Can "rocket-level" engineering capabilities achieve a dimensionality reduction strike? Musk claims: SpaceX holds advantages in the AI field that Google and other companies cannot reach

The Zhitong Finance APP noted that in the first quarterly report released by SpaceX (SPCX.US) as a publicly listed company, a particularly striking point is its massive investment in building new AI computing power. The company's capital expenditure related to AI doubled from the previous quarter to $15.8 billion, and management indicated that it expects to maintain a similar level of spending by the end of this year.

Indeed, compared to the investments in data centers by ultra-large cloud giants like Amazon (AMZN.US), Alphabet (GOOGL.US), and Microsoft (MSFT.US), this capital expenditure still pales in comparison. Just in the last quarter, each of these three companies spent between $41 billion and $55 billion. However, they also generated extremely large revenues from their respective cloud computing divisions and have a vast amount of signed but unfulfilled business.

However, SpaceX CEO Elon Musk believes that thanks to a unique competitive advantage, his company can deploy capital more effectively than those cloud giants.

SpaceX's Advantages in AI

Musk argues that in the competition within the cloud computing sector, SpaceX possesses engineering advantages that surpass all others. His premise is that the company can leverage its engineering talent pool and intellectual property accumulated to efficiently build new data centers that can generate high returns on investment.

During the company's first earnings call, Musk stated, "We found that even applying a small fraction of the experience we gained from building rockets (which is extremely difficult) to data centers can yield tremendous benefits," he specifically pointed out that the company's cooling systems far exceed current requirements.

Significant Return Potential

CFO Bret Johnson stated in prepared remarks, "The current economic benefits have already translated into an investment payback period of less than a year for our new computing power capital deployment."

This sharply contrasts with the statements made by Amazon CEO Andy Jassy, who outlined the economic benefits of Amazon Web Services (AWS) large-scale expansion.

He explained, "The capital investment in data centers began two years ago when we could start putting in servers and monetizing," which is merely the physical limitations faced in building new data centers; it couldn't even start monetizing within two years, let alone achieve breakeven within a year.

"For servers and network equipment, it typically takes just under three years to break even on these investments," Jassy added. Even if the construction costs of data centers were zero, Jassy stated that just purchasing the equipment for these buildings would take nearly three years to break even. The results published by Microsoft and Alphabet also indicate that their operations have a similar timeline.

There is a significant gap between SpaceX's accounting treatment and that of Amazon. Can a company like SpaceX, which has only recently entered the cloud computing field, really be so much more efficient in engineering and capital deployment that it can generate three times the performance of the cloud giants?

Investors are skeptical of Musk's claims Elon Musk never shies away from making bold assertions about the direction of his business and when various milestones will be achieved. However, the actual results often fall short of his predictions. We have reason to doubt whether SpaceX, relying solely on engineering talent, has the capability to establish a substantial competitive advantage in the field of AI computing power.

While Johnson's claim about the current rapid investment payback period may be accurate, it remains unclear whether it can scale efficiently. SpaceX has existing infrastructure that was used last quarter to provide additional computing power to third-party clients such as Anthropic and Google, a subsidiary of Alphabet. This may not be replicable.

More importantly, it is logically difficult for SpaceX to easily retain such top talent in the face of competition from cloud giants. As mentioned earlier, Alphabet, Amazon, and Microsoft invest three times more in capital expenditures than SpaceX. Their businesses heavily rely on the efficient returns of that capital expenditure. If they can multiply their capital return rates, they will undoubtedly pay top talent high salaries.

SpaceX is merely in a position to provide some computing power during a severe shortage of AI computing resources. Alphabet signed a computing power contract with the company because the long-term potential of leveraging its infrastructure to immediately serve large clients like Anthropic by offloading some internal AI computing needs to third parties is too good to pass up.

Once Alphabet builds sufficient capacity for itself, it can terminate its contract with SpaceX. This may bring some excellent short-term revenue to SpaceX, but it does not indicate that it has a long-term competitive advantage in the cloud computing field