Core Viewpoint - The article discusses the recent decline in the usage and market share of DeepSeek, questioning the validity of the reported statistics and emphasizing the importance of considering third-party API usage in evaluating its performance [2][4][10]. Summary by Sections Market Share and Usage Statistics - Reports from Semianalysis indicate that DeepSeek's market share has dropped to below 5%, with a significant decline noted since January [4][10]. - The statistics cited by Semianalysis primarily focus on the official API usage, potentially overlooking significant third-party integrations and deployments [10][12]. Third-Party API Usage - DeepSeek's third-party API calls have reportedly increased nearly 20 times since the release of versions V3 and R1, indicating sustained interest from developers [11][12]. - The article argues that the decline in official API usage does not reflect the overall demand for DeepSeek, as many applications integrate it without being captured in the official statistics [10][12]. Comparative Performance - Data from OpenRouter shows that DeepSeek V3 has a tokens consumption of 378 billion, ranking it third behind Claude Sonnet 4 and ahead of Google’s Gemini [17][22]. - Despite a decline in market share, DeepSeek maintains over 50% of the domestic B-end demand, indicating its strong position in the market [33]. User Preference and Community Engagement - A survey by Artificial Analysis found that 53% of respondents still prefer DeepSeek, placing it fourth among AI product providers [39]. - DeepSeek-R1 continues to lead in popularity on platforms like Hugging Face, indicating strong community support despite market fluctuations [44]. Industry Context and Future Outlook - The rapid evolution of AI technology suggests that a decline in DeepSeek's market share may not indicate a loss of relevance but rather reflects the dynamic nature of the industry [49]. - The article highlights the importance of open-source contributions from DeepSeek in promoting AI equity, contrasting it with other companies that are moving away from open-source models [49][50].

13年创业上市路，AI"老兵"的长期主义与现实生存法则。 作者｜甲小姐 创业13年，筹备上市5年，审核博弈649天——云知声创始人黄伟的故事像一面镜子，照见了中国AI企业的生存真相：在资本浪潮与技术理想之间，"留在 牌桌"需要勇气，更需要耐力。 在AI从"听音识人"到"通用智能"的十余年跋涉里，云知声是一家不够喧哗的公司。它没有站在算法竞赛最热闹的讲台上，也没有卷入千亿补贴的算力战 场。 更多人关注的是云知声上市之路的波折： 1."'老兵'能够活得老，还是有原因的" 甲小姐：敲钟那一刻，心情如何？ 黄伟：有一点喜悦。 甲小姐：不是超级喜悦？ 黄伟：我创业13年，筹备上市5年，中间经历过很多事，几年前想象的喜悦，在时间推移下都被抹平了。上市只是一个里程碑，是5年前的目标实现了，没 有更特殊的意味。 甲小姐：你有预料到上市后市值大涨吗？ 黄伟：没有。其实我头一两天会关注市值，后来就不关注了。市值只是资本市场阶段性对不同板块企业的认知差异，得用平常心看待。 2025年6月30日，云知声终于以"港股AGI第一股"的身份完成上市。敲钟台上，创始人黄伟的表情平静得近乎克制——没有振臂欢呼，没有热泪盈眶，只 有一句轻描淡写 ...

克雷西 发自 凹非寺 量子位 | 公众号 QbitAI 只需一句话，就能生成可实时交互的3D世界。 刚刚，谷歌DeepMind发布了 新一代通用世界模型Genie 3 。 性能上，Genie 3相比上一代大幅升级，支持 720P画质，每秒24帧实时导航，以及分钟级的一致性保持 。 | Genie 2 | Genie 3 | | --- | --- | | 360p | 720p | | 3D Environments | General | | Limited keyboard / mouse actions | Navigation; Promptable world events | | 10-20 seconds | Multiple minutes | | Not real time | Real time | 前DeepMind科学家、AI 3D生成创业者Tejas Kulkarni受邀体验了Genie 3。 他使用Genie 3，生成了长达57秒的城市高空漫游场景 （下图为节选） ： Tejas评价，Genie通用性强，还能学习物理，而且拥有强大的记忆力。 看了Tejas的测试，Reddit网友直 ...

Core Insights - OpenAI is increasingly cautious about protecting its talent from competitors like Meta, especially in light of the upcoming GPT-5 release [5][9][10] - The recent interview with OpenAI's chief scientist and technical expert highlighted the company's shift in approach towards employee confidentiality, contrasting with its previous openness [7][9] - Other tech companies, including Google and Apple, are actively enhancing their talent retention strategies in response to Meta's aggressive recruitment efforts [13][15] Group 1: OpenAI's Talent Protection - OpenAI has implemented strict measures to prevent the disclosure of employee names during interviews, indicating a heightened concern over talent poaching by Meta [5][7] - The shift in OpenAI's culture reflects a significant change from its earlier practices, where employees were freely discussed and celebrated [9][10] - The urgency to protect talent is particularly pronounced with the imminent launch of GPT-5, as OpenAI fears further recruitment attempts from Meta [9][10] Group 2: Competitive Landscape - Meta's recruitment strategy has led to a significant number of former OpenAI employees joining its ranks, forming a strong talent pool [9][13] - Google has responded by promoting internal talent to higher positions to prevent them from being poached by Meta [13][15] - Apple has increased salaries for its foundational model team to retain key members, while other companies like Anthropic are taking a different approach by refusing to raise salaries indiscriminately [13][15] Group 3: Employee Perspectives - OpenAI employees, such as Jakub and Szymon, have expressed their commitment to the company due to its supportive environment and funding, which they believe fosters innovation [9][10] - The discussion around AGI has evolved within OpenAI, with employees now viewing it as a critical milestone rather than just a system for economic value [12] - Employee retention remains a challenge across the industry, with Meta's retention rate at 64%, significantly lower than Anthropic's 80% and OpenAI's 67% [15][16]

