Core Viewpoint - Tesla stock has reached a new all-time high of $488, with predictions suggesting it could hit $500 soon, potentially within weeks [1][2]. Group 1: Stock Performance and Predictions - Tesla shares were trading above $490 in pre-market as of December 17, indicating strong market interest [2]. - The AI model ChatGPT-5.2 predicts that Tesla stock could reach $500 in the next two to six weeks if market conditions remain favorable, suggesting a timeframe of late January to mid-February 2026 for this milestone [6]. - Mizuho Securities upgraded Tesla's 12-month price target from $475 to $530, while other analysts like Morgan Stanley and Goldman Sachs maintain lower targets of $425 and $400, indicating potential downsides of 13.24% and 18.35% respectively [8]. Group 2: Catalysts and Market Sentiment - Key catalysts for the optimistic outlook include advancements in Robotaxi testing and CEO Elon Musk's focus on technological trends, particularly in autonomous driving and AI [3]. - Despite the bullish sentiment, there are concerns regarding investor sentiment and market momentum, with some analysts arguing that the stock may be overpriced based on fundamentals such as earnings and demand for electric vehicles, which is slowing in Europe [4].

Core Insights - Innoviz Technologies will showcase its new InnovizThree LiDAR, along with the InnovizTwo LiDAR platform and InnovizSMART, at CES 2026 from January 6 to 9 in Las Vegas [1][2] Group 1: Product Offerings - The InnovizThree LiDAR features behind-the-windshield placement for seamless vehicle integration and superior performance [7] - The InnovizTwo LiDAR platform includes short- to mid-range capabilities for urban maneuvering and long-range capabilities for highway autonomy, with innovative blockage solutions for reliable performance [7] - InnovizSMART is designed for smart applications, integrated with NVIDIA Jetson Orin for real-time edge AI processing [7] Group 2: Strategic Partnerships and Vision - The recent partnership with Daimler Truck and Torc Robotics underscores Innoviz's commitment to providing top-tier LiDAR solutions for autonomous driving and smart applications [2] - The company aims to set high standards for safety and performance in the automotive industry, contributing to the development of safe autonomous vehicles [6] Group 3: Engagement Opportunities - Attendees at CES 2026 can meet Innoviz's management team to gain insights into the company's vision and upcoming projects [4] - Live demonstrations of Innoviz's LiDAR technology will be available at booth 7318, along with opportunities for pre-scheduled test drives [3][7]

Core Insights - Pony AI is focused on developing autonomous driving technology for various commercial applications, indicating a strong market positioning in the growing field of autonomous vehicles [1] Company Overview - Pony AI is currently experiencing a hypergrowth phase, suggesting significant expansion plans and potential for increased market share in the autonomous driving sector [1] Investment Perspective - The company is part of a broader trend in technology and financial markets, particularly in AI infrastructure and digital transformation, which may present investment opportunities [1]

Group 1 - Waymo, the autonomous driving unit of Alphabet, is in discussions to raise over $15 billion [2] - The financing round is expected to value Waymo at nearly $100 billion [2] - The funding round is being led by Alphabet, Waymo's parent company [2]

12月17日，小鹏汽车董事长何小鹏在朋友圈发文谈到了机器人和AI泡沫的话题。他指出，人形机器人 将来会是巨头的竞争，而不同的专用机器人则会有大量不同领域的选手，且会有非常多的成功机会。他 认为，当前没有AI泡沫，未来AI的市场有着巨大机遇。 以下为全文： 关于美国的新创业和机器人 感觉美国在AI，生物，金融方面的创业多。而AI方面的创业，从硅谷看，在SaaS和物理AI机器人领域 特别多，估值也特别高，我这次聊了其中一半都是做机器人创业的朋友，非常集中。中国的机器人公司 很多是从关节和控制入手，而美国的机器人公司很多是从模型入手。我认为人形机器人将来会是巨头的 竞争，而不同的专用机器人则会有大量不同领域的选手，且会有非常多的成功机会。 关于物理AI 维特根斯坦曾经说过，"语言即世界"。语言是知识的总结，信息的传递，但某种程度也束缚了对世界的 想象。很多事情靠语言很难想象和学习，例如小朋友如何学习行走，如何学习游泳，主要都必须依靠对 世界的观察，模仿和强化。所以，大家都在期望从语言大模型，走向多模态甚至世界模型。我认为未来 3年，最有可能有大突破的不一定是在数字世界（可以很明显看到OpenAI虽然核心在说AGI， ...

