Core Insights - The current data industry in China is entering a "fast lane" of development, with the value of data as a key production factor becoming increasingly prominent [1][2] - High-quality datasets are essential for the reliable development of AI models, as low-quality data can lead to misleading outputs known as "hallucinations" [2][3] Data Quality and AI Models - The training data for large language models (LLMs) often comes from the internet, resulting in varying quality and leading to outputs based on "probabilistic matching" rather than "factual judgment" [2] - A study indicates that when training datasets contain only 0.01% false text, harmful content output by the model increases by 11.2%, highlighting the critical issue of insufficient high-quality data supply [2] - High-quality datasets are categorized into general datasets, industry general datasets, and industry-specific datasets, which are foundational for the application of both general and industry models [2][3] Industry-Specific Data - Industry general datasets include knowledge that requires a certain level of professional background to understand, such as healthcare data encompassing personal attributes, health status, and medical application data [3] - Industry-specific datasets require deeper professional knowledge and are crucial for specific business scenarios, such as medical AI relying on high-quality expert-annotated data [3] AI and Data Integration - The trend is shifting towards a data-centric approach in AI development, which does not diminish the value of model-centric AI but rather complements it [3] Prompt Engineering - The ability to ask questions and discern answers is emphasized as crucial in the AI era, with the concept of prompt engineering introduced to guide LLMs in generating useful content [4] - Skilled prompt engineers can enhance AI model efficiency by over 30% in fields like healthcare by designing precise prompts [4] Policy and Industry Development - The Chinese government has issued guidelines to strengthen the construction of high-quality AI datasets, emphasizing application-oriented approaches and the development of data processing and service industries [5] - The shift from "data-entity integration" to "entity-data integration" reflects a focus on promoting high-quality development driven by the needs of the real economy [5]

近日， 未可知人工智能研究院高级授课专家吴小楠老师受浙江省科学技术协会邀请 ，在 2025年浙江省科协系统信息员和新媒体工作人员培训班上 作题为 《DeepSeek提示词技巧与新闻宣传写作》 的专题培训 。全省120余名科协系统宣传骨干参与学习。 在培训中，吴小楠老师立足智媒时代传播特性， 系统解析了AI辅助写作的核心方法论 。课程通过提示词工程优化、科学叙事逻辑重构、多场景宣传文案 生成三大模块，演示了如何借助DeepSeek等智能工具提升新闻稿件的传播效能。现场学员通过实时操作，掌握了精准控制AI输出风格、快速生成适配内容 的实操技能。 合作联系微信： duyuaigc — END — 【关于未可知人工智能研究院】 未可知人工智能研究院聚焦 AI前沿趋势、商业落地与人才发展，致力于成为"AI时代的认知基础设施"。 更多精彩内容，欢迎关注官方微信公众号：未可知 人工智能研究院 。 合作伙伴 TATE GRID 第二三十六号 深圳職業技術大學 SHENZHEN POLYTECHNIC UNIVERSITY 作为人工智能领域的领先机构， 未可知人工智能研究院始终践行AI科普传播先锋的职责 ，开发DeepSeek ...

Core Insights - GPT-5 has received mixed reviews since its launch on August 8, with users expressing dissatisfaction despite its technical advancements [1][5][7] - The official stance from OpenAI is that the issues stem from users not adapting to the new interaction model required by GPT-5, which has evolved into a more autonomous "digital mind" [9][78] - The release of a prompt guide by OpenAI aims to help users better engage with GPT-5, emphasizing the importance of updated communication methods [8][9] Group 1: Performance and Capabilities - GPT-5 demonstrates significant improvements in areas such as mathematics, coding, and multi-modal understanding, showcasing its capabilities as a "full-stack engineer" [4][13] - Despite its higher IQ, GPT-5 exhibits instability, sometimes making errors on simple tasks and lacking emotional intelligence, which has led to concerns about its practical usability [5][6][10] - OpenAI has reported a performance increase in the Tau-Bench test, with scores rising from 73.9% to 78.2%, indicating better efficiency and lower costs [23][24] Group 2: User Interaction and Guidelines - The prompt guide outlines four key areas of evolution for GPT-5: agentic task performance, coding ability, raw intelligence, and steerability, which are crucial for effective user interaction [10][15][17] - Users are encouraged to adjust parameters like reasoning effort and verbosity to optimize GPT-5's performance based on task complexity [53][70] - The guide suggests methods for users to either constrain or empower GPT-5's capabilities, depending on the task requirements, highlighting the need for a more nuanced approach to AI interaction [29][32][36] Group 3: Challenges and Solutions - The dual-edged nature of GPT-5's capabilities means that improper use can lead to inefficiencies, necessitating users to become adept "trainers" of the AI [26][27] - OpenAI emphasizes the importance of clear and structured prompts to avoid conflicts that could lead to performance degradation [54][56] - The guide provides practical solutions for common user challenges, such as managing verbosity and reasoning depth, to enhance the overall interaction experience [50][52][68]