Model Overview - GPT-OSS is a state-of-the-art open-weights reasoning model with strong real-world performance comparable to O4-Mini [1] - The model can be run locally on personal computers or phones [1] - The company believes this is the best and most usable open model currently available [1] Impact and Availability - The model, resulting from billions of dollars of research, is being made available to the world to democratize AI access [2] - The company anticipates more good than bad will result from its release, citing performance on challenging health issues comparable to O3 [2] - The company has worked to mitigate serious safety issues, especially around biosecurity, with GPT-OSS models performing comparably to frontier models on internal safety benchmarks [2] - The company hopes this release will enable new kinds of research and the creation of new kinds of products, expecting a meaningful uptick in the rate of innovation [4] Strategic Vision - The company believes in individual empowerment, providing users with direct control and modification capabilities over their own AI, highlighting privacy benefits [3] - The company's mission is to ensure AGI benefits all of humanity [5] - The company is excited for the world to build on an open AI stack created in the United States, based on democratic values, available for free to all and for wide benefit [5]

Group 1: AI Model Developments - Claude Opus 4.1 is currently in internal testing and is expected to be released within two weeks, focusing on enhancing reasoning and planning capabilities [1] - Anthropic's annual revenue has increased fivefold to $5 billion, with programming clients like Cursor and GitHub Copilot contributing $1.4 billion in API revenue [1] - Alibaba has open-sourced the Qwen-Image model, which has 20 billion parameters and excels in rendering complex text in images, achieving state-of-the-art performance in multiple benchmarks [3] Group 2: New Features and Innovations - Tencent's ima has introduced new features including AI podcast capabilities that convert articles into dialogue format and a one-click folder import function that retains file hierarchy [2] - Huawei has open-sourced three Pangu models with sizes of 1 billion, 7 billion, and 718 billion parameters, including the Ultra MoE model, which utilizes a mixed expert architecture [4] - Nanom AI has launched a multi-agent swarm capable of generating high-quality AI videos lasting up to 10 minutes, significantly reducing production costs by 95% [5] Group 3: Competitive Landscape - Google has initiated the first large model competition, featuring eight top AI models competing in chess, including those from OpenAI, DeepSeek, and Anthropic [6][7] - A warning from former Google executive Mo Gawdat predicts that by 2027, AI will lead to a "hell period" where the middle class will be eradicated, leaving only the top 0.1% and the lower class [10] Group 4: Company Strategies and Future Outlook - Jieyue CEO announced the first open-source base model, Step 3, which has a total of 321 billion parameters and focuses on multi-modal reasoning [11] - The company is committed to the integration of multi-modal generation and understanding as a pathway to AGI, despite facing resource challenges [11] - Yushu Technology has introduced the Unitree A2 quadruped robot, designed for industry applications, and is preparing for an IPO with projected revenue exceeding 1 billion in 2024 [9]

RT Jack Parker-Holder (@jparkerholder)Genie 3 feels like a watershed moment for world models 🌐: we can now generate multi-minute, real-time interactive simulations of any imaginable world. This could be the key missing piece for embodied AGI… and it can also create beautiful beaches with my dog, playable real time https://t.co/YSIAwV3GiS ...

DeepMind thinks its new Genie 3 world model presents a stepping stone towards AGI | TechCrunch https://t.co/sADQm6kQCJ ...

DeepMind reveals Genie 3, a world model that could be the key to reaching AGI | TechCrunch https://t.co/i2UDsxXFqv ...

Model Overview - Google announced Genie 3, a general-purpose world model for generating diverse interactive environments [1][8] - Genie 3 allows real-time interaction with improved consistency and realism compared to Genie 2 [12] - The model generates 720p high-quality environments [3] Technical Aspects - Genie 3 considers the entire previously generated trajectory, not just the previous frame, for autoregressive generation [15] - Consistency in Genie 3 is an emergent capability resulting from training scale, not pre-programming [19] - Genie 3 generates dynamic and rich worlds frame by frame based on world description and user actions, unlike methods relying on explicit 3D representation [20] Potential Applications - World models like Genie 3 can be used for training robots and agents [9] - The technology has potential applications in creating video games, movies, and television shows [9] - Google positions world models as a key step towards AGI by providing AI agents with unlimited simulation environments for training [9][10] Comparison with Previous Models - Genie 3 demonstrates significant improvements in consistency, detail, and generation length compared to Genie 2 [22][23] - Genie 3 allows for deeper world exploration than Genie 2 [23] Interactive Features - Users can prompt events in real-time, adding elements to the scene [21] - The model demonstrates realistic interactions, such as light moving out of the way of a jet ski and reflections in mirrors [6] - The model can simulate actions like painting, with paint only being applied when the brush touches the wall [29][30]