The new AI space race is defined by two, directly opposing environmental claims. In one corner, startup Starcloud claims a “10 times carbon dioxide savings.” In the other, astronomers object that the “rising numbers of satellites” are “like bugs on a windshield.”
This “great dilemma” is at the heart of Google’s “Project Suncatcher.” The project is a “moonshot” to solve the $3 trillion, carbon-intensive datacentre problem on Earth. The “10x CO2 saving” is the “pro” argument, based on 8x-more-efficient solar power.
The “bugs on a windshield” complaint is the “con.” It argues that “clean” AI for Earth is coming at the cost of “polluting” the night sky, a vital scientific and cultural resource. This, combined with the “hundreds of tonnes of CO2” from each launch, creates a complex “cost-benefit” analysis.
Google, Elon Musk, and Nvidia/Starcloud are all caught in this dilemma. They are promoting the “10x saving” narrative while trying to manage the “bugs” narrative.
This is a new kind of environmental debate. It’s not about land vs. industry, but about sky vs. data. As Google’s 2027 prototypes get closer, this dilemma will only become more intense.