Core Viewpoints - The article presents four logical chains regarding the understanding of "predict the next token," which reflects different perceptions of the potential and essence of LLMs or AI [1] - Those who believe that predicting the next token is more than just probability distributions are more likely to recognize the significant potential of LLMs and AI [1] - A deeper consideration of AI and ideals can lead to an underestimation of the value of what ideals accomplish [1] - The ideal VLA essentially focuses on reinforcement learning dominating the continuous prediction of the next action token, similar to OpenAI's O1O3, with auxiliary driving being more suitable for reinforcement learning than chatbots [1] Summary by Sections Introduction - The article emphasizes the importance of Ilya's viewpoints, highlighting his significant contributions to the AI field over the past decade [2][3] - Ilya's background includes pivotal roles in major AI advancements, such as the development of AlexNet, AlphaGo, and TensorFlow [3] Q&A Insights - Ilya challenges the notion that next token prediction cannot surpass human performance, suggesting that a sufficiently advanced neural network could extrapolate behaviors of an idealized person [4][5] - He argues that predicting the next token well involves understanding the underlying reality that leads to the creation of that token, which goes beyond mere statistics [6][7] Ideal VLA and Reinforcement Learning - The ideal VLA operates by continuously predicting the next action token based on sensor information, indicating a real understanding of the physical world rather than just statistical probabilities [10] - Ilya posits that the reasoning process in the ideal VLA can be seen as a form of consciousness, differing from human consciousness in significant ways [11] Comparisons and Controversial Points - The article asserts that auxiliary driving is more suited for reinforcement learning compared to chatbots due to clearer reward functions [12][13] - It highlights the fundamental differences in the skills required for developing AI software versus hardware, emphasizing the unique challenges and innovations in AI software development [13]

Core Viewpoint - The release of GPT-5.0 has sparked discussions on the importance of open-source AI, highlighting the tension between innovation and control in the AI industry [1][3]. Group 1: Open Source vs. Closed Source - OpenAI's shift from open-source to closed-source with GPT-4 reflects broader uncertainties in the AI landscape, indicating a dynamic adjustment of productivity and production relations [3]. - The debate over open-source AI has evolved beyond technical governance to become a critical issue regarding the future direction of AI technology [3][20]. Group 2: Value of Open Source - Open-source software is estimated to provide a value of $8.8 trillion, significantly contributing to digital transformation [2]. - The open-source philosophy, emphasizing the "four freedoms," is increasingly recognized as essential for continuous innovation in software development [2][4]. Group 3: Challenges of Open Source in AI - Open-source AI faces criticism for being less transparent than traditional open-source software, with limitations on resource sharing that hinder technical replication and community learning [4][5]. - The licensing agreements for open-source AI often include restrictive clauses, contrasting with the traditional open-source spirit that promotes maximum inclusivity [5][6]. Group 4: Legal and Ethical Implications - The definition of "open-source AI" is contentious, with implications for legal responsibilities and protections under regulations like the EU's AI Act [7][20]. - The ongoing debate over the definition of open-source AI reflects deeper issues of public versus private interests and the evolving power dynamics in international relations [20]. Group 5: Geopolitical Context - The discourse surrounding open-source AI is increasingly intertwined with geopolitical considerations, as it can either foster international cooperation or exacerbate competition among nations [17][18]. - The U.S. government's approach to regulating open-source AI has shifted, indicating a complex interplay between national security and technological advancement [15][18]. Group 6: Future of Open Source in AI - The ongoing controversies surrounding open-source AI are not merely technical disagreements but are indicative of broader societal impacts and the future trajectory of AI development [20].

