图片来源： Sequoia Capital Z Highlights Bob McGrew ， OpenAI 前首席研究官，主导推动 GPT ‑ 3 、 GPT ‑ 4 以及内部称为 o1/o3 模型的研发，提出预训练（ pre-training ）、后训练（ post-training ）和推理（ reasoning ）的 " 三位一体 " 模型。现为多家 AI 初创企业的顾问或投资人，持续推动 AGI 的落地。本次访谈视频由 Sequoia Capital 在 2025 年 6 月 17 日发布，和 Bob 共同探讨了 从模型训练重点、 Agent 和机器人的未来发展、 AI 时代的教育心得与管理经验等主题，洞察人工智能的发展轨迹，并 指出初创企业依然可以挖掘并构建可持续竞争优势的领域。 预训练、后训练和推理，未来如何发展？ Stephanie Zhan ： 欢迎来到 Training Data 。今天我们非常高兴邀请到 Bob McGrew——OpenAI 前首席研究官，带我们深入探讨 frontier AI 的幕后发 展。 Bob 分享了预训练（ pre-training ）、后训练（ post-tr ...

机器之心报道 机器之心编辑部 NLP 领域被引用次数最多的研究者之一、斯坦福人工智能实验室（Stanford AI Lab）前主任克里斯托弗・曼宁（Christopher Manning）已从斯坦福大学休假，加入 风险投资公司 AIX Ventures 担任普通合伙人。 来源：https://www.wsj.com/articles/ai-researcher-christopher-manning-takes-leave-from-stanford-for-aix-ventures-0ab3cb4e?st=gLsy7t 对于曼宁的加入，AIX Ventures 的创始合伙人 Shaun Johnson 表示：「所有顶尖的 AI 原生工程师都认识 Chris，并且他们都希望与他合作。」 NLP 领域的先驱 曼宁教授是将深度学习应用于 NLP 领域的早期领军人物，在词向量 GloVe 模型、注意力、机器翻译、问题解答、自监督模型预训练、树递归神经网络、机器推 理、依存解析、情感分析和总结等方面都有著名的研究。 此前，曼宁自 2021 年起以兼职投资人身份与 AIX 展开合作，现在将全职投入为公司提供咨询服务。 ...

Welcome back to the week in AI. I'm Kevin Cook, your field guide and storyteller for the fascinating arena of artificial intelligence. We did another space on Friday, my myself and my colleague Ethan Feller on um the 28th, Friday the 28th. I had to remember now that we're in July here. Um here were here was the the post basically the invitation for people to to come with my big topics. What's more doomer AI AGI imminent meaning tomorrow or AGI never? That's an important question. And then welcome to the sca ...

本文为IPO早知道原创 作者｜ Stone Jin 微信公众号｜ipozaozhidao 据 IPO早知道消息，智谱在 7月2日 举行的 智谱开放平台产业生态大会 宣布获得浦 东创投集团和 张江集团联合战略投资，总额 10亿元 。 同时， 智谱 CEO张鹏在主题演讲中发布了智谱携手生态伙伴迈向AGI的两项最新成果：一是开源发 布新一代通用视觉语言模型GLM-4.1V-Thinking，以推理能力为核心突破，刷新10B级别多模态模 型性能上限；二是MaaS全新上线Agent聚合平台「应用空间」，全面激活行业场景中的AI能力，联 动Z基金启动Agent开拓者数亿元专项扶持计划。 为智谱构建可信的人工智能基础设施注入坚实动能。 其中， 视觉语言大模型 GLM-4.1V-Thinking 的发布并开源 ，标志着 GLM系列视觉模型实现从 感知走向认知的关键跃迁 。 GLM-4.1V-Thinking是一款支持图像、视频、文档等多模态输入的通用推理型大模型，专为复杂认 知 任 务 设 计 。 它 在 GLM-4V 架 构 基 础 上 引 入 " 思 维 链 推 理 机 制 （ Chain-of-Thought Reas ...

Group 1 - Meta successfully recruited three core researchers from OpenAI, indicating the effectiveness of its aggressive hiring strategy led by CEO Mark Zuckerberg [1] - The recruited researchers, Lucas Beyer, Alexander Kolesnikov, and Xiaohua Zhai, were previously responsible for establishing OpenAI's Zurich office and joined Meta's Superintelligence team [1] - Zuckerberg's recruitment strategy includes offering over $100 million compensation packages to attract top talent from competitors like OpenAI [2] Group 2 - OpenAI CEO Sam Altman acknowledged the high offers from Meta but expressed confidence that their best talent has not accepted these proposals [2] - Meta's recent hiring of Alexandr Wang, CEO of Scale AI, for $14 billion marks one of the most expensive hires in tech history, although it has not successfully recruited other key figures from OpenAI [2] - Meta faced setbacks in the AI field, particularly with the disappointing performance of the Llama 4 model, which led to internal and external criticism regarding its capabilities [3] Group 3 - The launch of Meta's large model "Behemoth" has been delayed, raising concerns within the leadership about its competitive edge compared to products from OpenAI, Anthropic, and Google [3] - Zuckerberg's ambition for Meta to have the best AI product by year-end has resulted in increased pressure on the AI team, leading to long hours and unmet expectations [3]

