面向医生的AI搜索工具OpenEvidence以120亿美元估值完成新一轮融资，首席执行官兼联合创始人丹尼尔·纳德勒身家翻倍，达76亿美元。图片来源： MAURICIO CANDELA FOR FORBES 目前，约74万名医生——约占美国医生总数的45%——使用这款搜索引擎，快速检索顶级医学期刊中数百万篇经同行评审的科研论文，几秒钟就可找到有 用（有时甚至能挽救生命）的信息，而不是像过去那样耗费数小时甚至数天。纳德勒表示，上个月医生在约1800万次临床问诊中使用了该软件。他对《福 布斯》表示："我们已经成为医生群体的默认操作系统。" 纳德勒称，OpenEvidence在2025年的年化收入突破1亿美元，已通过广告实现盈利，尽管大部分广告资源尚未开放出售。他表示，如果公司全面开放所有 广告资源，年收入可达约10亿美元。但为优先保障用户体验，他并不打算这么做——这与谷歌早期的做法颇为相似。制药公司和医疗器械公司可以在应用 内购买5秒的视频广告，广告会根据医生输入的搜索关键词进行精准投放。 凭借先发优势以及在美国医疗专业人士中的快速普及，纳德勒并不太担心OpenAI的ChatGPT等竞品会赶超上来。这款热门聊天机 ...

国内产品追赶OpenEvidence的步伐开始加速。 OpenEvidence凭借着短期内覆盖40%美国医生群体的成绩，让国内从业者意识到只有像OpenEvidence这样融入医生的工作生态，AI产品才更有可能获得医 生和院方的认可。因此，国内众多产品开始从医生视角出发，在诊疗辅助和科研辅助这两条道路上进行发力，试图用AI赋能，让医生从繁重而耗费精力 的循证过程中解放出来。 另一方面，国务院2025年8月份发布的《关于深入实施"人工智能+"行动的意见》和11月国家卫健委联合五部委发布的《人工智能+医疗卫生实施意见》 中，明确提出在2027年，AI智能体等应用普及率超过70%，并且在2030年将应用普及率超90%，同时也提出了推动临床智能体应用建设。 这些政策的潜台词是包括医疗在内的多个领域，都将迎来一场"AI协同"的大变革。从2025年到2027年，再到2030年，每一个三年，都将是一次台阶式跃 升。 OpenEvidence的破局点正是构建了从数据源头到结论输出的全链路信任体系，这恰恰是之前中国医疗AI大模型较为稀缺的能力。这一实践直指中国医疗AI 的核心痛点，数据量足但质差并且权威缺失。中国虽拥有全球最庞 ...

