重磅 ！ 重磅 ！ 重磅消息来了 ！ 京东竟然也造车了 ！ 京东宣布联合广汽集团、宁德时代推出 一款神秘国民好车 ，瞬间在汽车行业激起千层浪。京东 在造车项目中的角 色定位为 生态整合者 ， 主要 通过 机器学习模型 （如DIN、MMoE） 挖掘消费者偏好，精准定义车型配置与定价策略， 优化 购车流程。 探索 大模型 （如LLaMA、GPT） 在客服、风控等场景的应用， 提升 后市场 服务智能化 水平。 刚刚看了下京东汽车发布的招聘岗位，近半年京东汽车零售事业部发布了大量 算法工程师 的岗位 ，年薪可达123万元（35k-65k， 19薪） ， 但对应聘者的技术能力要求较高，需具备 算法、深度学习及大模型 等领域的 成熟经验 。 现在不仅仅是京东，其他大厂也在大力 扩充AI岗位， 吸收AI人才 ，很多算法岗的 薪资甚至比往年提高40%， 而对于正在求 职，转型的各位同学来说，现在正是 最适合 拿下AI算法岗的红利期！ 为了帮助大家更好的冲刺 算法岗 ，拿到 高薪offer ，我们特地打造了一个 《AI算法工程师培养计划》 ，它是 由国内一线 大厂在职的算法leader亲自主讲 ，确保你学习完以后，能够具备在实 ...

Core Insights - Microsoft Azure experienced a significant global outage on October 29, 2025, affecting numerous services and users worldwide [6][10][18] - The outage coincided with the release of Microsoft's Q1 FY2026 earnings report, which showed a 40% year-over-year revenue growth for Azure and other cloud services, marking it as the fastest-growing segment [8][10] - The incident raised concerns about the centralized risks associated with cloud services, as both AWS and Azure faced outages within a short timeframe, highlighting systemic vulnerabilities in the industry [29][30][32] Summary by Sections Outage Details - The outage began at 16:00 UTC and was expected to last until 23:20 UTC, totaling nearly 9 hours [6][26] - Downdetector reported thousands of outage reports from various regions, indicating the widespread impact of the incident [2] Affected Services - Key services impacted included Azure Active Directory, Azure SQL Database, and Microsoft 365 applications, among others, affecting both Microsoft’s core services and third-party applications [11][10] - Major companies like Alaska Airlines and healthcare institutions in Quebec reported service disruptions due to the Azure outage [13][15] Cause of the Outage - The outage was traced back to an accidental configuration change in Azure Front Door, which disrupted content delivery and caused downstream service failures [18][19] - Microsoft halted all new configuration changes to prevent further issues and began deploying a known stable configuration [22] Industry Implications - The recent outages at both AWS and Azure have intensified discussions about the risks of cloud service centralization, with experts warning that such incidents could lead to significant financial losses for businesses [30][32] - The need for improved resilience in cloud infrastructure has been emphasized, suggesting that companies should consider redundancy and multi-cloud strategies to mitigate risks [32]

转自：喵大白话 以为月薪过万都很忙很认真工作 我来这个公司前半生 一直被牛马 转岗后 虽然不是天天 但是忙一阵松一阵 他们一直都很闲 以上是我的嫉妒 ükel ükel ükel okel okol 2 公众号 · 喵大白话 电脑只是摆设 每天到处晃悠 2 公众号 · 喵大白话 昨天我意外知道他 月薪居然是w 首先他不是搞技术的 也不是搞研发 7496 我日常60% 的时间摸鱼,过得浑浑噩 噩,真觉得没意思,双休,过万。天天 不是刷小红书就是水群,太难受了。 我跟你拼了 1.7万 我月薪大概1.7w,每天就工作2小时 @早 3879 九晚五,公司管饭圈因为玩手机太多, 眼睛干涩,上周还去查了眼科 @再这么 呆着人都要废了: 那好吧 好的家人们我破放了困 5260 % 公众号 · 喵大白话 感觉很多都是工资越低的越肯干,工资 越高的越清闲合当然也分岗位,有很多 低门槛的岗位本身就是钱少事多,都是 一些烦人的琐事,但有的岗位确实是真 才实学真干事换的工资 一般工资高的都做决策性工作啦,工 838 资低的做执行工作 月薪到手2w+，招下属会给他们争取工 资,尽量给到区间内最大(1w+),每天 工作4h左右,强调效率 ...

