In 10 Years, Data Centers Will Be in Space: The World of 'SpaceX is Coming', Why Google Pays 1 Trillion Monthly and Waits in Line | Author Lee Jae-hoon

Summary

Author Lee Jae-hoon discusses SpaceX's stock as a long-term, high-risk play on space and AI convergence. He breaks down the company's cash cow Starlink (1% penetration), defense Starshield, and AI data center leasing revenue, while highlighting Starship's cost-reduction potential as key to unlocking a 28 trillion dollar market by 2035. He cautions about short-term downside risks due to volatility and AI bubble fears.

• SpaceX is positioned as a decade-long bet, not a short-term trade.
• Starlink offers massive growth with only 1% of the addressable market penetrated.
• Defense variant Starshield enjoys 5x price power due to lack of alternatives.
• AI data center leasing (Colossus) generates ~2 trillion won monthly from Google and Anthropic.
• Starship V3 aims to cut launch costs from $2,000-$3,000/kg to $200-$300/kg, making orbital AI data centers viable.
• Orbital data centers solve ground issues (power, regulation) and could reach 1GW with 8,000 satellites.
• Full commercial deployment of space data centers still a decade away, but feasibility is becoming clearer.