有时候我会直接拒绝别人提交的合并请求（MR），甚至不做代码审查以及过多的解释，就是因为这份 代码是用 AI 写的，而且写得很糟糕，反而给团队添了麻烦。在工作的时候，我经常遇到这几种情况： 如果你看到我拒绝了你的 AI 代码，而且没有额外说明，只是把你丢到这个页面来，那就是因为它踩中 了上面的坑。 虽然现在有不少研究和讨论都承认 AI 在写代码时能帮上忙，但滥用 AI 也是个新问题。我们需要一些 规则来识别这种情况。 一些词解释 Merge Request（MR，合并请求）：就像一个人写好了一部分东西，然后把它交给团队，问"能不能把 这部分合进我们的项目里？" Code Review（CR，代码审查）：另一位成员检查这个请求，给出意见，决定要不要采纳。 为什么我来写这个 在 AI 编程工具越来越普及的当下，加拿大政府的一位高级计算机科学家兼教育工作者写下了一篇题为 《为什么我拒绝你的 AI 生成的 MR》的文章。他并非全盘否定 AI，而是希望厘清边界：哪些情况下 AI 代码可以被接受，哪些情况下则必须坚决拒绝。在文中，他不仅分享了自己作为团队负责人的真实困境 ——如何面对新人对 AI 的依赖和滥用，也给出了他在 ...

Group 1 - The article questions the validity of the claim that AI can lead to a tenfold increase in programming efficiency, suggesting that such assertions may be exaggerated [1][10][24] - It highlights the author's personal experience of anxiety regarding the rapid advancement of AI and its implications for software engineers [1][3][30] - The author critiques the performance of AI programming tools, stating that while they can generate template code, they often struggle with understanding larger codebases and can produce insecure code [4][5][15] Group 2 - The article argues that the notion of a "10x engineer" is often misunderstood, emphasizing that true productivity gains come from preventing unnecessary work rather than simply writing code faster [19][20][23] - It discusses the limitations of AI in software development, noting that while AI can assist in certain tasks, it does not fundamentally change the human processes involved in software engineering [12][18][24] - The author warns against the pressure to adopt AI tools hastily, advocating for a balanced approach that prioritizes quality and enjoyment in coding over mere speed [31][32][33]

Core Insights - The launch of DeepSeek-V3.1 marks a significant step towards the era of AI agents, with developers now able to build their own intelligent agents [1] - Alibaba's introduction of the Qoder programming platform highlights the competitive landscape in AI programming, with major players like ByteDance and Tencent also entering the market [2] - The AI programming sector is rapidly growing, with at least seven unicorns valued over $1 billion and total funding exceeding 240 billion RMB [2][3] Group 1: Product Developments - DeepSeek-V3.1 achieved a score of 76.3% in Aider coding tests, outperforming competitors like Claude 4 Opus and Gemini 2.5 Pro [1] - Qoder integrates top programming models and can search through 100,000 code files at once, significantly enhancing software development efficiency [1] - Anysphere's Cursor has gained approximately 30,000 enterprise clients and reached an annual recurring revenue (ARR) of over $500 million, showcasing its rapid growth in the AI programming space [3] Group 2: Market Dynamics - The AI programming race has intensified, with major tech companies vying for control over the ecosystem rather than just competing on product features [2] - The potential market for personalized software development could reach up to $15 billion by 2030, driven by reduced costs and barriers to entry in software development [6] - The rise of open-source strategies among domestic companies, such as Qwen3-Coder and DeepSeek-V3.1, is attracting global developers and fostering ecosystem growth [5][6] Group 3: Competitive Landscape - The AI programming sector is characterized by a unique advantage for domestic tech firms, which includes performance catch-up and ecosystem collaboration [4] - The market share of domestic models like Tongyi Qianwen has increased from 5% to 22% in the AI programming field within a month [6] - The competition is not only about faster coding but also about establishing a stronghold in the next wave of AI and computational power [5]

Core Viewpoint - Vibe coding, popularized by Andrej Karpathy, has gained traction in the tech industry, particularly among FAANG companies, although its definition and implementation remain contentious [1][5]. Group 1: Vibe Coding Popularity - A Reddit post suggests that vibe coding may be more prevalent than expected, with many employees at FAANG companies engaging in this practice [1][5]. - The post's author, an AI software engineer with over 15 years of experience, highlights the integration of AI in coding processes [3][4]. Group 2: Coding Process and Methodology - The coding process begins with reliable design documents and architecture, followed by writing tests before development [4][6]. - Key steps in the process include design reviews, task planning, software development using Test Driven Development (TDD), code review, and pre-release testing [6][13]. - Despite the involvement of AI, the process still requires significant human input, leading to debates about whether it truly qualifies as vibe coding [9][11]. Group 3: Perspectives on the Process - Some developers see value in the structured approach, advocating for detailed technical specifications and pre-development reviews [14][15]. - Others argue that the complexity of the process can hinder development speed, which may benefit independent founders [13][14].