Core Viewpoint - The debate surrounding whether artificial intelligence (AI) should be open-sourced reflects broader concerns about the evolution of technology, its governance, and the balance between public and private interests in the AI landscape [2][18]. Group 1: Open Source AI Concept and Controversies - Open source software has historically been a foundation for digital technology, contributing an estimated $8.8 trillion in value to society, surpassing Japan's GDP [1]. - The shift from open-sourcing to closed-sourcing by companies like OpenAI highlights the dynamic adjustments in productivity and production relations within the AI sector [2]. - The complexity of open-sourcing AI involves multiple dimensions, including the openness of training frameworks, model weights, and the resources required for training, which differ from traditional open-source software [4][5]. Group 2: Ethical and Legal Implications - Critics argue that the open-sourcing behavior of AI companies may be more about public relations than genuine openness, leading to the term "openwashing" [5]. - The definition of "open source AI" is contentious, particularly regarding data sharing, as training data often involves copyright issues, complicating the push for transparency [6][5]. - The European Union's AI Act introduces legal responsibilities and exemptions for open-source AI, emphasizing the importance of defining its boundaries [6]. Group 3: Value and Performance of Open Source AI - The effectiveness of open-source AI in driving innovation is debated, with concerns that it may not match the performance of closed-source models due to resource constraints [8][9]. - The success of models like DeepSeek demonstrates that high performance can be achieved under limited resources, challenging the notion that only closed-source models can excel [9]. - Open-source AI is seen as a means to democratize technology and enhance productivity, with studies indicating higher investment returns for companies utilizing open-source AI [10]. Group 4: Risks and Governance - Concerns about the risks associated with open-source AI include potential misuse and the inability to ensure model safety, as highlighted by experts in the field [12][14]. - The Biden administration's regulatory approach to open-source AI has been criticized for imposing heavier compliance burdens compared to closed-source models, reflecting a perceived asymmetry in risk [14]. - The ongoing discourse around open-source AI risks will likely evolve, addressing broader societal impacts beyond traditional technical concerns [15]. Group 5: Geopolitical Context - The debate over open-source AI is intertwined with geopolitical dynamics, where it can either facilitate international cooperation or exacerbate competition among nations [16][17]. - The emergence of high-performance open-source models like DeepSeek challenges existing government controls over technology flow, indicating a shift in the landscape of AI development [17]. - The future trajectory of open-source AI amidst geopolitical tensions remains uncertain, with potential implications for global competition and collaboration [18].

Core Viewpoint - The open-source project ZLUDA, which enables non-NVIDIA chips to run CUDA, has been revived after facing near bankruptcy due to the withdrawal of AMD's support. A mysterious organization has stepped in to provide assistance, allowing the project to continue its development and support for large model workloads [1][2][12]. Historical Development - ZLUDA was initiated by Andrzej Janik, who previously worked at Intel, aiming to allow CUDA programs to run on non-NVIDIA platforms [4][5]. - Initially, ZLUDA was taken over by Intel as an internal project to run CUDA programs on Intel GPUs, but it was soon terminated [6][9]. - In 2022, ZLUDA received support from AMD but was again halted in February 2024 after NVIDIA released CUDA 11.6, which restricted reverse engineering on non-NVIDIA platforms [10][11][12]. Recent Developments - In October 2024, Janik announced that ZLUDA had received support from a mysterious organization, focusing on machine learning and aiming to restore the project to its previous state by Q3 2025 [13][15]. - The project has added a new full-time developer, Violet, who has made significant improvements, particularly in supporting large language model workloads [17]. Technical Progress - ZLUDA is working on enabling 32-bit PhysX support, with community contributors identifying and fixing errors that may also affect 64-bit CUDA functionality [19]. - A test project named llm.c is being developed to run the GPT-2 model using CUDA, marking ZLUDA's first attempt to handle both standard CUDA functions and specialized libraries like cuBLAS [20][22]. - The team has made progress in supporting 16 out of 44 required functions for the test program, indicating a step closer to full functionality [25]. Accuracy and Logging Improvements - ZLUDA aims to run standard CUDA programs on non-NVIDIA GPUs while matching NVIDIA hardware as closely as possible. Recent efforts have focused on improving accuracy by implementing PTX "scan" tests to ensure correct results across all inputs [26][28]. - The logging system has been significantly upgraded to track previously invisible activities and internal behaviors, which is crucial for running any CUDA-based software on ZLUDA [31][33]. Runtime Compiler Compatibility - ZLUDA has addressed issues related to the dynamic compilation of device code necessary for compatibility with modern GPU frameworks. Recent changes in the ROCm/HIP ecosystem have led to unexpected errors, but the ZLUDA team has resolved these problems [34][36][38].