点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近30个 方向 学习 路线 >>自动驾驶前沿信息获取 → 自动驾驶之心知识星球 现有自动驾驶系统非常依赖车载传感器进行实时精确的环境感知。然而，这种模式受行驶过程中的感知范围限制，在视野受限、遮挡或黑 暗、降雨等极端条件下常出现性能失效。相比之下，人类驾驶员即使在能见度不佳的情况下，仍能回忆起道路结构。为了让模型具备这种"回 忆"能力，针对这个特点， 复旦可信具身智能和上交等合作 ，将离线检索的地理图像作为额外输入引入系统。这些图像可从离线缓存（如谷 歌地图或已存储的自动驾驶数据集）中轻松获取，无需额外传感器，是现有自动驾驶任务的即插即用型扩展方案。 在实验中，首先通过谷歌地图API检索地理图像，扩展了nuScenes数据集，并将新数据与自车轨迹对齐。并在五个核心自动驾驶任务上建立了 基准：目标检测、在线建图、占用预测、端到端规划和生成式世界模型。其中在线建图mAP提升13.4%，占用预测静态类mIoU +2.57%，夜间 规划碰撞率从0.55%降至0.48%，为复杂场景自动驾驶提供低成本、高鲁棒的感知增强方案。大量实验表明，该扩展模态 ...

Core Viewpoint - A California Administrative Law Judge has determined that Tesla's marketing of its Autopilot and full self-driving systems is misleading, leading to potential punitive actions against the company in California [1][4]. Group 1: Regulatory Actions - The California Department of Motor Vehicles (DMV) has recommended a 30-day suspension of Tesla's license to sell and manufacture vehicles in the state [2]. - Judge Juliet Cox's proposed decision, which addresses whether Tesla misled consumers regarding its autonomous driving features, is based on a lawsuit from the California DMV [3]. - The DMV has opted not to pursue the suspension of Tesla's manufacturing license and has instead imposed a 90-day stay on the suspension of its sales license, allowing Tesla time to adjust its marketing practices [5]. Group 2: Marketing and Legal Findings - The state regulator found Tesla's branding of its driver assistance technologies as "Full Self-Driving" and "Autopilot" to be misleading, with claims that the FSD system could conduct trips without driver intervention [4]. - Tesla's legal team acknowledged during hearings that vehicles with Full Self-Driving capabilities are not currently able to drive themselves [4]. Group 3: Market Reaction - Despite the regulatory challenges, Tesla shares increased by 3.07% to a record high of $489.88 per share, although they experienced a slight decline of 0.89% overnight [6]. - There is growing optimism regarding Tesla's autonomous driving features, as CEO Elon Musk announced the testing of robotaxis in Austin without a human safety monitor [6].

When you look at physical AI plays, there's Nvidia and Tesla's. I think the two best out there. And now it's about proving it out.I'd call it a make or break year. You need to be in 30 cities. I mean, when so when you look at Austin's and now what we're seeing in California and some other cities, you need to be in 30 cities by some with the robo taxi >> with robo taxi.So So that's one. >> Two, new safety driver. Now you're starting to see that now in Austin.We saw it firsthand. You need to see the geoence a ...

点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近30个 方向 学习 路线 >>自动驾驶前沿信息获取 → 自动驾驶之心知识星球 论文作者 | Xiaosong Jia等 编辑 | 自动驾驶之心 自动驾驶中的4D场景重建是实现环境感知与运动规划的关键环节，然而传统视觉几何模型在多相机、低重叠的自动驾驶场景中往往表现不佳。 来自上海交大、复旦等机构的研究者提出 DriveVGGT，一种专为自动驾驶设计的视觉几何Transformer，通过显式引入相机相对位姿先验，显著提升了多相机系统的几 何预测一致性与推理效率。 更多自动驾驶的行业信息、技术进展，欢迎加入自动驾驶之心知识星球获取！ 背景介绍 4D重建是一项从视觉传感器预测几何信息的计算机视觉任务。与其他传感器相比，基于相机的重建因其低成本而在各个领域，尤其是在自动驾驶和机器人学中，得到 了广泛的研究和应用。通常，重建方法有两种类型。第一种是基于迭代的方法，例如。这些方法需要选择特定的场景或物体，并通过迭代重建来获得优化结果。然 而，由于泛化能力不足，当场景或物体发生变化或修改时，基于迭代的方法需要重新训练模型。第二种是前向方法。这些方法 ...

Core Viewpoint - Alphabet's autonomous driving subsidiary Waymo is negotiating with potential investors to raise funds with a valuation of at least $100 billion [1] Group 1: Funding Details - The funding round is expected to be at least several billion dollars, potentially exceeding $10 billion [1] - The exact valuation is not yet determined, and the funding round is anticipated to be arranged in early next year [1]