Production AI
Facing the AI Wave: Turning “Repeated Defeat” into “Certain Victory” > Summary: > Through three rounds of internal AI capability upgrades, this article reveals the pitfalls of formalism and proposes a “competition-as-training” model to help enterprises truly transform their thinking, moving from traditional engineering logic to AI-first workflows. --- Background:
AI news
In the Fast-Changing Automotive Era, What Is Li Auto’s Real Moat? In a world where technology evolves daily and new models debut monthly, every automaker — and every investor — is asking the same core question: What is the true moat of a car company? Li Auto is seen as one
AI news
The Paradox of "More vs. Less" in Tech For decades, Moore’s Law shaped the tech industry’s mindset — more power, more features, bigger everything. Exponential growth equaled progress. Chips grew denser, screens expanded, and features multiplied. Yet history shows that the most memorable products often took the
AI news
Introduction: When AI Answers Too Much When the AI’s blinking cursor stops and leaves a long stretch of gray text, I often feel uncertain about where to begin reading — not relieved for having solved a problem. If you’re also a heavy AI user, this feeling may be familiar.
Production AI
Bridging Software Engineering and Data: A Practical Guide Introduction Matthias Niehoff: I'm Matthias, working at Codecentric — a consultancy based in Germany. My career began in software engineering, later moving into data-driven projects with Apache Spark and analytics platforms. My foundation is still in engineering — but applied to data
AI news
# Spring Tools: Bringing Logical Structure and Stereotypes to the Forefront When working on **Spring projects**, developers typically think about **high-level concepts** — such as **services**, **repositories**, **configuration classes**, **entities**, and **aggregate roots** — rather than just low-level constructs like classes and interfaces. **Spring Tools** enhances this focus by analyzing your source code
AI news
U.S. Power Inherits China’s “Legacy” > History often writes the answers long before the right person appears to ask the question. --- A $3.5 Billion Deal That Shook Wall Street In June 2024, U.S. Bitcoin mining company Core Scientific announced a $3.5 billion agreement with
AI news
Summary of Key Tech News * Zig migrates its main repository from GitHub to Codeberg, making the old repo read-only and encouraging sponsors to transfer support to Every.org. * Bring Back Doors campaign urges hotels to restore closable, opaque bathroom doors, promoting guest privacy via a database and public reporting. * Penpot
AI news
NeurIPS 2025 Best Paper Awards Announced The AI flagship conference NeurIPS 2025 has revealed its Best Paper recipients — four exceptional works recognized for their rigor, insight, and impact on advancing AI research in both societal and industrial domains. --- Personal Highlight Among the winners, my standout favorite is Liwei Jiang’
AI news
Podcast Overview In this episode, Shane Hastie, Lead Editor for Culture & Methods, interviews Joe Cassavaugh (Puzzles by Joe) about: * Building a sustainable solo game development business. * Why authenticity can outperform conventional market wisdom. * Learning from both successes and failures in the digital content market. --- 🎯 Key Takeaways * Sustainability in
Production AI
# AI-Powered Group Chat Understanding for Community Operations ## 1. Why We Need AI for Group Chat Analysis At **Bilibili**, our operations team manages numerous creator group chats — including category support groups, growth bootcamps, specialized forums, and Q&A channels. These groups generate **massive amounts of daily messages**. Manual tracking is:
AI news
作者 | 周国睿 编辑｜李忠良、蔡芳芳 策划｜AICon 全球人工智能开发与应用大会 传统级联式推荐将召回、粗排、精排与重排割裂，反馈难以闭环，模块间不一致持续累积，令算力利用效率（MFU：总消耗算力 /OPEX 可购最大算力）长期偏低，成为技术与业务演进的共同瓶颈。快手以生成式端到端架构 OneRec 重构推荐链路，在核心场景实现“更大模型、却更低成本”的跃迁：服务成本降至原系统的 1/10，并重新定义推荐系统的智能边界与效率标尺。快手科技副总裁、基础大模型及推荐模型负责人周国睿在 AICon 全球人工智能开发与应用大会上，系统解析了推荐系统的范式革新、 OneRecV2的scaling定制优化，以及OneRec-Think的“生成—理解”统一进展，为 AI 原生时代的推荐提供可复制的方法论。 以下是演讲实录（经 InfoQ 进行不改变原意的编辑整理）。 传统推荐架构的规模化瓶颈与范式局限 传统推荐架构正成为技术与业务演进的瓶颈。