Group 1: Manus 1.6 Release - Manus 1.6 Max has transitioned from an "auxiliary tool" to an "independent contractor," resulting in a 19.2% increase in user satisfaction, capable of independently completing complex Excel financial modeling and data analysis [1] - New mobile development features support end-to-end app development processes, allowing users to generate runnable iOS and Android applications simply by describing their needs [1] - The introduction of Design View allows for localized image editing, precise text rendering, and multi-layer composition, addressing the uncontrollable issues of AI-generated images [1] Group 2: OpenAI Circuit-Sparsity Model - OpenAI has released the Circuit-Sparsity model with only 0.4 billion parameters, enforcing 99.9% of weights to be zero, retaining only 0.1% non-zero weights, which addresses model interpretability issues [2] - The sparse model forms a compact and readable "circuit," reducing the scale by 16 times compared to dense models, although it operates 100 to 1000 times slower [2] - The research team proposed a "bridge network" solution to insert encoder-decoder pairs between sparse and dense models, enabling interpretable behavior editing of existing large models [2] Group 3: Thinking Machines Product Update - Thinking Machines, founded by former OpenAI CTO Mira Murati, has opened access to its Tinker product, an API for developers to fine-tune language models [3] - The update includes support for Kimi K2 Thinking fine-tuning (designed for long-chain reasoning) and Qwen3-VL visual input (available in 30B and 235B models) [3] - A new inference interface compatible with OpenAI API has been introduced, allowing users to easily integrate with any platform that supports OpenAI API, simplifying the post-training process for LLMs [3] Group 4: NotebookLM Integration with Gemini - NotebookLM has officially integrated with the Gemini system, allowing users to add NotebookLM notes as data sources for Q&A within Gemini conversations [4] - Gemini acts as a "hub" connecting multiple NotebookLM notes, resolving the issue of NotebookLM not supporting notebook merging, enabling simultaneous queries across multiple notes [4] - The content from NotebookLM can now be used alongside online information, facilitating a mixed analysis of "personal data + global information," integrating into Google's core AI product line [4] Group 5: Tongyi's Model Releases - Tongyi Bailing has upgraded the Fun-CosyVoice3 model, reducing initial latency by 50% and doubling the accuracy of mixed Chinese-English recognition, supporting 9 languages and 18 dialects for cross-lingual cloning and emotional control [5] - The Fun-ASR model achieves a 93% accuracy rate in noisy environments, supports lyrics and rap recognition, and covers 31 languages for free mixing, with the initial word latency reduced to 160ms [5][6] - The open-source Fun-CosyVoice3-0.5B provides zero-shot voice cloning capabilities, while the lightweight Fun-ASR-Nano-0.8B version offers lower inference costs [6] Group 6: Zoom's AI Claims - Zoom claims to have achieved a score of 48.1% on the "Human Last Exam" HLE benchmark, surpassing Google Gemini 3 Pro's score of 45.8% by 2.3 percentage points [7] - The company employs a "federated AI approach," combining its small language model with both open-source and closed-source models from OpenAI, Anthropic, and Google, using a Z-scorer scoring system for output selection [7] - This score has not appeared on the official HLE leaderboard, and on the same day, Sup AI announced a score of 52.15%, indicating Zoom's ambition to become the AI hub in enterprise workflows [7] Group 7: Gemini 3's CFA Exam Performance - Recent research indicates that reasoning models have passed all levels of the CFA exam, with Gemini 3.0 Pro achieving a historic high of 97.6% on Level 1 and GPT-5 leading Level 2 with 94.3% [8] - In Level 3, Gemini 2.5 Pro scored 86.4% on multiple-choice questions, while Gemini 3.0 Pro reached 92.0% on open-ended questions, showing significant improvement from previous years [8] - Experts caution that passing exams does not equate to practical capability, noting that AI struggles with ethical questions and cannot replace analysts' strategic thinking and client communication [8] Group 8: OpenEvidence Valuation Surge - OpenEvidence is undergoing a $250 million equity financing round, with a post-money valuation reaching $12 billion, doubling from its previous round two months ago [9] - The company generates revenue by selling advertising space for chatbots to pharmaceutical companies, with an annual advertising income of approximately $150 million, tripling since August, and a gross margin exceeding 90% [9] - An OffCall survey indicates that about 45% of U.S. doctors use OpenEvidence, answering approximately 20 million questions monthly, with its medical journal information being more accurate than general chatbots [9] Group 9: OpenAI's Sora Development Insights - OpenAI's development of the Android version of Sora was completed in just 28 days by a team of 4 engineers collaborating with the AI agent Codex, consuming around 5 billion tokens, with approximately 85% of the code generated by AI [10] - The team utilized an "exploration-validation-federation" workflow, allowing Codex to handle heavy coding tasks while engineers focused on architecture, user experience, and quality control, achieving a 99.9% crash-free rate [10] - Codex is responsible for 70% of OpenAI's internal PR weekly, capable of monitoring its training process and handling user feedback, creating a self-evolving model of "AI iterating AI" [10]

Core Viewpoint - The article discusses the alarming decline in critical thinking and cognitive abilities among the population, particularly in the context of the rise of artificial intelligence, which is predicted to significantly impact the job market within the next 18 months [1][2][18]. Group 1: Impact of AI on Employment - Predictions suggest that by the summer of 2027, AI capabilities will explode, potentially eliminating up to half of entry-level white-collar jobs [1]. - The article emphasizes that the real crisis lies not in AI taking jobs, but in the decline of human cognitive abilities as individuals outsource their thinking to machines [2][18]. Group 2: Decline in Writing and Reading Skills - A significant concern is the decline in writing skills, as many students are using AI to complete assignments, leading to a generation of graduates who may lack essential writing abilities [3][4]. - The average reading scores in the U.S. have reached a 32-year low, indicating a broader decline in literacy and comprehension skills [7]. - The fragmentation of reading habits, with a decrease in leisure reading by nearly 50% since the early 2000s, reflects a troubling trend in cognitive engagement [8][11]. Group 3: Consequences of Cognitive Decline - The decline in writing and reading is seen as detrimental to deep thinking, which is essential for modern economic practices [12]. - The article warns that the over-reliance on AI tools may erode independent thinking skills, particularly in fields like medicine, where students may rely on AI for diagnosis rather than developing their own analytical skills [13][14]. - The potential societal implications include a loss of critical thinking skills, which could lead to a populace that is more susceptible to manipulation and less capable of informed decision-making [15][18].

