Core Insights - The article discusses the evolution of computer vision towards a new paradigm, emphasizing the transition from basic pixel perception to complex spatial reasoning and world modeling, facilitated by Test-time Scaling (TTS) [2][5] - The upcoming ViSCALE 2026 conference aims to gather leading scholars to explore breakthroughs in visual models through computational expansion, focusing on deep reasoning rather than mere static outputs [4][5] Group 1: Conference Highlights - ViSCALE 2026 will feature discussions on spatial intelligence and world models, with contributions from top scholars including Sergey Levine, Manling Li, and Ziwei Liu [5] - The conference encourages innovative research submissions that challenge existing visual model limitations, providing a platform for both theoretical and application-focused studies [7] Group 2: Key Topics of Discussion - The conference will cover various topics, including: - Enhancing video generation's physical consistency and long-term causal reasoning through TTS [6] - Breaking 2D limitations to enable models to navigate and operate in 3D spaces like humans [6] - Developing visual reasoning chains that allow models to self-correct and engage in multi-step reasoning [6] - Exploring scaling laws that relate computational load during testing to visual reasoning performance [6] Group 3: Submission Details - The conference invites submissions in two tracks: Full Papers (8 pages) and Extended Abstracts (up to 4 pages), with specific formatting requirements [9] - Important deadlines include submission by March 10, 2026, and notification of acceptance by March 18, 2026 [9]

Core Viewpoint - The article discusses how NetEase has chosen a pragmatic approach in the AI era, avoiding the costly competition of developing general-purpose AI models while focusing on application-level innovations and maintaining a strong R&D investment [3][20]. Group 1: Market Context - During the recent Spring Festival, major tech companies like Alibaba, Tencent, ByteDance, and Baidu spent over 4.5 billion yuan on "red envelopes," marking one of the most expensive tech competitions in history [2]. - NetEase, however, did not participate in this "red envelope war" or the race for large AI models, leading to questions about whether it is falling behind in the AI era [2][3]. Group 2: Business Strategy - NetEase's strategy is characterized by a focus on practical applications rather than competing in the foundational AI model space, which is seen as a more sustainable approach for most companies [3][20]. - The company has maintained a consistent R&D expenditure of over 15% of its revenue, with a projected R&D budget of 17.7 billion yuan for 2025, focusing on application layers rather than general model training [4][20]. Group 3: AI Integration in Products - NetEase has developed thousands of AI production pipelines that enhance various aspects of game development, achieving significant efficiency improvements, such as a 70% increase in design efficiency and a 50% boost in development efficiency through AI tools [6][8]. - The company has also applied AI in its educational and music platforms, enhancing user experience and operational efficiency without pursuing a general-purpose model [6][8]. Group 4: Financial Performance - In 2025, NetEase's total revenue is expected to reach 112.6 billion yuan, with operating profit at 35.8 billion yuan, driven primarily by its gaming segment, which saw a net revenue increase of 11% year-on-year [13][20]. Group 5: Future Growth Potential - The potential for growth in the gaming industry is seen in AI-native games, which are expected to generate over 30 billion yuan by 2027, contributing to a 10% market increase [13][20]. - NetEase's focus on integrating AI into its gaming products positions it well to capitalize on this emerging market, as it transitions from traditional gaming to AI-driven experiences [13][20].

Core Insights - The industry of embodied intelligence is at a crossroads of capital and industrial focus, with increasing financing and frequent technological demonstrations, yet facing challenges in stability, scalability, and cost control [1] Group 1: Financing and Valuation - Starry Sea has completed a Series B financing round of 1 billion yuan, bringing its total financing to nearly 3 billion yuan and achieving a valuation of 10 billion yuan, making it a unicorn in the embodied intelligence sector [1] - The CFO of Starry Sea emphasizes that the success in the AI industry is driven by Scaling Law, where the efficiency of capital utilization is more critical than the amount of financing [1][2] Group 2: Industry Dynamics - The current phase of the embodied intelligence industry is compared to the "Hundred Groups War," where companies are advised to focus on understanding the essence of business rather than just technology [2] - The industry is transitioning from early-stage technology exploration to resource-intensive competition, with a shift in capital logic from broad investment to focusing on leading companies [2] Group 3: Commercialization and Technology - The commercialization of embodied intelligence is divided into technology-driven and business-driven aspects, with specific operational boundaries that need to be met for successful deployment [4] - The CFO believes that the industry is still in the early stages of a technological race, and companies must retain sufficient funds to cope with the increasing costs of data and model training [2][4] Group 4: Financial Potential and Business Model - The ToB (business-to-business) segment of embodied intelligence has significant revenue potential, with large orders capable of generating substantial income, but the focus should be on revenue quality metrics [5] - The long-term business model in this industry is likened to selling "tokens of the physical world," with the real barriers being intelligence levels and the ability to design and manufacture hardware [5] Group 5: Competitive Advantages - China is recognized for its data supply chain advantages, which are significantly more cost-effective than those in the U.S., allowing for greater data collection at lower costs [6] - The CFO highlights that the unique aspect of embodied intelligence companies lies in developing their foundational models for physical world execution, emphasizing the need to focus resources on building these capabilities [7]