Core Insights - The article discusses the evolving landscape of AI-assisted programming, emphasizing the shift in value from writing code to deleting it in a low-cost code generation environment [1][19] - Andrej Karpathy shares his experiences and methods for maximizing AI's assistance in programming, highlighting the importance of integrating multiple tools for different tasks [2][3] Tool Usage Philosophy - The philosophy of using tools is centered around the idea that tools should serve people, advocating for a combination of various workflows rather than relying on a single "perfect" tool [3] - Different tools excel at different levels of tasks, with tools like Claude Code and Codex being suitable for larger, less complex tasks, while Tab completion requires initial human input [3] Cursor (Tab Auto-Completion) - Karpathy indicates that Tab auto-completion is the primary method used in daily work, accounting for approximately 75% of his coding activities [4] - Writing code blocks or comments in the correct position can effectively communicate task specifications to AI [6] Auxiliary Tools (Claude Code / Codex) - Karpathy notes that while tools like Claude Code and Codex can generate code, they often lack "taste" in coding style, producing overly defensive or complex code [9] - These tools are particularly useful for tasks in unfamiliar areas, such as Rust or SQL, where they can generate extensive code quickly for debugging purposes [9][14] - The concept of the "post-scarcity era" of code is introduced, where the ability to create and discard vast amounts of customized code diminishes the perceived value of code itself [9][19] GPT-5 Pro: The Final Line of Defense - GPT-5 Pro is described as the ultimate tool for addressing the most challenging bugs that other tools cannot resolve, demonstrating its capability to identify subtle issues [14] - It can also assist in optimizing code abstraction and providing high-quality resources for specific topics [14] Characteristics of the Post-Scarcity Era of Code - The programming field is seen as being radically transformed by various paradigms and tools, leading to a sense of urgency to keep up with technological advancements [15] - The article highlights the potential for exploratory and experimental programming due to the lowered barriers to writing code [19]

Core Viewpoint - The article discusses Coinbase's controversial decision to fire engineers who refused to adopt AI programming tools, emphasizing the company's stance that AI is essential for their operations [5][11]. Group 1: AI Adoption in Programming - The use of AI in programming has become standard among developers, with Google claiming that 50% of its code is AI-generated [2]. - There is a growing community of developers who rely entirely on AI for coding, known as Vibe Coders, while some programmers still prefer traditional coding methods [4]. Group 2: Coinbase's Decision - Coinbase CEO Brian Armstrong announced the firing of engineers who did not use AI programming tools, stating that the company had purchased enterprise licenses for GitHub Copilot and Cursor [6]. - Armstrong expressed shock at the slow adoption rate of AI among engineers and implemented a mandatory trial period for AI tools, leading to the dismissal of those who did not comply [8][10]. Group 3: Reactions and Implications - The decision sparked significant discussion online, with mixed reactions from the tech community, including claims that the prevalence of AI programming is overestimated [13][14]. - Armstrong acknowledged that his approach was high-pressure and not well-received by some employees, but he aimed to convey that using AI is not optional [11].

Core Insights - Alibaba has launched a new AI programming platform called Qoder, which integrates top programming models and enhances software development efficiency significantly [1][2] - Qoder can reduce the time required to develop a full-stack e-commerce website from several days to just ten minutes [1] - The platform addresses challenges in real software development, such as high complexity and uncertainty, by upgrading its contextual engineering capabilities [1] Features and Capabilities - Qoder includes a built-in code search engine capable of retrieving 100,000 code files, improving recall rates by 12% compared to industry benchmarks [2] - The platform supports Repo Wiki to make implicit knowledge explicit, aiding both developers and AI in understanding code projects [1][2] - Qoder features a long-term and short-term memory system that summarizes project experiences and user preferences, allowing for self-learning and evolution [1] User Experience - Qoder offers different modes, including Ask Mode, Agent Mode, and a new Quest Mode, which allows the AI to act as a full-stack engineer [2] - In Quest Mode, developers can delegate tasks to the AI, significantly increasing development efficiency by over 10 times [2] - Qoder is currently available for both Mac and Windows systems, with users able to download and experience it for free from the official website [3]