Group 1: Event Overview - The fourth global elite "Generation Z" leaders online event was successfully held, gathering over 40 youth representatives from 15 renowned universities, including Shanghai Jiao Tong University and the University of California, Berkeley, to discuss "AI technology and future applications" [1][4] Group 2: AI Technology Insights - Yang Jian, a former core researcher from Alibaba's Tongyi team, highlighted the breakthrough in code intelligence technology, emphasizing that AI models have democratized programming, allowing code generation through natural language descriptions [4] - Echo Zhang from University College London stated that the core value of AIGC (AI-generated content) lies in "co-creation between humans and algorithms," illustrating its impact on personalized education and medical diagnostics with examples like Google DeepMind's "MedGemma" model [5] - Erum Yasmeen from Shanghai Jiao Tong University referenced a World Economic Forum statistic predicting that 85 million jobs will be displaced by AI, while new jobs will be created, stressing the importance of adapting faster than technology [9][10] Group 3: Educational Technology Evolution - Hua Xiaowen from Shanghai Jiao Tong University reviewed the evolution of educational technology, advocating that technology should enhance learners' individual expression and multiple intelligences rather than replace teachers [7] - The discussion included the introduction of AI courses in countries like Finland, encouraging students to engage with global issues such as sustainable development goals [7] Group 4: Data Analysis and AI Development - Duan Yuqing from the University of Auckland shared a thought-provoking perspective that "dirty data" can sometimes be more valuable than "clean data" for training AI models, particularly in financial fraud detection [12]

Core Insights - Google has made significant strides in the AI competition, particularly with the launch of Gemini 2.5, positioning itself on par with OpenAI [1][4] - The future of Google Search is envisioned to integrate advanced AI models that will enhance user experience by providing valuable content through multi-path retrieval [4][13] - The company is currently in the AJI (Artificial Jagged Intelligence) phase, indicating notable progress but also existing limitations in AI capabilities [4][42] Group 1: AI Development and Integration - Google aims to deploy the strongest models in search, executing multi-path retrieval for each query to deliver valuable content [4][13] - Approximately 30% of code is generated with the assistance of AI prompts, leading to a 10% increase in overall engineering efficiency [32][34] - The company is focused on creating a seamless integration of AI into its products, with plans to migrate AI Mode to the main search page [4][18] Group 2: Search and Advertising Evolution - The traditional search interface is evolving, with AI becoming an auxiliary layer that provides context and summaries while still directing users to human-created content [14][19] - AI Mode is currently being tested by millions, showing promising early indicators of user engagement and satisfaction [15][18] - Future advertising strategies will be rethought to align with AI capabilities, ensuring that ads are presented in a natural and unobtrusive manner [16][17] Group 3: Challenges and Future Outlook - Scaling laws remain effective, but the company acknowledges limitations in computational power affecting model deployment [29][30] - The integration of AR (Augmented Reality) is seen as the next significant interaction paradigm, with Project Astra being crucial for the Android XR ecosystem [36][38] - The company anticipates that while AGI may not be achieved by 2030, significant advancements will occur across various dimensions of AI [42][44]

Core Points - The article discusses the architecture and technology choices for social applications, emphasizing the importance of selecting the right frameworks and services for development [5][8][9]. Group 1: Frontend Development - The frontend of a social app consists of mobile (iOS/Android) and web applications, utilizing frameworks like React.js, Vue.js, and Angular for single-page applications [3][5]. - Mobile app development can be native (using Swift for iOS and Kotlin for Android) or cross-platform (using React Native, Flutter, uni-app, or Taro), each with its own advantages and disadvantages [6][8]. Group 2: Backend Development - The backend handles business logic, data storage, user authentication, and API interfaces, with popular frameworks including Spring Boot for Java, Django for Python, and Express.js for Node.js [9]. - Java is noted for its high performance and stability, making it suitable for large-scale applications, while Python offers rapid development capabilities for smaller projects [9]. Group 3: Database and Storage Solutions - Relational databases like MySQL and PostgreSQL are commonly used for structured data, while NoSQL databases like MongoDB and Redis are preferred for unstructured data and high-speed access [9]. - Object storage services from providers like Alibaba Cloud and Tencent Cloud are essential for managing user-generated content such as images and videos [9]. Group 4: Cloud Services and Compliance - For the Chinese market, compliance with local regulations, including ICP filing and app registration, is crucial, along with the selection of domestic cloud service providers like Alibaba Cloud and Tencent Cloud [8]. - The article highlights the importance of integrating third-party SDKs for functionalities like instant messaging and content moderation, with a focus on local providers [8][9]. Group 5: Development Tools and Technologies - The use of message queues (e.g., Kafka, RabbitMQ) and search engines (e.g., Elasticsearch) is recommended for system decoupling and enhancing user experience through personalized content [9]. - Containerization technologies like Docker and Kubernetes are suggested for efficient application deployment and management [9].