Core Insights - Runway, an AI video company, has shifted its focus to world models and has secured significant investment from Nvidia and AMD, indicating strong industry confidence in its new direction [1][2][12]. Company Overview - Runway was founded in 2018 by three art students and has undergone two major transformations, leading to a current valuation of $5.3 billion with only 140 employees [1][4]. - The company initially focused on video editing tools, gaining traction with its "green screen" feature, which led to early funding rounds totaling $200 million [6][8]. Recent Developments - Runway completed its Series E funding round, raising $315 million (approximately 2.17 billion RMB) to develop the next generation of world models [2][4]. - The latest funding round was led by General Atlantic, with participation from Nvidia and AMD, reflecting a strong belief in Runway's potential [2][12]. Valuation Growth - Following the recent funding, Runway's post-money valuation nearly doubled to $5.3 billion (approximately 36.58 billion RMB) [4][12]. - The company has seen a steady increase in valuation through its strategic pivots, particularly its entry into generative AI following the launch of ChatGPT [6][8]. Product Evolution - Runway's product evolution includes the introduction of the Gen-1 and Gen-2 models, with Gen-2 being the first commercially viable text-to-video model [8][10]. - The company has recently launched the GWM-1 (General World Models-1), which allows for interactive control and real-time image generation, marking a significant advancement in its technology [10][12]. Industry Context - The world model technology is gaining traction across various sectors, including autonomous driving, with companies like Tesla and Waymo developing their own models [13][17][22]. - Nvidia's investments in Runway and other companies utilizing world models highlight the growing importance of this technology in the AI landscape [12][22].

以下文章来源于极客公园 ，作者金光浩 这两天，AI视频圈被偷摸摸上线的Seedance2.0刷屏了。 在AI视频领域颇有影响力的博主海辛，在即刻分享了自己对它的观点： 「Seedance2.0是我26年来最大的震撼」、「我觉得它碾压Sora2」。 真的如此吗？一点都不夸张。 这是它做出来的视频，一句话音画同出，几乎无限逼近于影院里看到的电影。 极客公园 . 用极客视角，追踪你最不可错过的科技圈。欢迎同步关注极客公园视频号 「 AI春运 」 ，开始了。 文 ｜ 金光浩 编辑 ｜ 靖宇 来源｜ 极客公园（ID：geekpark） 封面来源 ｜ IC Photo 虚拟科幻眼镜幻想视频｜视频来源：Seedance2.0飞书文档 字节自己在飞书里发了一份产品介绍文档，标题只有几个字，但意味重大： 视频Seedance2.0正式上线！Kill the game（杀死比赛）。 我在2月7号下午看到了这份文档，出于好奇点进去想快速扫一遍，结果一看就到了晚上。文档右上角显示的同时在线人数， 从下午两点到晚上十二点，几 乎没有掉到300人以下 。我凌晨四点关掉页面的时候，还有90多人同时在线读文档呢（可能是周日的缘故？）。 2月 ...

