critical for the success of these businesses [20][22]. Group 3: Financial Dynamics - The business model of 2 yuan bread and 7 yuan beef noodle shops is based on low customer spending and high turnover, with claims of daily revenues reaching up to 20,000 yuan [15][16]. - Initial investment costs for these ventures are relatively low, with 2 yuan bread shops requiring around 10,000 to 20,000 yuan to set up, while beef noodle shops can range from 15,000 to 25,000 yuan [17]. - Many entrepreneurs are misled by inflated revenue claims during the franchise recruitment process, leading to unrealistic expectations and subsequent financial losses [20].