Google 首席科学家Jeff Dean 今年4月于在苏黎世联邦理工学院发表关于人工智能重要趋势的演讲，本次演讲回顾 了奠定现代AI基础的一系列关键技术里程碑，包括神经网络与反向传播、早期大规模训练、硬件加速、开源生 态、架构革命、训练范式、模型效率、推理优化等。算力、数据量、模型规模扩展以及算法和模型架构创新对AI 能力提升的关键作用。 以下是本次演讲 实录 经数字开物团队编译整理 01 AI 正以前所未有的规模和算法进步改变计算范式 Jeff Dean: 今天我将和大家探讨 AI 的重要趋势。我们会回顾：这个领域是如何发展到今天这个模型能力水平的？在当前的技 术水平下，我们能做些什么？以及，我们该如何塑造 AI 的未来发展方向？ 这项工作是与 Google 内外的众多同仁共同完成的，所以并非全是我个人的成果，其中许多是合作研究。有些工作 甚至并非由我主导，但我认为它们都非常重要，值得在此与大家分享和探讨。 我们先来看一些观察发现，其中大部分对在座各位而言可能显而易见。首先，我认为最重要的一点是，机器学习 彻底改变了我们对计算机能力的认知和期待。回想十年前，当时的计算机视觉技术尚处初级阶段，计算机几乎谈 ...

图片来源： Youtube 据The Information透露， 英伟达已完成对Lepton AI的收购， Lepton 联合创始人贾扬清和白俊杰均 已加入英伟达。 交易达成时，Lepton约有20名员工。 不久之前，英伟达被曝出即将达成收购 Lepton AI 的协议， 交易价值数亿美元。 这笔交易是英伟达 通过收购小型 AI 初创公司来推动其云和软件业务的一部分，使得使用英伟达芯片构建新的 AI 模型 成为可能。 根据公开信息，Lepton AI是一家成立于2023年的人工智能公司，致力于为企业提供高效、可扩展的 AI应用平台。 公司由阿里前 高管贾扬清创办。 Lepton AI支持模型的开发、训练和部署，具备生产级性能和成本效益，提供全面的机器学习工具和 灵活的GPU选项，满足企业级服务水平协议（SLA）的要求。 Lepton 之前曾为游戏初创公司 Latitude.io 和科学研究初创公司 SciSpace 提供 AI 云服务。 据钛媒体报道，Lepton AI在2023年5月成功完成天使轮融资，投资方包括硅谷知名风投CRV、红杉中 国和Fusion Fund。三家投资机构在短短不到两年时间内就实现 ...

以下文章来源于南风窗 ，作者荣智慧 南风窗 . 冷静地思考，热情地生活。 芯片必须越来越小。 作者 | 荣智慧 来源 | 南风窗 （ID：SouthReviews） 半导体领域的事儿，越来越"矛盾"。 晶体管的通道、软硬件之间的"次元壁"越来越小。而国家之间的"墙"越来越大。 在越来越小的领域，英伟达、AMD和台积电赚得盆满钵满。在越来越大的领域，金钱像 筹码一样押在跷跷板的两头——一头是美国，身后坐着拉美"后院"伙伴，非洲国家跟随 其后；一头是中国，东南亚和南亚正等着溢出的供应链；中国台湾、日本和韩国首鼠两 端。 更顽固的是消费者，今年大伙儿牢牢捂紧钱包，什么也不想买。随着当选总统特朗普第 二任期的逼近，更多的出口禁令、更高的关税、供应过剩和更富创造性的制裁规避方法 将在2025年出现。 越来越小 按价值计算，半导体现在是世界上交易量第三大的商品，仅次于石油和汽车。 处理能力每两年翻一番的摩尔定律，成功运行了半个多世纪。2 0 1 7年，英伟达创始人黄 仁勋宣布摩尔定律已死。2 024年，摩尔创立的英特尔的首席执行官帕特·基辛格坚称摩尔 定律还活着，年底，基辛格被大失所望的股东"炒了鱿鱼"。 在2024年， ...