一水 鹭羽 发自 凹非寺 量子位 | 公众号 QbitAI 不妨来直观感受一下它的生成速度： 最新开源 LongCat-Flash-Omni ，从名字你也能看出来了 （ Omni 意为 " 全能的 " ） ，没错，这款模型终于支持多模态了！ 任务虽更复杂，但模型实力不减，一出手依旧是"开源即SOTA"： 在综合性的全模态基准测试 （如 Omni-Bench, WorldSense ） 上，超越Qwen3-Omni、Gemini-2.5-Flash，这款模型直接达到了开源SOTA 水准，而且能和闭源的Gemini-2.5-Pro相媲美。 即使单拉出来文本、图像、音频、视频等各项模态能力，它也依旧能打 （单项能力均位居开源模型前列） ，真正实现了"全模态不降智"。 而且啊，这款模型还有一个亮点，那就是"快" （天下武功唯快不破？） —— 继承LongCat-Flash系列"快"的基因，这款Omni模型总参数560B，激活参数仅27B，这种"大总参小激活"的MoE架构，使其在保持庞大知识 容量的同时，实现了极高的推理效率。 用外卖的打法做AI模型？美团这是跟"又快又稳"杠上了（doge）。 两个月哐哐发了多款模型后 ...

Group 1 - The article discusses various personal anecdotes related to family professions, highlighting unique and humorous perspectives on traditional jobs such as funeral services and poultry supply [2][5][12]. - It features a light-hearted tone, with readers sharing their experiences and reactions to different professions, showcasing a blend of humor and cultural commentary [20][24][30]. Group 2 - The content includes interactions and comments from readers, reflecting a community engagement around the shared experiences and humorous takes on everyday life [22][37][45]. - There are references to social media trends and the impact of digital culture on personal narratives, indicating a shift in how people communicate and share their stories [8][14][40].

Core Viewpoint - The article discusses the organizational structure and key personnel reporting directly to CEO Jensen Huang at Nvidia, highlighting the strategic focus on hardware and AI, as well as the implications of a flat management structure in a rapidly growing company [1][4]. Group 1: Organizational Structure - Nvidia has 36 direct reports to CEO Jensen Huang, divided into seven functional areas: strategy, hardware, software, AI, public relations, networking, and an executive assistant [1][3]. - The hardware segment remains a cornerstone of Nvidia's business, with one-third of Huang's direct reports focusing on hardware-related operations, including GPUs and systems [6][7]. - AI and emerging technologies are becoming a second pillar in Huang's business strategy, with seven key leaders in this area [8][10]. Group 2: Key Personnel - Jonah Alben, a long-time Nvidia executive, is recognized as the "soul of GPU architecture," having been with the company for 28 years and holding 34 patents [23][30]. - Dwight Diercks, another veteran with 31 years at Nvidia, oversees software engineering and has significantly expanded his team over the years [32][37]. - Bill Dally, Nvidia's Chief Scientist, has been instrumental in evolving GPUs into general-purpose parallel computing platforms and has a strong academic background [42][47]. Group 3: New Additions - Wu Xinzhou, the only Chinese executive directly reporting to Huang, is responsible for Nvidia's automotive business and has a strong background in autonomous driving technology [56][59]. - Under Wu's leadership, Nvidia's automotive revenue is projected to nearly double from $281 million to $567 million in the 2024-2025 fiscal year [71]. Group 4: Management Philosophy - Huang advocates for a flat organizational structure to enhance information flow and decision-making speed, contrasting with traditional hierarchical models [80][86]. - Despite the increase in direct reports, Huang's management style emphasizes transparency and collective communication rather than one-on-one meetings [82][83]. - The rapid growth of Nvidia's workforce, from 29,600 to 36,000 employees in one year, has led to challenges in maintaining a flat structure, suggesting a potential shift towards a more vertical organization [100][114].