Core Insights - The AI programming industry is facing significant losses due to high costs and low profit margins, with many companies relying on subscription models that do not adequately cover their expenses [1][3][4] - Despite rapid revenue growth in some companies, the underlying business models are often unsustainable, leading to concerns about long-term viability [2][4][10] Group 1: Financial Performance - Cursor achieved $100 million in annual recurring revenue (ARR) in just 21 months, with a current ARR of $500 million and revenue per employee at $3.2 million [2] - Replit grew from $10 million to $100 million ARR in only 6 months, while Lovable reached $100 million ARR in 8 months, with a projected ARR of $250 million by year-end [2] - Many AI programming companies exhibit high growth rates but have low or negative gross margins, indicating that growth is often at the expense of profitability [4][12] Group 2: Cost Structure and Pricing Challenges - AI programming companies face a mismatch between fixed subscription fees and variable costs associated with high usage, leading to significant financial strain [3][6][12] - Users can exploit subscription models to incur costs far exceeding their subscription fees, creating a situation where companies are effectively subsidizing heavy users [3][11] - Attempts to raise prices have met with backlash from users, highlighting the fragile customer retention rates in the industry [7][8] Group 3: Market Dynamics and Competition - The competitive landscape is intensifying, with traditional software companies entering the AI space, further complicating the market for AI programming firms [8][9] - High customer churn rates, estimated between 20% to 40%, pose a significant challenge for AI programming companies, making it difficult to maintain a stable revenue base [8][10] Group 4: Business Model Viability - The concept of Business Model and Product Fit (BMPF) is critical for the sustainability of AI programming companies, as many are currently operating under flawed business models [10][12] - Companies that fail to establish a clear path to profitability may find themselves in a "scale trap," where growth does not translate into financial health [12][13] - The reliance on subsidies to attract users is not a viable long-term strategy, as it masks underlying issues with profitability and market demand [12][13]

Group 1: Course Overview - The course "AI Programming Assistant Cursor: Building a Mini Program Similar to Xiaohongshu from Scratch" provides a systematic learning platform for developers to master core skills in mini program development [2] - The course emphasizes a structured learning path that includes understanding the basic architecture of mini program development, effective application of AI tools, implementation of core functionalities, full-stack development capabilities, and cross-platform thinking [6] Group 2: Mini Program Development Framework - The mini program development framework consists of three main components: view layer (WXML and WXSS), logic layer (JavaScript), and configuration layer (JSON) [3] - The course utilizes a case-based teaching method with over 70 teaching cases, significantly enhancing learning efficiency [3] Group 3: AI Programming Assistant Cursor - The course highlights the integration of AI programming assistant Cursor, which represents a new direction in programming [4] - Key skills include natural language description of requirements, understanding and modifying generated code, and debugging and optimization using Cursor [5] Group 4: Core Functionalities of Xiaohongshu - The core functionalities of the Xiaohongshu mini program include content display, social interaction, and personal center modules [5][7] - Important implementation aspects include content waterfall flow display, user interaction design, and multimedia processing [7] Group 5: Full-Stack Development Skills - The course aims to cultivate full-stack development capabilities, expanding from front-end development to full-stack thinking [5] - Data shows that through 32 hours of systematic learning, developers can comprehensively master skills from basics to cloud development [5] Group 6: Cross-Platform and Commercialization - The course encourages learners to develop cross-platform thinking, understanding how to extend core business logic to other platforms [6] - Techniques for traffic conversion and understanding the Xiaohongshu open platform's entry process are also covered [7]

Group 1: Software ETF Performance - The software ETF has a turnover rate of 9.25% during trading, with a transaction volume of 491 million yuan [3] - As of August 15, the software ETF has seen an average daily transaction volume of 488 million yuan over the past week, ranking first among comparable funds [3] - The software ETF has experienced a net inflow of 709 million yuan over the last 21 trading days, with inflows on 12 of those days [3] Group 2: Growth and Returns - The software ETF's net asset value has increased by 10.39% over the past three years [3] - The highest monthly return since inception is 39.35%, with the longest consecutive monthly gain being three months and a maximum cumulative increase of 69.40% [3] - The average return during the months of increase is 9.75% [3] Group 3: AI Programming and Market Potential - AI programming is identified as one of the fastest-growing and most valuable applications in the AI sector, addressing the imbalance between unlimited software demand and limited developer supply [4] - The potential market size for AI programming is projected to reach 15 billion USD by 2030, serving as foundational infrastructure for AI agents [4] Group 4: Key Stocks in Software Service Index - The top ten weighted stocks in the CSI Software Service Index include iFlytek, Kingsoft Office, Tonghuashun, and others, collectively accounting for 61.39% of the index [4] - Notable stock performances include Tonghuashun with a 15.74% increase and Kingsoft Office with a 5.57% increase [6]