Core Insights - Alibaba has launched Qwen-Image, an image generation foundational model designed to tackle complex text rendering and precise image editing challenges through systematic data engineering and advanced training paradigms [1][4] - The model aims to enhance the understanding and alignment capabilities of complex, multi-dimensional text instructions in image generation tasks, addressing long-standing challenges in the AI field [3][5] Data Processing and Model Architecture - Qwen-Image employs a comprehensive data processing system that collects billions of high-quality text-image pairs, emphasizing quality over quantity, and utilizes a seven-stage filtering pipeline to enhance data quality and alignment [5][6] - The model features a dual encoding design, utilizing high-level semantic features and low-level reconstruction features to balance semantic coherence and visual fidelity during image editing [6][5] Training and Performance - The training process is progressive, moving from low-resolution to high-resolution images, and incorporates reinforcement learning methods to optimize the quality of generated results and adherence to instructions [6][5] - Benchmark tests and human evaluations indicate that Qwen-Image achieves industry-leading performance in general image generation, complex text rendering, and directive image editing tasks [6] Comparison with Traditional Tools - Qwen-Image exhibits core editing capabilities similar to Photoshop but operates through natural language instructions rather than manual tools, allowing users to describe edits instead of executing them through traditional methods [25][26] - The model's ability to understand and execute complex instructions, such as adjusting poses while maintaining visual and semantic consistency, surpasses traditional tools that require manual adjustments [26][27] User Experience and Accessibility - Qwen-Image lowers the technical barrier for image editing by enabling users to express visual intentions through clear language, contrasting with Photoshop's requirement for mastery of complex tools and color theory [28][29] - While Qwen-Image is not a direct replacement for Photoshop, it represents a new paradigm in image content creation and editing, catering to different user needs and scenarios [29]

Core Viewpoint - The article discusses the perspective of Jad Tarifi, a founding member of Google's generative AI team, who advises against pursuing a PhD in AI due to the rapid evolution of the field, suggesting that by the time one graduates, the AI landscape may have drastically changed [1][8]. Group 1: AI Talent Market - Major tech companies like Meta are offering signing bonuses reaching hundreds of millions to attract AI talent [2]. - Tarifi's comments serve as a stark contrast to the ongoing talent war in the AI sector, highlighting the urgency and volatility of the field [3][4]. - The job market is being reshaped by AI, with over 1 million jobs in the U.S. announced for layoffs due to generative AI adoption in 2025 alone [14][15]. Group 2: Employment Impact - The technology sector has been particularly affected, with over 89,000 layoffs attributed directly to AI-driven redundancies since 2023 [16]. - Entry-level positions, especially in knowledge-intensive roles, are at risk as AI can perform tasks traditionally handled by junior employees [19]. - Nearly half of U.S. Gen Z job seekers feel that AI has devalued their degrees, reflecting a significant shift in the job market [21]. Group 3: Future Skills and Adaptation - Tarifi emphasizes the importance of developing social skills and empathy as essential competencies in the AI era [23]. - He suggests that while technical knowledge is valuable, understanding how to effectively use AI tools and having a good sense of taste in their application is crucial [24]. - The article also notes that individuals should focus on excelling in specific areas rather than trying to master every detail of AI technology [28].

