詹锟本次人事调整标志着理想具身智能战略是动真格的
理想TOP2·2026-01-14 16:27

Core Viewpoint - The article emphasizes the distinction between two approaches to vehicle development: one centered on autonomous driving and the other on embodied intelligence. The founder of the company has shifted from viewing autonomous driving as the ultimate goal to recognizing embodied intelligence as the core focus, with autonomous driving being a subset of this broader concept [2]. Group 1: Company Strategy - The company plans to build a complete agent system based on VLA, positioning vehicles as the largest embodied intelligence terminals in the coming years [2]. - The company has undergone a management shift, returning to a startup model from a professional manager model, which is deemed more suitable given the current industry dynamics [4]. - The management approach will focus on deeper dialogues, user value, efficiency, and identifying key issues rather than creating information asymmetry [4]. Group 2: Product Development - The future of valuable products in the next decade will be vehicles that possess both autonomous and proactive capabilities, offering premium services akin to first-class airline experiences [5]. - The company aims to eliminate traditional vehicle cockpit systems, focusing instead on a vehicle that integrates autonomous driving, spatial intelligence, and embodied control [3]. Group 3: Technological Advancements - The company identifies a critical need for advancements in perception technology, aiming to enhance effective 3D perception distance from over 100 meters to 200-300 meters, which would significantly improve common issues in assisted driving [6]. - Current computational models are limited, with a 40 billion parameter model running at only 10 Hz. Enhancements in model performance are necessary to address comfort and responsiveness issues in assisted driving [6]. - The response time for human braking and steering is around 450 milliseconds, while the complete chain for autonomous driving is currently at 550 milliseconds. The company aims to reduce this to 350 milliseconds through a line control system, potentially halving accident rates [7].