Group 1 - DeepSeek's new model, DeepSeek R2, is expected to launch in early May and will reduce costs by 97% compared to GPT-4, utilizing Ascend cards for training [1] - DeepSeek R2 will feature a hybrid expert model (MoE) with a total parameter count of 1.2 trillion, doubling the parameters of DeepSeek R1, which had 671 billion [1] - The model aims to achieve breakthroughs in programming capabilities, multilingual reasoning, and higher accuracy at lower costs [1] Group 2 - Jin Modern is actively expanding its standardized, general-purpose software product business centered around an "AI low-code" development platform, having developed several standardized platform software products [2] - Haoyun Technology continues to invest in low-code technology research and development, with its low-code platform "Haoyida" deeply integrated with AI and IoT to customize AI agents for enterprises [2]

Core Viewpoint - Alfa Bank and Skolkovo Institute of Science and Technology are launching the SMILES-2025 Machine Learning Summer School at Harbin Institute of Technology in China, aimed at enhancing machine learning education and collaboration between Russia and China [1] Group 1 - The Machine Learning Summer School will take place from July 14 to July 27 in China [1] - A total of 100 Russian undergraduate students, graduate students, and young scholars will be funded by Alfa Bank to attend the summer school, with an additional 300 participants joining online [1]

Group 1 - The report by Airwallex and Statista highlights the rapid growth of the Hong Kong e-commerce market, with an expected industry valuation increase of 52% from $4.77 billion in 2024 to $7.25 billion by 2029 [14][16][20] - 92% of consumers in Hong Kong shop online at least once a month, and 95% prefer purchasing overseas products, indicating a strong demand for cross-border shopping [16][20] - Payment preferences are diverse, with credit cards being the most common payment method at 82%, followed by electronic wallets at 64% and Buy Now Pay Later (BNPL) services at 71%, particularly favored by luxury consumers [17][18][20] Group 2 - Key strategies for market expansion include understanding local tax structures, providing localized checkout experiences, and supporting multiple payment methods such as Visa and Alipay [2][25][31] - Airwallex offers a one-stop service that aids businesses in compliance, providing global accounts, security measures, and financial tools to facilitate local currency transactions [3][28][29] - The report emphasizes the importance of a seamless checkout experience, with 85% of consumers wanting prices displayed in Hong Kong dollars and over 90% expecting preferred payment options and clear tax details [23][31][32] Group 3 - The report outlines the regulatory environment in Hong Kong, noting a simple tax structure with a corporate tax rate of 16.5%, and no Goods and Services Tax (GST) or Value Added Tax (VAT), making it attractive for international sellers [25][26] - Companies must register within a month of starting operations and comply with local product safety and labeling standards, particularly in sectors like electronics and cosmetics [25][26] - Data privacy regulations require businesses to obtain consent before collecting personal information and to implement security measures to prevent misuse [51][56]

Core Insights - Despite tariffs imposed by multiple countries on Chinese automotive brands, the overall impact remains limited, with a 24% increase in export costs amounting to $46 billion, which only represents 3.8% of China's total automotive industry output [1] Group 1: Tariff Impact - The U.S. tariffs on imported vehicles and specific auto parts, effective from April 3, 2023, have caused panic within the global automotive supply chain [1] - The impact of U.S. tariffs on China's complete vehicle exports is minimal, while the effect on auto parts exports is more significant, potentially affecting $20 billion to $30 billion in exports [2] - In 2024, China exported only 107,000 complete vehicles to the U.S., accounting for 1.81% of its total automotive exports [2] Group 2: Market Dynamics - China's automotive exports have become a crucial growth engine, with exports surpassing 6.4 million units in 2024, maintaining its position as the world's largest automotive exporter [1] - The demand for Chinese vehicles in Russia has doubled over the past five years, helping to mitigate the impact of global tariff fluctuations [3] - In 2024, Russia and the Middle East accounted for 35% of China's total automotive exports, surpassing exports to Europe and North America [3] Group 3: Technological Advancements - China leads in the adoption of intelligent driving features, with nearly 60% of passenger cars sold in 2024 equipped with L2 or higher assistance systems, compared to less than 40% in the U.S. [5] - The unique advantages of Chinese automotive products in electric and intelligent driving technologies have established a competitive edge in overseas markets [4][5] - Two-thirds of surveyed executives believe that China is at the forefront of intelligent driving systems, with other markets lacking the necessary conditions to replicate this success [5] Group 4: Future Projections - China's automotive exports are projected to grow by 23% in 2024, with passenger car exports reaching 6.4 million units, significantly outpacing Japan [6] - The growth rate is expected to slow to 4% in 2025, while the domestic market is also projected to grow by 4%, reaching 26.8 million vehicles [6] - By 2030, Chinese brands are anticipated to capture approximately 30% of the global automotive market, up from 21% in 2024 [6]

