The three business segments in SpaceX's prospectus are Starlink, which is very profitable, and Rocket, which appears to be losing money but actually makes a lot of money (mainly due to Starship's massive research and development). As long as Starship is successfully commercialized by the end of the year, the second business will also shift from surface losses to profits (considering the huge success of Falcon 9 in 25 years, accounting for half of the global launch volume, we can expect the success of Starship's business). The main culprit causing SpaceX's current losses is still the xAi business, which was initially only for its own use, so it did not consider building a cloud computing center for external rental. But its self use effect is also very average. Buying 550000 high-end NVIDIA GPUs is the most among all large model enterprises, but its utilization efficiency is the lowest among all large model enterprises. Nowadays, big model companies are considering how to make money and monetize, but with so many giants, it won't be a win-win situation. There will always be lagging giants who lose, so this money burning business needs to continue. It has now leased 220000 high-end graphics cards to Anthropic for $1.25 billion per month, with a lease period of 29 years until May, recovering some of its losses. But it also deeply binds SpaceX's stock price to its AI business. If the long-term burning of money does not yield the expected return, it can still have a significant impact on the stock price. Additionally, there is potential room for future stock prices. According to the previous plan of the board of directors to reward a huge market value of $7.5 trillion, if the market value after going public approaches $2 trillion, the potential space will be 3-4 times. Especially if predicted by Wood Sister's ARK Fund, SpaceX's valuation will reach $7.5 trillion by 2035. That is to say, the potential return is 4 times that of 9 years. At least for retail investors, it's not a sexy investment. So the emotions after going public can be hyped up in the short term, but it's not cost-effective to invest for three to five years as a long-term investment. The above is my personal judgment. If you have different opinions, please feel free to discuss.