AI data centres in space raise new environmental questions
Thinking of the future in space
As artificial intelligence drives explosive growth in energy-hungry data centres, technology companies are looking to space as a potential solution to mounting environmental pressures on land, reported The New York Times on 1st January.
Technology companies racing to build artificial intelligence systems are increasingly warning that the environmental limits of Earth may not be enough to sustain the industry’s growth. As demand for computing power surges, some executives and scientists are now promoting an unconventional idea: placing A.I. data centres in space.
The proposal is driven by concern that terrestrial data centres already consume vast amounts of land, electricity and water. These facilities, which house servers that train and run A.I. models, are expanding rapidly across the world. Major technology firms including Meta, OpenAI, Microsoft, Amazon and Google are investing hundreds of billions of dollars in new infrastructure, with OpenAI alone committing an estimated $1.4 trillion to data centre projects.
But these investments are increasingly running into environmental and social resistance. In many regions, data centres face power shortages, rising electricity prices and criticism over heavy water use. Local communities have objected to projects they say worsen water scarcity and strain energy grids.
Against this backdrop, some industry leaders argue that space could offer environmental relief. “The main benefit to building a data center in space is abundant energy, with nearly 24/7 access to the sun and no clouds to obstruct the project’s solar panels,” said Philip Johnston, the chief executive of Starcloud, a space data centre start-up. He added that space also avoids local environmental regulations and public opposition.
Google announced in November that it was working on Project Suncatcher, a space-based data centre initiative that could begin test launches in 2027. Elon Musk has also promoted the idea, saying at a recent conference that space data centres would be the cheapest way to train A.I. “not more than five years from now.”
Supporters say orbiting facilities could rely almost entirely on solar power, reducing pressure on Earth’s energy systems. However, critics argue that the environmental benefits are uncertain. Launching heavy equipment into space is extremely carbon-intensive, and current costs remain high. A kilogram of material costs around $8,000 to launch into orbit, according to Pierre Lionnet, a space economist at Eurospace.
“There are fewer environmental regulations than on Earth,” Mr. Johnston said, but skeptics question whether moving pollution off-planet truly solves environmental harm. “It’s completely nonsensical,” Mr. Lionnet said, referring to claims that massive space-based systems could soon be viable.
There are also technical challenges with environmental implications. Modern computer chips are vulnerable to space radiation, and cooling systems would require large radiator panels rather than air or water. While space is extremely cold, the vacuum makes heat dissipation difficult.
Despite the hurdles, interest continues to grow as environmental limits tighten on Earth. “The hottest thing to invest in right now is A.I., and the second-hottest thing is space,” said Tom Mueller, a former SpaceX executive. “Now they’re converging.”
Whether space data centres reduce environmental damage or simply shift it elsewhere remains an open question, as the A.I. industry searches for ways to sustain its rapid expansion.
The New York Times and Maghrebi.org
Want to chase the pulse of North Africa?
Subscribe to receive our FREE weekly PDF magazine