Core Viewpoint - NetEase's Q4 2025 financial results showed a revenue of 27.5 billion RMB, a 3% year-on-year increase, but a net profit attributable to shareholders of 6.2 billion RMB, down nearly 30% from 8.8 billion RMB in the same period last year, falling short of market expectations [2][4]. Financial Performance - For Q4 2025, NetEase's revenue was 27.5 billion RMB, with a year-on-year growth of 3% [2]. - The net profit attributable to shareholders for Q4 was 6.2 billion RMB, a decrease of approximately 30% compared to 8.8 billion RMB in Q4 2024 [2][4]. - Total revenue for the full year 2025 reached 112.6 billion RMB, representing a year-on-year increase of about 7% [4]. - The net profit attributable to shareholders for the full year 2025 was 33.8 billion RMB, up 13.8% year-on-year [4]. Cost and Expenses - Sales and marketing expenses in Q4 increased by approximately 1.07 billion RMB to 3.89 billion RMB compared to the same period in 2024 [2]. - Overall investment losses reached 1.67 billion RMB, an increase of about 1.2 billion RMB, along with foreign exchange losses exceeding 500 million RMB [2]. Business Segments - The core gaming and related value-added services generated revenue of 22 billion RMB in Q4, a year-on-year increase of 3.4%, accounting for 80% of total revenue [4]. - The flagship games such as "Fantasy Westward Journey" and "Identity V" supported the revenue base, while new games like "Yanyun Sixteen Sounds" and "Marvel Showdown" contributed to revenue growth [4]. - Other business segments included NetEase Youdao with Q4 revenue of 1.6 billion RMB, up 16.8% year-on-year, and NetEase Cloud Music with revenue of 2 billion RMB, a 4.7% increase year-on-year [4]. Cash Flow and Financial Health - As of December 31, 2025, NetEase's net cash balance was 163.5 billion RMB, up from 131.5 billion RMB in 2024 [7]. - The net cash flow from operating activities for 2025 was 50.7 billion RMB, compared to 39.7 billion RMB in 2024 [7]. Industry Insights - Management discussed the impact of AI on the gaming industry, noting that while AI lowers the entry barrier for game development, it raises the success threshold for major commercial titles [5][6]. - The introduction of generative models like Google's Genie 3 has caused significant market reactions, but management believes the market may misunderstand its implications for the gaming industry [5][6].

Core Viewpoint - NetEase's Q4 2025 financial results showed a revenue of 27.5 billion RMB, a year-on-year increase of 3%, but a net profit attributable to shareholders of 6.2 billion RMB, down nearly 30% from 8.8 billion RMB in the same period last year, falling short of market expectations [3] Financial Performance Summary - For Q4 2025, NetEase's total revenue was 27.5 billion RMB, with a net profit of 6.2 billion RMB, which is a significant decline compared to the previous year's 8.8 billion RMB [3][5] - The total revenue for the year 2025 reached 112.6 billion RMB, representing a year-on-year growth of approximately 7%, while the net profit attributable to shareholders was 33.8 billion RMB, up 13.8% [5] - The increase in sales and marketing expenses in Q4 2025 was approximately 1.07 billion RMB, reaching 3.89 billion RMB, and overall investment losses amounted to 1.67 billion RMB, a significant increase of about 1.2 billion RMB [3][5] Business Segment Performance - The core gaming and related value-added services generated revenue of 22 billion RMB in Q4 2025, a year-on-year increase of 3.4%, accounting for 80% of total revenue [5][6] - New game launches, including "Yanyun Sixteen Sounds" and "Marvel Showdown," contributed to revenue growth, while established titles like "Fantasy Westward Journey" and "Identity V" supported the performance [6] - Other business segments showed varied performance, with NetEase Youdao's revenue increasing by 16.8% to 1.6 billion RMB, while NetEase Cloud Music's revenue rose by 4.7% to 2 billion RMB, and other innovative businesses saw a decline of 10.4% to 2 billion RMB [6] Management Insights - NetEase's management discussed the impact of AI on the gaming industry, emphasizing that while AI lowers the entry barrier for game development, it raises the success threshold for high-quality products [6][7] - The management believes that the true potential of AI lies in creating new entertainment types distinct from traditional games, although current AI models are not yet suitable for conventional gaming [7] Cash Flow and Financial Health - As of December 31, 2025, NetEase's net cash balance was 163.5 billion RMB, up from 131.5 billion RMB in 2024, with net cash flow from operating activities amounting to 50.7 billion RMB, compared to 39.7 billion RMB in 2024 [7]

