SoftBank CEO Masayoshi Son has publicly challenged the viability of Elon Musk's orbital data center concept, joining a growing chorus of skepticism regarding space-based AI infrastructure. During a recent SoftBank shareholder meeting, Son argued that deploying data centers in orbit offers negligible cost advantages and fails to address urgent industry needs. He emphasized that the critical window for AI leadership lies within the next few years, rendering long-term orbital projects irrelevant compared to the immediate necessity of expanding terrestrial compute capacity. SoftBank remains heavily invested in Earth-based data center development, reinforcing the strategic preference for ground-level solutions. The critique emerges against a backdrop of severe compute constraints driving a "neo-cloud" trend. Industry analysts observe that companies ranging from specialized chipmakers like Groq to formerly unrelated firms are pivoting to lease excess processing power. While SpaceX has initiated compute leasing deals, experts question the long-term durability of this model. Sean O'Kane of TechCrunch noted that SpaceX's orbital ambitions may primarily function to guarantee recurring revenue for its launch business. By proposing a constellation of satellites requiring regular replacement, Musk effectively creates a self-sustaining demand cycle for SpaceX launches. O'Kane highlighted that SpaceX's dominance in the global launch market is heavily tied to the Starlink constellation; orbital data centers represent another mechanism to expand launch volume and solidify SpaceX's ecosystem. The debate highlights the conflicting financial interests influencing the discourse. Kirsten Korosec noted the irony of SoftBank, known for historically bold investments, adopting a cautious stance. However, industry observers stress that few commentators offer impartial analysis. While Musk promotes orbital infrastructure to benefit SpaceX, Son's skepticism aligns with SoftBank's terrestrial data center portfolio. Sam Altman has also expressed reservations about orbital computing. This convergence of interests suggests that predictions regarding space-based AI are often shaped by the promoters' existing business imperatives rather than objective technical assessment. Practical challenges further undermine the orbital data center proposition. The concept faces significant engineering and economic barriers, including the high costs of deployment and the logistical complexity of maintaining hardware in space. These factors make orbital solutions impractical for resolving current AI bottlenecks, which require rapid, scalable infrastructure. As the industry navigates compute scarcity, the focus remains on accelerating terrestrial expansion, with orbital data centers viewed as a distant prospect rather than a viable near-term strategy.