I discussed with Gemini why embodied intelligence is the AI with the potential to penetrate the mass market, and why autonomous driving may be the first embodied AI to penetrate the mass market? I feel that the answer is quite enlightening: 1: To become a 'mass AI product', it must have extremely high practicality Today's "application AI" products are mostly tools that require users to actively use them (such as opening the app, entering prompt words). But what the mass market needs is not tools, but services. 2: The reason why App based AI cannot maximize its capabilities and completely penetrate the mass market is precisely because it relies too heavily on human intervention. This dependency, which we call the 'last mile problem of AI', greatly limits the scalability value of AI. The biggest advantage of embodied AI is its closed-loop capability. The open-loop nature of applying AI means that it cannot autonomously and continuously create value. It only works when actively activated by users, and its value creation is passive and fragmented. 3. The advantages of autonomous driving: the problem domain is relatively concentrated: although the road environment is extremely complex, the car only needs to handle specific tasks (driving) and specific environments (roads). In contrast, humanoid robots have higher complexity and greater variability in handling general tasks (washing dishes, folding clothes, climbing stairs) and general environments (irregular home spaces). By the way, I also asked about the standards of mass consumption: The core requirement is to meet the following three dimensions 1. Penetration rate Definition: The actual proportion of users in the target audience. • Standard: At least reach hundreds of millions of users worldwide, and have a household or individual penetration rate of over 50% in developed economies (or higher in specific core populations). 2. Indispensability Definition: Once a product is lost, it will have a serious impact on daily work or life. Standard: The product must solve a common, fundamental, and repetitive problem to become a new infrastructure. Example: Losing a smartphone will cut off social connections and payments; Losing a car can affect mobility. The current situation of AI: Losing me will reduce efficiency, but life will not stop. 3. Independent Business Value Definition: The product itself can create huge, predictable, and standardized revenue, without relying on advertising or third-party subsidies.