Core Insights - Baichuan Intelligent has launched the Baichuan-M2 Plus, an enhanced medical large model, which significantly reduces the hallucination rate compared to general models and outperforms the popular US medical product OpenEvidence, achieving a credibility level comparable to experienced clinical doctors [2][3]. Group 1: Product Performance - The M2 Plus achieved a remarkable score of 97 on the USMLE, matching the performance of GPT-5 and surpassing the average human test-taker score, showcasing its world-class clinical problem-solving capabilities [4]. - In the Chinese Medical Licensing Examination, M2 Plus scored 568, far exceeding the passing score of 360 and ranking first among mainstream large models [5][6]. - The model also scored 282 in the Chinese Master's Degree Entrance Examination for Clinical Medicine, demonstrating its advanced understanding of complex medical knowledge [6]. Group 2: Market Position and Usage - OpenEvidence has registered 40% of US doctors for clinical use, with a monthly consultation volume of 16.5 million, indicating a strong market presence [2]. - Baichuan-M2 Plus is positioned as a "doctor version of ChatGPT," facilitating clinical decision-making and addressing the challenges posed by patients using models like DeepSeek for self-diagnosis [7]. - The model's API allows integration into various medical services, enhancing the professionalism of AI in healthcare [8].

Core Viewpoint - Baichuan Intelligent has launched the Baichuan-M2 Plus, an evidence-enhanced medical large model, which significantly reduces the hallucination rate compared to general models, achieving credibility comparable to experienced clinical doctors [3][15]. Group 1: Product Launch and Features - Baichuan-M2 Plus is an upgrade from the previously open-sourced Baichuan-M2, featuring a significant reduction in hallucination rates, outperforming both DeepSeek and OpenEvidence [3][4]. - The model introduces a six-source evidence reasoning (EAR) paradigm, making it suitable for clinical decision support in various healthcare environments, including China, the US, Japan, and the UK [4][22]. - The model's architecture is designed to ensure that it only uses authoritative medical evidence, avoiding non-professional information from the internet [6][9]. Group 2: Evidence Framework - The six-source evidence framework consists of layers that evolve from original research to real-world feedback, ensuring a comprehensive knowledge system [5][8]. - The layers include original research, evidence reviews, guidelines, practical knowledge, public health education, and regulatory information, creating a robust evidence chain [8][9]. Group 3: Retrieval and Reasoning Mechanisms - M2 Plus employs a PICO framework for structured medical queries, enhancing the precision of evidence retrieval [11][12]. - The model incorporates a "evidence-enhanced training" mechanism that prioritizes citation over speculation, fundamentally changing its response logic [13][15]. - The model's ability to evaluate evidence quality ensures that it prioritizes high-trust information, embedding it seamlessly into its reasoning process [13][15]. Group 4: Performance Metrics - M2 Plus achieved a score of 97 in the USMLE, surpassing the average human score and matching GPT-5, demonstrating its clinical problem-solving capabilities [19][21]. - In the Chinese medical licensing exam, M2 Plus scored 568, significantly higher than the average passing score, showcasing its mastery of clinical guidelines and practices [21]. - The model also performed well in various international medical qualification exams, achieving over 85% accuracy [20][21]. Group 5: Market Position and Applications - Baichuan-M2 Plus is positioned as a "doctor's version of ChatGPT," enhancing the usability of AI in serious medical scenarios [22][23]. - The model is integrated into the Baixiao app, providing a tool for doctors to counteract the challenges posed by general models like DeepSeek [23][24]. - The company aims to continuously improve the applicability of AI in real clinical settings through open-source initiatives and API offerings [24].

Core Insights - The article emphasizes the transformative impact of artificial intelligence (AI) on various sectors, highlighting significant investments and advancements in AI technologies [2][3] - The AI Investment Summit 2025 is set to take place on November 2 at UC Berkeley, aiming to gather leaders from academia, industry, and investment sectors to discuss the future of AI [2][3] Audience Composition - The summit will feature over 150 researchers from fields such as AI, economics, robotics, and cognitive science [8] - More than 150 founders from sectors including healthcare and machine learning will participate [8] - The event will also attract over 400 students from prestigious institutions like UC Berkeley, Stanford, and MIT [8] Featured Speakers - Notable speakers include Konstantine Buhler from Sequoia Capital, Rohit Patel from Meta Superintelligence Labs, and Tianfu Fu from OpenAI [10][11][12] - The lineup includes experts from various leading organizations, such as NVIDIA, Google DeepMind, and BlackRock [21] Summit Agenda - The summit will cover a range of topics, including intelligence infrastructure, AI-native products, and the future of human-AI interaction [23][24] - Discussions will focus on economic and industrial landscapes in the morning, followed by topics like incentive mechanisms and multimodal breakthroughs in the afternoon [22] Ticket Information - Early bird tickets are available at discounted rates, with student tickets priced at $29 and general tickets ranging from $69 to $89 [26][28] - Limited seating is emphasized, encouraging prompt registration to secure attendance [26]