Core Insights - A unique machine learning "periodic table" has been created by a team from the Massachusetts Institute of Technology, showcasing the relationships between over 20 classic machine learning algorithms [1] - This framework reveals strategies for scientists to integrate different methods, which can enhance existing AI models or propose entirely new models, further promoting the development and application of artificial intelligence technology [1]

Financial Data and Key Metrics Changes - Revenue was relatively flat year over year, while EBITDA growth accelerated to 4.8%, driven by strong mobile growth and improved service quality [6][40] - Adjusted EBITDA grew by 4.8% year over year, with net income attributable to shareholders at $1.2 billion, compared to $1.1 billion last year [40][46] - First quarter free cash flow totaled $1.6 billion, an increase of approximately $1.2 billion compared to last year's first quarter [44] Business Line Data and Key Metrics Changes - The company added over 500,000 Spectrum Mobile lines, with a total of over 2.1 million lines added over the last year, resulting in line growth of over 25% [5] - Internet customer results showed a decline of 60,000 customers in the first quarter, while video customers declined by 181,000, an improvement from a loss of 405,000 in the same quarter last year [28][29] - Mobile revenue growth was driven by a 3.9% increase in mid-market and large business revenue, while small business revenue declined by 0.2% [34] Market Data and Key Metrics Changes - Monthly data usage by non-video Internet customers grew to approximately 825 gigabytes per month, with over 30% of those customers using over one terabyte of data [8] - The company ended the quarter with 902,000 subsidized rural pass lines, growing those passings by 89,000 in the first quarter [31] - Advertising revenue declined by 12.9% primarily due to less political revenue, with total consolidated first quarter revenue up 0.4% year over year [35][36] Company Strategy and Development Direction - The company continues to focus on delivering the best networks and products at the best value, with a unique set of assets and significant scale [11][12] - The launch of the "Life Unlimited" brand and new customer commitment aims to enhance reliability and service quality, driving higher customer satisfaction [21][24] - The company is investing in machine learning and AI to improve service efficiency and customer experience [18][20] Management's Comments on Operating Environment and Future Outlook - Management noted that the operating environment remains competitive, but the impact of the Affordable Connectivity Program (ACP) elimination is behind them [6] - The company expects to gradually increase leverage to the middle of the four to four and a half times range pro forma for the Liberty transaction over the next several quarters [49] - Management expressed confidence in returning to positive broadband subscriber growth despite current market conditions [126] Other Important Information - The company incurred approximately 9,000 disconnects related to the Los Angeles wildfires, but first quarter adjusted EBITDA was not significantly impacted [25][26] - Capital expenditures totaled $2.4 billion in the first quarter, down about $400 million from last year's first quarter [41] - The company added two Liberty-nominated members to its board of directors [24] Q&A Session Summary Question: Differences in converged households and impact on broadband numbers - Management noted that there is a substantial difference in Internet churn rates for customers who also take mobile lines, with significant benefits observed [55][56] Question: Impact of tariffs on capital spending - Management does not expect tariffs to have a meaningful impact on capital expenditures, reiterating guidance for the year at $12 billion [63] Question: Update on Seamless Entertainment rollout - Management provided an update on the rollout of direct-to-consumer apps and the digital storefront, with significant progress made [70][74] Question: Consumer behavior and mobile substitution - Management indicated that sales are up and churn is stable, despite some mobile substitution trends [133] Question: Fiber competition and broadband penetration - Management stated that fiber overbuilders have historically impacted penetration rates, but the current market dynamics are more influenced by mobile substitution [150]

新加坡城市景观。 新华社发 就读于电子信息科技大学的刘胜然对新加坡科技设计大学很感兴趣。他说："这所学校相对年轻，它由 新加坡政府与美国麻省理工学院合作设立，注重跨学科融合，是一所专注设计与创新研究、工程领域的 新概念大学。" 刘胜然坦言，申请新加坡排名靠前的公立大学硕士项目竞争压力不小。"据我所知，申请新加坡国立大 学和南洋理工大学的硕士，要有不错的本科GPA成绩，且雅思成绩不低于6.5或托福100分，同时要求具 备科研或相关职场经验。"刘胜然说。 《2025新加坡留学报告》中一组数据显示，在2024入学季，新加坡公立大学硕士录取的学生中，八成以 上的申请者GPA在80分以上，雅思成绩达到6.5分以上，近七成的录取学生GMAT成绩在650分以上。相 关留学专家建议，背景条件较薄弱的学生，不妨选择新加坡私立大学就读语言课程或预科类课程，逐步 完善自身条件和能力。 此外，高中生可凭高考成绩直接申请新加坡公立大学本科。启德教育相关负责人介绍："对于新加坡排 名前三的公立大学（新加坡国立大学、南洋理工大学、新加坡管理大学），建议申请者高考成绩高于当 地一本线100分以上，并具备托福93分或雅思6.5分（阅读和写作单 ...

