种一棵树，最好的时间是十年前，其次是现在。 综合来看，理想决定全面步入AI深水区的动作，呈现出从容布局和"箭在弦上"的双重特征——一方面是软硬件技术积淀成熟所带来的"水到渠成"，另一方 面则是内外部挑战之下的压力倒逼。 万事俱备 从"汽车本身就是机器人"，到将人形机器人提上日程，理想在具身智能赛道的布局早有端倪。 2024年初，机器人领域资深专家廖平平加入理想，负责开发轮式双臂人形机器人，主要用于理想工厂的部分制造环节，但因技术限制、供应链不成熟等原 因，该项目宣告暂停。 去年的Livis眼镜，是理想穿戴部门量产的第一代产品，也是其AI大模型能力在车外空间交互里长出的第一个"样本"，也被视作理想同学走出车外的第一 步。之所以选择AI眼镜这样一个更加轻量化的路径来作为突破口，也隐藏着理想将AI能力向物理世界延伸的一次尝试。 在2024理想AI Talk中，李想曾表示一定会做具身智能，但"节奏不是现在"：L4自动驾驶的汽车还解决不了，怎么去解决更复杂的？ 尽管如今看来或许还没到完美的时机，但理想还是决定踏出这一步。据21世纪经济报道，1月26日，李想召开线上全员会，全程几乎不谈汽车，只谈AI， 强调2026是所有 ...

1 、 很多人都很担心新能源汽车行业会不会光伏化 ， 内卷永无止境 ？ 新能源乘用车20万以下会光伏化 ， 20万以上会和智能手机的格局走势类似 。 这里面的原因有以下几点 ： （ 1 ） 20万以下 ， 智能化在这个价位主流消费者购买决策中排名更靠后 。 这并不是说这个区间的消费者不想要智能化 ， 而是智能化的成本 （ 举个例子包括激光雷达 、 大算力芯片 、 研发投入摊销等等两万人民币 ） 占车价超过10％ ， 这个比例 太高了 ， 这个价位主流的消费者宁愿选择便宜一两万 。 这也是尚界第一款车失败的原因之一 。 因此20万以下的竞争会更接 近光伏化内卷竞争 ， 完全比硬件 、 比价格 ， 没有任何一项投入可以获得边际成本递减效应 。 （ 2 ） 20万以上有很多投入可以获得边际成本递减效应 ， 比如品牌 、 智能驾驶 、 智能座舱 。 消费者买一辆20万以上的车 ， 价格是很重要的考虑因素 ， 但不是唯一重要的了 ， 他要综合考虑品牌 、 体验 、 智能化升级空间等等 。 厂家对品牌建 设 、 AI团队的投入 ， 第一是有门槛的 ， 第二是无形资产 ， 第三在卖车的时候又是消费者重点决策因素 。 这里面 ...

Core Insights - The report discusses the recent market adjustments driven by global liquidity pressures, concerns over an "AI bubble," and the A-share Q4 settlement season, highlighting a potential market recovery following the easing of these uncertainties [1] - It emphasizes the importance of structural adjustments in the upcoming 2025 Central Economic Work Conference, suggesting a focus on expanding domestic demand while acknowledging the need for a solid economic recovery [1] - The report identifies potential investment opportunities in the AI sector, particularly in application areas such as AI healthcare, AI edge computing, intelligent driving, and embodied intelligence, as well as in key industries related to the 14th Five-Year Plan, including commercial aerospace, nuclear power, hydrogen energy, quantum communication, and brain-computer interfaces [1][2] AI Sector Opportunities - AI healthcare shows significant potential for deep applications in pathology detection and pharmaceuticals, with a focus on building high-quality datasets and breaking down data barriers [2] - The report notes the recent release of AI medical models and their rising popularity, indicating a strong consumer demand for accessible healthcare services [2] - AI edge computing is being explored by companies like ByteDance, with new AI phone models being introduced, although challenges related to data privacy and competition remain [3] - The report highlights advancements in AI glasses, which are expected to see increased consumer adoption and order releases in the near future [5] Intelligent Driving - The initiation of L3 autonomous driving trials in China is expected to enhance the value of intelligent driving hardware, with companies like Changan and Xiaopeng leading the way [6] - The report mentions the successful commercialization of Robotaxi models, indicating a positive trend in the intelligent driving sector [6] - It also notes the importance of smart traffic infrastructure as part of the broader "Transportation Power" goals outlined in the 14th Five-Year Plan [6] Embodied Intelligence - The humanoid robot sector is poised for significant developments, with key catalysts including IPOs and product launches from companies like Tesla [7] - The report emphasizes the need for practical applications of humanoid robots in industrial settings to validate their market potential [7] Key Industries from the 14th Five-Year Plan - Commercial aerospace is highlighted as a critical area for national strategy, with ongoing developments in satellite internet and 6G technology [8] - The report discusses the increasing focus on nuclear power and controlled nuclear fusion projects, with significant funding and project announcements expected [11] - Hydrogen energy is noted for its growing economic viability due to decreasing production costs and increasing demand for green energy solutions [12] - Quantum communication is positioned as a strategic area for national competition, with ongoing efforts to standardize and commercialize technologies [13] - Brain-computer interfaces are advancing rapidly, with recent breakthroughs in clinical trials and applications in healthcare [14]