编辑：桃子 【导读】 AI大神何恺明正式入职谷歌DeepMind，担任杰出科学家，同时保留MIT终身副教授身 份。从Meta到MIT，再到如今的谷歌，这位「学界+业界」双修的大牛，将为DeepMind的AGI注 入一针强心剂。 AI圈炸了！CV大牛何恺明正式官宣入职谷歌。 已更新的个人主页上，明确写着：兼职谷歌DeepMind杰出科学家。 转自：新智元 与此同时，他依然保留MIT EECS终身教授的身份。 这位CV领域的传奇人物，因提出ResNet而名震江湖，彻底改变了深度学习的发展轨迹，成为现代AI 模型的基石。 如今，这位「学界+业界」双轨并行的跨界大神，再次用行动证明了他的无限可能！ 对于谷歌DeepMind而言，何恺明的加入更是如虎添翼。 他的技术专长，涵盖了计算机视觉、深度学习等核心领域，学术影响力在全球范围内有目共睹。 Demis Hassabis曾公开表示，AGI可能在未来5-10年内实现。 何恺明的到来，无疑将助力这一终极目标的加速实现。 ResNet之父再跨界，DeepMind迎超级大脑 何恺明曾是微软亚研院研究院、Meta「明星实验室」FAIR的研究科学家，专注的研究领域包括深度 学习和计 ...

OpenAI创始人Sam Altman近日在与YC总裁Garry Tan(陈嘉兴)的访谈中，重磅预告了GPT大模型的最新进展：OpenAI即将推出强大的开源模型，并暗示了 GPT-5的部分能力。 这次访谈信息量极大，我们给大家划下重点： 一、OpenAI即将推出一个开源模型，GPT-5迈向完全多模态 OpenAI最终的完全多模态，将具备深度推理能力，能够进行深入研究，生成实时视频，并能够编写大量代码，即时为用户创建全新的应用程序，甚至渲 染提供用户交互的实时视频。当OpenAI实现完全多模态时，这将带来一种全新的计算机界面。 二、AI发展现状与产品创新：潜力巨大，成本骤降。 Sam Altman认为，当前AI模型的能力（如GPT-3）显然超越了现有产品的应用水平，存在巨大的"产品溢出"空间。这意味着，即使模型能力不再提升，也 能支撑海量的新产品开发。 AI模型的使用成本正在急剧下降，例如GPT-3的成本在短短一周内就下降了五倍，且这种趋势将持续下去，价格与性能之比的下降幅度让人震惊。 Sam Altman提到ChatGPT的内存功能，将让AI发展成为一个了解用户、持续运行、主动提供帮助并整合到各种设备与服务中 ...

Core Viewpoint - Nvidia's upcoming shareholder meeting is unlikely to be a significant catalyst for its stock, but long-term investors may still consider buying shares before the meeting due to the potential growth in the AI sector [1][10]. Company Events - The annual shareholder meeting will primarily focus on electing the board of directors, with expected approval for the nominated slate [4]. - A vote on the advisory approval of executive compensation will take place, although it is non-binding [5]. - The meeting will also address the elimination of supermajority voting provisions, which requires a 66 2/3% shareholder vote to become effective [6]. - Three stockholder proposals will be discussed, including the elimination of the one-year holding period for special meetings and a request for enhanced public reporting on employee demographics [7][8]. Future Catalysts - The next significant event that could impact Nvidia's stock is the fiscal year 2026 second-quarter results scheduled for August 27, 2025, although it may not significantly affect the stock unless earnings estimates are exceeded [9]. - The adoption of AI and advancements in artificial general intelligence (AGI) are expected to provide substantial growth opportunities for Nvidia in the long term [10][11]. Valuation Perspective - Concerns about Nvidia's valuation persist, but if the AI market develops as anticipated, the company's growth could justify its current valuation [12].

Microsoft Today MSFT Microsoft $476.84 -1.20 (-0.25%) 52-Week Range $344.79 ▼ $480.69 Dividend Yield 0.70% Add to Watchlist Microsoft Corporation NASDAQ: MSFT stock is up 23% in the last three months and is now at an all-time high of around $478. At times like this, it's normal for investors to consider taking some profits. Some technical signals point to the potential that the rally is getting tired. However, MSFT stock is one of the clear winners in the market this year. Get Microsoft alerts: That means i ...

Core Insights - Unisound Intelligent Technology Co., Ltd. has passed its listing hearing with the Hong Kong Stock Exchange, aiming to become the first AGI stock to debut in Hong Kong this year [2][3] - The company specializes in voice intelligence and integrated AI solutions, with a flagship model launched in 2023 featuring 60 billion parameters [4] - Unisound has raised over $340 million through 11 funding rounds, with significant backing from notable investors [8] Company Overview - Founded in 2012, Unisound focuses on AI solutions for healthcare and lifestyle services, ranking as the fourth-largest AI solutions provider in China by 2024 revenue [4][7] - The company is led by CEO Huang Wei, who emphasizes the importance of pursuing challenging innovations in AI [5][6] Financial Performance - Unisound's revenue increased from RMB 601 million in 2022 to RMB 939 million in 2024, reflecting a compound annual growth rate of 25% [9] - Despite revenue growth, the company reported widening net losses from RMB 375 million in 2022 to RMB 454 million in 2024 [9] - R&D expenses rose by 29% year-on-year to RMB 370 million in 2024, contributing to ongoing losses [10] Business Model and Revenue Streams - The lifestyle AI segment is the primary revenue driver, contributing nearly 79% of total revenue in 2024 [11] - A shift in business model is noted, with revenue from system integrators/agents surpassing that from direct users [11] Future Outlook - Proceeds from the IPO will be allocated to scaling AI infrastructure, funding R&D, pursuing international expansion, and investing in new verticals [11] - CEO Huang remains optimistic about the future of AI in China, highlighting the inevitability of large model innovations [12]