SoftBank founder and CEO Masayoshi Son has publicly questioned the viability of orbital data centers, arguing that the cost and timeline involved make them irrelevant to the AI industry’s immediate compute needs. Son made the remarks at a recent shareholder meeting, contending that the next few years of the AI race matter far more than infrastructure that could take a decade or more to deliver meaningful capacity.
The comments carry notable irony given SoftBank’s long track record of high-risk bets on unproven technologies, a point raised by TechCrunch’s Kirsten Korosec, who noted that Son’s skepticism is significant precisely because of his profile, even as his firm remains heavily invested in terrestrial data center projects.
Sean O’Kane argued that SpaceX’s orbital data center ambitions serve a dual commercial purpose, as building a satellite constellation to host compute would simultaneously generate guaranteed launch revenue for SpaceX’s own rocket business, which currently depends heavily on Starlink for its dominant market share.
Anthony Ha observed that the broader debate is complicated by the fact that every major figure weighing in, from Elon Musk to Sam Altman to Son, has significant financial interests shaping their predictions. Musk’s orbital vision benefits SpaceX directly, Son’s skepticism aligns with SoftBank’s terrestrial investments, and Altman’s dismissiveness reflects his complicated history with Musk personally.
The underlying driver of the debate remains the AI industry’s acute compute shortage, which has pushed investors and founders toward increasingly unconventional infrastructure solutions as land-based data center development faces growing constraints.