Core Insights - The performance improvement of AI models is attributed equally to model upgrades and the optimization of user prompts, with 51% of the enhancement coming from the model and 49% from prompt optimization [2][28]. Group 1: Research Findings - A collaborative study by institutions such as the University of Maryland, MIT, and Stanford demonstrated that user prompts significantly influence AI performance, specifically in image generation tasks using DALL-E models [2][4]. - The concept of "prompt adaptation" was introduced, highlighting the importance of user input in maximizing the capabilities of AI models [3][12]. - The study involved 1,893 participants who generated images using DALL-E 2 and DALL-E 3, revealing that DALL-E 3 outperformed DALL-E 2 due to both model improvements and user prompt adjustments [4][21]. Group 2: Experimental Design - Participants were tasked with generating images based on specific target images, with their performance measured by the cosine similarity between generated and target images [14][15]. - The experiment aimed to separate the effects of model upgrades and prompt optimization on overall performance, using a replay analysis method to assess contributions from both factors [16][26]. - Results indicated that users of DALL-E 3 produced images with a cosine similarity average higher by 0.0164 compared to DALL-E 2 users, demonstrating the model's superior capabilities [22][25]. Group 3: User Behavior and Prompting Strategies - Users of DALL-E 3 tended to create longer and more descriptive prompts, indicating a shift in strategy as they adapted to the model's enhanced capabilities [25][30]. - The study found that the effectiveness of prompt optimization is contingent upon the model's ability to handle complex instructions, suggesting that user input must evolve alongside technological advancements [30][32]. - The research highlighted that lower-skilled users benefited more from model upgrades, while high-skilled users experienced diminishing returns, emphasizing the need for tailored prompting strategies [31][32].

Core Viewpoint - The article discusses the tension between the grand narrative of AI and the practical challenges faced by product managers in implementing AI solutions, highlighting the gap between theoretical concepts and real-world applications [1][2][9]. Group 1: AI Product Development Challenges - Product managers are overwhelmed by the rapid advancements in AI technologies, such as GPT-5 and Kimi K2, while struggling to deliver a successful AI-native product that meets user expectations [1][2]. - There is a significant divide between those discussing the ultimate forms of AGI and those working with unstable model APIs, seeking product-market fit (PMF) [2][3]. - The current AI wave is likened to a "gold rush," where not everyone will find success, and many may face challenges or be eliminated in the process [3]. Group 2: Upcoming Global Product Manager Conference - The Global Product Manager Conference scheduled for August 15-16 aims to address these challenges by bringing together industry leaders to share insights and experiences [2][4]. - Attendees will hear firsthand accounts from pioneers in the AI field, discussing the pitfalls and lessons learned in transforming AI concepts into viable products [5][6]. - The event will feature a live broadcast for those unable to attend in person, allowing broader participation and engagement with the discussions [2][11]. Group 3: Evolving Role of Product Managers - The skills traditionally relied upon by product managers, such as prototyping and documentation, are becoming less relevant due to the rapid evolution of AI technologies [9]. - Future product managers will need to adopt new roles, acting as strategists, directors, and psychologists to navigate the complexities of AI integration and user needs [9][10]. - The article emphasizes the importance of collaboration and networking in this uncertain "great maritime era" of AI development [12].

Core Viewpoint - The collaboration between two Tsinghua University alumni has successfully enhanced the Gemini 2.5 Pro model to achieve a gold medal level in the International Mathematical Olympiad (IMO) through a self-iterative verification process and prompt optimization [1][4][10]. Group 1: Model Performance and Methodology - Gemini 2.5 Pro achieved a 31.55% accuracy rate in solving IMO problems, significantly outperforming other models like O3 and Grok 4 [9]. - The research team utilized a structured six-step self-verification process to improve the model's performance, which includes generating initial solutions, self-improvement, and validating solutions [16][18]. - The model was able to generate complete and mathematically rigorous solutions for 5 out of 6 IMO problems, demonstrating the effectiveness of the structured iterative process [24][23]. Group 2: Importance of Prompt Design - The use of specific prompt designs significantly improved the model's ability to solve complex mathematical problems, highlighting the importance of prompt engineering in AI model performance [12][14]. - The research indicated that detailed prompts could reduce the computational search space and enhance efficiency without granting the model new capabilities [23]. Group 3: Research Team Background - The authors, Huang Yichen and Yang Lin, are both Tsinghua University alumni with extensive academic backgrounds in physics and computer science, contributing to the credibility of the research [26][28][33]. - Yang Lin is currently an associate professor at UCLA, focusing on reinforcement learning and generative AI, while Huang Yichen has a strong background in quantum physics and machine learning [30][35]. Group 4: Future Directions and Insights - The research team plans to enhance the model's capabilities through additional training data and fine-tuning, indicating a commitment to ongoing improvement [42]. - Yang Lin expressed the potential for AI to play a more significant role in mathematical research, especially in addressing long-standing unresolved problems [44].