如果您希望可以时常见面，欢迎标星收藏哦~ 来源：内容 编译 Business Insider ，谢谢。 英伟达（Nvidia）不需要经历剧烈的文化转型来让员工"卷"起来——这种氛围早已根植于企业文化 之中。 相比之下，像 Shopify、微软（Microsoft）和 Meta 等公司正在加大对员工的要求，力求在AI赛 道上抢占先机、提升效率。科技公司的这一转变伴随着对"低绩效员工"的裁撤、强硬的返岗令，以 及员工福利的缩减。 尽管英伟达在过去几年迅速扩张了员工规模，市值也像坐上了过山车，但其企业文化的基石早在 AI浪潮之前就已奠定。过去四个月出版的两本关于英伟达及其CEO黄仁勋（Jensen Huang）的 书，也印证了多位前员工对《Business Insider》的说法。 英伟达拥有一种自上而下的高压工作文化，而这种文化源于那位广为人知的CEO，与那些靠行政 命令推动"高强度文化"的科技公司形成鲜明对比。 《The Thinking Machine: Jensen Huang, Nvidia, and the World's Most Coveted Microchip》一 书的作者 Stephen Wi ...

研究显示，亚太地区的医院经历了从稳步扩张、大流行冲击、上坡恢复到持续增长等阶段。在每个阶段， 我们都分析了医院收入、手术数量和病床占用率与 GDP 同比增长率的关系。 在新冠之前，该地区的医院通过收购和提供更多服务不断扩张。这使得医院收入稳定增长5%至10%，手 术数量增长 10%至 15%，病床占用率稳定在65%至70%，与该地区 4%至 5%的 GDP 增长率保持一致。 尽管医院规模稳步扩大，但大流行病暴露了医院系统的严重弱点，导致收入减少，病床占用率因可用病床 短缺而上升。手术也被推迟或取消，因为病人为了尽量减少接触病毒的风险而回避医院。 后新冠时代，虽然没有了集中性的疾病冲击，医院正面临着一系列趋势和挑战。为了解整个亚太地区目前 的医院经营环境，科尔尼公司在 2022 年至 2024年期间调研了25位医院高管和150多位患者。 与此同时，运营更加灵活的医院能够迅速做出调整。例如，新加坡陈笃生医院（Tan Tock Seng Hospital） 将膝关节置换手术转变为采用增强型术后恢复（ERAS）模式的非卧床手术，将住院时间从六天缩短为一 天。 随着疫苗接种率超过人口的40%，选择性手术出现反弹，到 2 ...

Core Insights - The article discusses the upcoming webinar focused on summer weather outlooks in the Americas and Asia, and their potential impacts on crops, including updates on U.S. spring planting activities and South American late-season progress [1]. Agenda Overview - The agenda includes a review of the South American season, early weather outlook for U.S. spring planting, U.S. planting and production outlook, ENSO (El Niño-Southern Oscillation) and its global impacts, and summer weather in Asia [5]. Event Information - The event is scheduled for April 16, 2025, at 09:00 AM Beijing time, and will be conducted in English [3]. Expert Speakers - The webinar features senior analysts from LSEG, including Isaac Hankes, a senior climate research analyst, and Dong Soon Choi, a senior agricultural analyst [4]. LSEG's Commodity Trading Solutions - LSEG offers market insights, data management solutions, and seamless trading execution capabilities to provide a competitive advantage in commodity trading [8]. - The company emphasizes the importance of obtaining accurate information at the right time and format for successful commodity trading [10]. - LSEG provides tools, fundamentals, forecasts, alternative data, and the latest news to help clients excel in the competitive landscape of commodity trading [11]. Specific Commodity Trading Areas - **Energy Trading**: LSEG covers a wide range of energy assets, including oil, gas, electricity, coal, and carbon, supporting global energy commodity trading platforms [13]. - **Metal Trading**: The company utilizes machine learning and AI to enhance data and analysis for global metal trading, helping clients discover trading opportunities [15]. - **Agricultural Trade**: LSEG employs robust fundamental data and alternative sources, such as weather tracking and satellite imagery, to predict market price trends in agriculture [17]. - **Shipping Trade**: The company provides unique insights into global shipping trade through real-time reporting of vessel movements and port congestion [19]. - **Data Aggregation and Digitalization**: LSEG specializes in standardizing and structuring multiple data sources to generate actionable insights for global trading companies [20].