Investment Rating - The report maintains a positive outlook on the AI industry, particularly focusing on advancements in large model technologies and their applications in various productivity scenarios [2][3]. Core Insights - In 2025, global large model capabilities advanced significantly, overcoming challenges in reasoning, programming, and multimodal abilities, although issues like stability and hallucination rates remain [2][3]. - Looking ahead to 2026, breakthroughs in reinforcement learning, model memory, and context engineering are anticipated, moving from short context generation to long reasoning chain tasks and from text interaction to native multimodal capabilities [2][3][4]. - The scaling law for pre-training is expected to continue, with flagship models achieving higher parameter counts and intelligence limits, driven by advancements in NVIDIA's GB series chips and the adoption of more efficient model architectures [3][4]. Summary by Sections Model Architecture and Optimization - The report emphasizes the continuation of the Transformer architecture, with a consensus on the efficiency of the Mixture of Experts (MoE) model, which balances performance and efficiency [40][41]. - Various attention mechanisms are being optimized to enhance computational efficiency, with a focus on hybrid approaches that combine different types of attention for better performance [49][50]. Model Capabilities - The report highlights significant improvements in reasoning, programming, agentic capabilities, and multimodal tasks, indicating that large models have reached a level of real productivity in various fields [13][31]. - The ability of models to perform complex reasoning tasks has improved, with the introduction of interleaved thinking chains allowing for seamless transitions between thought and action [24][28]. Market Dynamics - The competition among leading global model manufacturers remains intense, with companies like OpenAI, Anthropic, and Gemini pushing the boundaries of model intelligence and exploring AGI [31][32]. - Domestic models are catching up, maintaining a static gap of about six months behind their international counterparts, with significant advancements in capabilities [32][33]. Future Outlook - The report anticipates that the introduction of continuous learning and model memory will address the "catastrophic forgetting" problem, enabling models to adapt dynamically based on task importance [4][5]. - The integration of high-quality data and large-scale computing resources is crucial for enhancing the capabilities of reinforcement learning, which is expected to play a key role in unlocking advanced model functionalities [3][4].

图片来源： Runway 知情人士 透露， AI 视频生成初创公司 Runway 已完成 3.15 亿美元 E 轮融资，公司估值飙升至 53 亿美元，较之前水平近乎翻倍。 公司在其宣布融资的博客中表示，新资金将使 Runway 能够 " 预训练下一代世界模型，并将其引入新产品和行业 " 。 世界模型是一种能够构建环 境内部表征的人工智能系统，从而能够对未来事件进行规划，许多顶尖学者认为这类模型对突破大语言模型的局限至关重要。 据公司发言人透露，展望未来， Runway 计划运用新资金将其约 140 人的团队在研发、工程和市场拓展等岗位进行快速扩容。 本轮融资由 General Atlantic 领投，参投方包括英伟达、富达管理与研究公司、 AllianceBernstein 、 Adobe Ventures 、未来资产、 Emphatic Capital 、 Felicis 、 Premji 以及 AMD Ventures 。 参考资料： https://techcrunch.com/2026/02/10/ai-video-startup-runway-raises-315m-at-5-3b-valuatio ...

Core Insights - Runway, founded in 2018 by three NYU alumni, has raised a total of $815 million (approximately 5.6 billion RMB) in funding, with the latest round in April 2025 securing $308 million (approximately 2.1 billion RMB) from investors including SoftBank and NVIDIA, leading to a valuation exceeding $3 billion (approximately 20.7 billion RMB) [5][10] - The company is renowned for its video generation products and recently launched its latest model, Gen-4.5, which can produce high-fidelity outputs suitable for film, including complex scenes and realistic physical effects [5][10] - Gen-4.5 currently ranks third in the global AI text-to-video model performance leaderboard, outperforming notable models from Google and OpenAI [5][6] Funding and Future Plans - The new funding will be utilized to train the next generation of world models and expand into new products and industries [7] - Following the release of Gen-4.5, Runway introduced the General World Model (GWM-1), designed for real-time simulation and interaction, with three variants aimed at different applications [7][9] Technological Advancements - Runway is leveraging the NVIDIA Rubin platform to enhance its video generation and world model technologies, being one of the first teams to showcase video generation models on this platform [9] - The company has partnered with CoreWeave, a US AI cloud service provider, to expand its infrastructure and computational capabilities, with NVIDIA being a key supporter and supplier [9] Market Position and Competition - Runway's recent advancements have rekindled investor interest, especially after surpassing competitors like OpenAI Sora and Kuaishou Keling in benchmark tests [10] - The company is making significant investments in the world model sector, which is highly competitive, with advancements from Stanford's World Labs and Google's DeepMind [10]