The US gov't has been leaning into LLMs for a while now (really picked up under this administration) and (the useful and practical) potential of large, cheap (relative to current costs) satellite constellations, particularly for communication and surveillance networks. The LLM pushers in the gov't are letting out lots of contracts for development, either to fund entire data centers or to partially fund them for priority use.

This data center in space stuff is a marriage of the two fields that will likely lead to a lot of contracts and little output, but the not at all secret trick to government contracting is you don't have to successfully deliver to make a lot of money on government contracts, you just have to do well enough to avoid severe penalties and blacklisting.

A demonstration of my point in the last paragraph: https://www.spaceforce.mil/News/Article-Display/Article/4465... - Over $6 billion and nothing to show for it. Raytheon will be fine.