Masayoshi Son, the founder and CEO of SoftBank, has publicly pushed back against Elon Musk's concept of placing data centers in orbit, arguing instead that the future of artificial intelligence belongs on the ground. The comments, made during a recent discussion on AI infrastructure, mark a strategic divergence between two of tech's most influential figures.
Son's stance is clear: terrestrial data centers are the more practical and scalable path for the massive computing needs of AI. He didn't mince words, calling for a redirection of energy and capital toward building out land-based facilities rather than chasing space-based alternatives. The challenge comes as both companies—SoftBank through its Vision Fund and Musk through his ventures SpaceX and Tesla—jockey for position in the AI arms race.
The Ground-Based Argument
Son's advocacy for terrestrial infrastructure isn't new, but his direct opposition to Musk's orbital vision is. He argues that the physics and economics of space make it a poor fit for the dense, low-latency computing AI demands. Data centers on Earth can be built faster, upgraded easier, and connected to power grids without the hurdles of rocket launches and orbital maintenance.
SoftBank has been pouring billions into AI, from chip design to data center construction. Son sees the ground as the logical home for the compute clusters that will train and run tomorrow's models. “We need to maximize ground-based AI infrastructure,” he said, according to people familiar with his remarks. The message is a direct rebuttal to Musk's suggestion that space could host the backbone of AI processing.
Musk's Orbital Idea
Musk has long floated the idea of placing data centers in orbit, leveraging SpaceX's Starlink network and Starship rocket to create a space-based computing layer. The vision is grand: satellites or space stations hosting servers that could bypass terrestrial constraints like land costs, cooling water, and grid limitations. But critics point to the astronomical expense, the latency of space-to-ground links, and the difficulty of repairs.
Son's critique zeroes in on those weaknesses. He didn't dismiss Musk's ambition but questioned whether it's the right priority for the immediate AI boom. The subtext is clear: while Musk dreams of Mars and orbital factories, Son wants to win the AI race on Earth, right now.
What This Means for SoftBank
SoftBank is one of the world's largest tech investors, with a track record of placing big bets on infrastructure. Son's pivot toward terrestrial data centers could signal a shift in where the Vision Fund allocates its capital. Instead of backing space-based computing startups, SoftBank may double down on land-based AI projects, from fiber networks to massive server farms.
The company already holds stakes in Arm, the chip architecture powering many AI processors, and has invested in data center operators. A public clash with Musk could also be a positioning move—staking out a clear technological path to attract partners and customers who share Son's view.
The debate isn't academic. Both men have the resources and influence to shape where billions of dollars flow. Musk's SpaceX continues to launch satellites; SoftBank's Son is signaling that the real action is on the ground. For now, the battle lines are drawn: space versus earth for the future of AI computing. Which vision wins will depend on engineering, economics, and the clock—because AI waits for no one.