Core Viewpoint - The article discusses the rapid growth and innovative business model of OpenEvidence, a medical AI application that has gained significant traction among U.S. physicians, highlighting its unique approach to providing clinical decision support through AI-driven medical search capabilities [5][10][11]. Group 1: Company Overview - OpenEvidence has reached a valuation of $3.5 billion within three years of its inception, with its user base growing from a few thousand to over 430,000 registered physicians, covering more than 40% of practicing doctors in the U.S. [8][10][24]. - The platform processes approximately 850 million clinical consultations monthly, showcasing its high usage frequency among healthcare professionals [10][11]. Group 2: Problem Solving - OpenEvidence addresses the challenge of rapidly evolving medical knowledge, which doubles every 73 days, by providing a platform that allows doctors to quickly access the latest and most relevant medical evidence [5][7][11]. - The application enables physicians to ask clinical questions in everyday language and receive concise answers with authoritative citations within seconds, significantly reducing the time spent searching for information [13][14]. Group 3: Business Model - The company employs a "freemium + advertising" business model, offering its services for free to verified physicians while generating revenue through targeted advertising from pharmaceutical companies and medical device manufacturers [23][24][25]. - This approach allows OpenEvidence to bypass traditional B2B sales processes in the healthcare industry, facilitating rapid user acquisition and establishing a strong network effect among its users [24][25]. Group 4: Competitive Landscape - OpenEvidence operates in a competitive environment where other AI startups are emerging, such as DynaMed and Hippocratic AI, which also focus on providing accurate clinical decision support tools [32][33]. - The article contrasts OpenEvidence's success with the failure of IBM's Watson Health, emphasizing the importance of practical application and user trust in the medical AI sector [32]. Group 5: Founders and Team - OpenEvidence was co-founded by Daniel Nadler and Zachary Ziegler, both Harvard alumni, with Nadler previously selling his AI company Kensho for approximately $550 million [8][27][30]. - The team includes experts from top institutions, ensuring a strong foundation in both AI technology and medical knowledge [20][27].

Core Insights - OpenEvidence is a rapidly growing AI-driven clinical decision support platform that has gained significant traction among U.S. physicians, with over 100,000 daily users, up from a few thousand just a year ago [2][4] - The platform addresses the challenge of rapidly evolving medical knowledge, allowing doctors to quickly access the latest evidence-based information [2][6] - OpenEvidence's unique business model bypasses traditional healthcare software sales, offering free access to individual doctors and monetizing through targeted advertising [19][20] Company Overview - OpenEvidence was founded by Daniel Nadler and Zachary Ziegler, both Harvard alumni, with Nadler previously selling his financial AI company Kensho for approximately $550 million [22][25] - The platform has achieved a valuation of $3.5 billion within three years of its inception, with significant funding from top venture capital firms [25][28] - OpenEvidence's mission is to organize and expand global medical knowledge, providing verified physicians with quick access to relevant clinical information [6][10] User Engagement - The platform has registered over 430,000 doctors, covering over 40% of practicing physicians in the U.S., with a monthly user growth rate of 65,000 [4][5] - OpenEvidence processes approximately 850,000 clinical inquiries monthly, showcasing its high usage frequency among healthcare professionals [5][10] - The platform's core functionality includes intelligent search and instant Q&A, providing precise answers with authoritative citations in just 5-10 seconds [9][10] Technological Innovation - OpenEvidence utilizes a specialized AI model that avoids the common pitfalls of hallucination by relying on authoritative sources such as FDA and CDC data [12][13] - The platform has integrated advanced features like the DeepConsult AI agent, which can autonomously analyze hundreds of studies and generate comprehensive reports for physicians [10][15] - OpenEvidence is the first AI system to achieve a perfect score on the U.S. Medical Licensing Examination (USMLE), highlighting its advanced capabilities [14] Market Strategy - The company employs a "freemium + advertising" model, similar to early Google, to build a large user base before monetizing through targeted ads [19][20] - OpenEvidence's advertising strategy is designed to maintain trust among users by clearly distinguishing between organic results and advertisements [20] - The platform's approach has created a strong network effect, establishing itself as a standard within the medical community [19][20] Competitive Landscape - OpenEvidence operates in a competitive environment, with emerging startups like DynaMed and Hippocratic AI also focusing on clinical decision support tools [28][29] - The failure of IBM's Watson Health serves as a cautionary tale, emphasizing the importance of practical application and user trust in the success of medical AI solutions [28]