整理 | 屠敏 一次全球性宕机，波及各行各业 微软确认，Azure 自 2025 年 10 月 29 日（周三）UTC 时间 16:00（太平洋时间 09:00）起出现大范围中断，预计要到当日 UTC 时间 23:20（太平洋 时间 16:20）才能完全恢复。 出品 | CSDN（ID：CSDNnews） 亚马逊的 AWS 刚"崩"完没多久，微软的 Azure 也崩了......这让很多人好奇，最近的云厂商到底怎么了？ 日前，大量用户在 X（原 Twitter）、Hacker News、Reddit 等社交平台上报告称，微软 Azure 出现大规模故障，连 Azure 官网、 microsoft.com 都一度无法访 问。 根据故障追踪网站 Downdetector 的统计，仅数小时内，全球多个地区的报告就累计上千起报告，显示这次中断影响范围之广，堪称一次全球性事件。 尴尬的是，这次宕机恰逢微软发布 2026 财年第一季度财报之际 （微软的财年并不与日历年同步，其 2026 财年从 2025 年 7 月 1 日开始到 2026 年 6 月 30 日结束） 。 财报显示，Azure 及其他云服务的收入较去年同期 ...

Core Viewpoint - The article discusses a bizarre incident during the GTC 2025 conference where an AI-generated version of NVIDIA's CEO Jensen Huang outperformed the real Huang in a fraudulent live stream, attracting a significantly larger audience and engaging in scams [1][2][6]. Group 1 - During the GTC 2025 conference, Jensen Huang delivered an enthusiastic speech, while an AI-generated version of him was simultaneously live-streamed, drawing nearly 100,000 viewers [2][6]. - The fraudulent live stream, branded as "NVIDIA LIVE," attracted eight times more viewers than the official broadcast, highlighting the effectiveness of the scam [6][15]. - The AI-generated Huang not only mimicked the real Huang's appearance and voice but also engaged in fraudulent activities, deceiving many viewers [9][10][23]. Group 2 - The orchestrators of the scam, a channel named "Offxbeatz," utilized AI face-swapping and voice synthesis technologies to create a convincing imitation of Huang, capitalizing on the audience's interest in the GTC conference [19][20]. - Many viewers, including notable media outlets like CNBC, were misled by the realistic portrayal of the AI-generated Huang, resulting in significant financial losses, with approximately $115,000 (around 820,000) stolen [23][25]. - The incident underscores the advancements in AI technology, which now allow for the easy creation of realistic deepfake videos, raising concerns about the potential for similar scams in the future [42][44].

Group 1 - The article discusses the humorous and relatable experiences of individuals regarding their relationships and interactions, highlighting the quirks of human behavior in romantic contexts [1][2][3] - It emphasizes the idea that people often keep track of details about their partners to remind themselves of their feelings and priorities [2][4] - The commentary includes various anecdotes and reactions from readers, showcasing a mix of humor and personal reflections on relationships [6][7][8] Group 2 - The article features a nostalgic look at the influence of teachers and the lasting impact they have on students, even years after graduation [8][9] - It highlights the importance of educators who continue to support their former students by providing resources and guidance [8][10] - The emotional connection between teachers and students is underscored, illustrating how these relationships can shape one's life and career [8][11]

Core Viewpoint - The article discusses a peculiar bug in Windows systems where users are prompted to "eject" their NVIDIA graphics cards, which is not a standard feature and may lead to temporary issues with video output and performance [1][10]. Group 1: Bug Description and User Experience - Users reported seeing a prompt to "eject NVIDIA GeForce RTX XX graphics card" after updating their Windows system, indicating a misidentification of the graphics card as a removable device [1][10]. - Clicking the eject option may cause a brief black screen on the monitor, but typically does not damage the hardware [2]. - Desktop PCs with discrete graphics cards will lose video output if the card is mistakenly removed, while devices with integrated graphics may not notice any change [3]. Group 2: Impact on Laptop Users - For laptops, temporarily disabling the discrete graphics card can lead to decreased graphics performance, such as lag in gaming or design software, but functionality is restored upon reboot [4]. - Some laptop users have reported that this bug unexpectedly improves battery life, with one user noting that their Lenovo Y9000X gaming laptop achieved up to 8 hours of battery life after "ejecting" the discrete graphics card, compared to only 3-4 hours previously [5][6]. Group 3: Historical Context and Solutions - This bug has been present for several years, with reports dating back to the Windows 10 era, indicating a long-standing issue that has not been fully resolved [6]. - The problem is believed to be related to the collaboration between Microsoft and NVIDIA regarding driver compatibility, necessitating a joint resolution [10]. - Users can temporarily disable the eject option by modifying the Windows registry, although this process may not be user-friendly and could lead to further complications [10].