谷歌智能体业务主管Omar Shams认为芯片虽然重要，但能源供应才是AI长期发展的关键制约因素。他认为，美国电网 扩容缓慢，而中国每年新增电力产能已超英法总和。他甚至提出了在月球或太空部署太阳能电站的设想，为AI算力提供能 源支撑。 硬·AI 作者 | 鲍奕龙 编辑 | 硬 AI 7月3日，谷歌智能体业务主管Omar Shams在接受密歇根州立大学计算数学教授Steve Hsu专访时，分享 了关于AI产业前沿的解读。 以下是要点： 以下是精华内容： 理论物理学人才的稀缺价值。 Omar Shams以自己从弦理论研究转向AI创业的经历，阐释了物理直觉在AI 研究中的重要作用。损失函数的优化过程类似于能量流形上的球体滚动，信息论中的KL散度与物理学中 的哈密顿量存在直接对应关系。 芯片虽然重要，但能源供应才是AI长期发展的关键制约因素 。他认为，美国电网扩容缓慢，而中国每年 新增电力产能已超英法总和，中美在AI基础设施竞争中的能源差距日益凸显。 他甚至提出了在月球或太 空部署太阳能电站的设想，为AI算力提供能源支撑 。 AI领域不存在秘密，但存在价值连城的隐性知识 。人们通过社交、跳槽等方式，使得各家公司的研究 ...

Core Insights - The legal industry is undergoing a transformative change driven by technological advancements, particularly generative AI, which is reshaping traditional legal services [1][2] - By 2025, AI is expected to transition from a supportive role to a more autonomous function, capable of executing complex tasks independently [2] - The integration of AI into legal workflows is becoming a focal point, emphasizing the need for stability, security, and measurable business value [2] Industry Trends - AI tools like Thomson Reuters' Westlaw Precision and Harvey AI are enhancing legal research and document management, marking a shift from auxiliary tools to core business process drivers [3] - The emergence of AI in legal services is prompting law firms to collaborate closely with legal tech companies to address the supply-demand imbalance in legal services [8] - The legal tech landscape is evolving rapidly, with a significant focus on the integration of AI capabilities into daily legal practices [3][8] Key Events - The "Dawn of Legal AI" forum held in Shanghai gathered over 300 experts and representatives to discuss the integration of AI in legal services and share insights on future developments [3] - Shanghai Zhihui Network Technology Co., Ltd. was recognized as one of the top ten "Pioneering Enterprises in Legal AI" at the forum, highlighting its contributions to the field [4] Leadership Insights - Liu Qiming, co-founder of Zhihui, emphasized the importance of distinguishing between tasks suited for humans and those for machines in the evolving legal landscape [5][7] - The future of law firms will require professionals to possess a keen understanding of the differences between human and machine capabilities, which will be crucial for maintaining competitiveness [7][13] Digital Transformation - Law firms are encouraged to embrace digital transformation, with a focus on developing a "digital brain" that can accurately assess the capabilities of both humans and machines [13] - The digitalization of legal services is seen as an irreversible trend, necessitating a shift in how legal professionals are trained and selected [13][12] User Engagement - Legal tech companies are analyzing user segments, with professional lawyers showing higher engagement with legal tech products, indicating a shift towards machine-delivered legal services [15] - The validation of legal tech products can be effectively tested through trial periods with legal professionals to ensure market fit and usability [18][20]

1、区分客户的定制化需求是只针对其个人还是能够用在其他人身上，对定制化需求敢于说"不"； 2、Scale AI做的很好的一件事就是倾听前沿客户的声音； 3、如果找不到伟大的市场，就会失败。"渠道为王，产品为尊"，前期渠道可以通过蛮力进行早期牵引， 后期需要产品提供实质内容； 4、AI应该直接为用户选择，而不是让用户选择模型； 5、模型公司将不得不更多地成为产品公司 在人工智能领域，不乏快速增长的故事，但在不同的高速增长的AI公司担任产品主管，参与一个又一个独 角兽缔造的故事并不多，阿蒂什·纳亚克（Aatish Nayak）是其中一位。 在生成式AI时代，随着模型成本的下降，模型能力已不是门槛，但用户体验是，这与产品息息相关。 阿蒂什·纳亚克是专注法律领域人工智能初创公司HarveyAI的产品主管，负责产品愿景、战略、设计、分 析、营销和支持。这家获得红杉、OpenAI投资的独角兽，2024年收入增长4倍，ARR接近5000万美金，估 值达到30亿美元。而阿蒂什则是这家公司产品的架构者。 不仅如此，阿蒂什曾经还在人工智能数据标注公司Scale AI担任产品主管，这家公司的规模从40 人扩张到 800 人，如今估 ...