Post

Let me explain in detail. Most people still underestimate Tesla's true future business model. The first layer is FSD. FSD is essentially a software subscription. The biggest difference between software and cars is that selling software to the first user or the millionth user costs almost nothing extra; the marginal cost keeps decreasing while profit margins keep increasing. The second layer is Robotaxi. Every Tesla car sold helps collect real-world driving data; the more data, the faster FSD learns; the more mature FSD becomes, the safer Robotaxi is; the more profitable Robotaxi is, the more people buy Teslas and join the Robotaxi network, generating even more data. This is a self-reinforcing data flywheel. The third layer is the Optimus humanoid robot. When FSD learns how to perceive, make decisions, and avoid obstacles in the real world, this AI capability won’t just stay in cars but can be transferred to robots. Today, companies hire workers and bear costs like wages, social security, and management annually; after the initial purchase of a robot, the remaining costs are mostly electricity, maintenance, and depreciation, plus continuous iterative upgrades. So I believe Tesla’s future sales won’t be about cars or robots. Cars are just data entry points, FSD is a high-margin software subscription, Robotaxi is a platform generating continuous cash flow, and humanoid robots replicate the same AI capabilities across more industries. The real value isn’t in the hardware but in this AI flywheel: once it starts spinning, each additional user and each device sold lowers marginal costs, increases data, and deepens the moat. I think this is where many people underestimate Tesla: it’s not aiming to make one-time hardware sales profits but to build an AI ecosystem that continuously collects software subscription fees, platform commissions, and robot service fees. This is also why many investors are willing to give it a much higher valuation than traditional automakers.