Core Insights - The article discusses the evolution of prompt engineering in AI, emphasizing its importance in enhancing the interaction between users and AI models [4][16][65] - It highlights the differences in AI model performance based on the quality of prompts used, suggesting that effective prompt engineering can significantly improve AI outputs [3][16][65] Group 1: Evolution of Prompt Engineering - The evolution of prompts has progressed through three stages: "Magic Spell" era, "Enlightenment and Guidance" era, and "Systematic Engineering" era [10][11][14] - In the "Magic Spell" era, users treated AI like a search engine, leading to inconsistent results [10] - The "Enlightenment and Guidance" era introduced techniques like example learning and thinking chains, improving AI's reasoning and logic capabilities [12][13] - The current "Systematic Engineering" era requires structured prompts that include roles, objectives, constraints, examples, and steps to ensure stable and controllable AI outputs [14][15] Group 2: Importance of Prompt Engineering - Prompt engineering is defined as the science of designing and optimizing prompts to effectively communicate with large language models, directly impacting the quality of AI outputs [16] - High-quality prompts reduce the likelihood of AI generating "hallucinations" and help uncover the AI's potential for complex tasks [17] - The R.O.L.E.S. framework (Role, Objective, Limit & Constraint, Examples, Steps) is introduced as a method for creating effective prompts [17][18][20][22][26][28] Group 3: ByteDance's PromptPilot - ByteDance launched PromptPilot, a platform aimed at optimizing the entire process of AI model application, from concept to deployment and iteration [35] - The platform offers features for prompt generation and optimization, making it accessible for users without prior prompt writing experience [39] - Users can validate and refine prompts through various tuning modes, enhancing the effectiveness of AI-generated outputs [40][41][62] Group 4: Conclusion and Future Implications - The article concludes that mastering prompt engineering is essential for leveraging AI effectively, transforming it into a foundational skill for future interactions with AI [65][66] - While PromptPilot is not perfect, it serves as a valuable tool for users to develop structured thinking and improve their interactions with AI [67][70]

Core Viewpoint - The article discusses the challenges of creating valuable AI Native products, emphasizing that user experience has evolved from a design-centric issue to a technical one, where both user needs and value delivery are at risk of "loss of control" [3][4]. Group 1: User Experience Challenges - The transition from mobile internet to AI Native products has made it more difficult to deliver a valuable user experience, as it now involves complex technical considerations rather than just aesthetic design [3]. - The current bottleneck in AI Native product experience is fundamentally a technical issue, requiring advancements in both product engineering and model technology to reach a market breakthrough [4]. Group 2: Input and Output Dynamics - AI products are structured around the concept of Input > Output, where the AI acts as a "Magic Box" that needs to manage uncertainty effectively [6]. - The focus should be on enhancing the input side to provide better context and clarity, as many users struggle to articulate their needs clearly [7][8]. Group 3: Proposed Solutions - Two key approaches are highlighted: "Context Engineering" by Andrej Karpathy, which emphasizes optimizing the input context for AI, and "Spec-writing" by Sean Grove, which advocates for structured documentation to clarify user intentions [7][8]. - The article argues that the future of AI products should not rely on users becoming experts in context management but rather on AI developing the capability to autonomously understand and predict user intentions [11][12]. Group 4: The Role of AI - The article posits that AI must evolve to become a proactive partner that can interpret and respond to the chaotic nature of human communication and intent, rather than depending on users to provide clear instructions [11][12]. - The ultimate goal is to achieve a "wide input" system that captures high-resolution data from users' lives, creating a feedback loop between input and output for continuous improvement [11].