After a few years the orbit will degrade enough that it'll start to fall back to earth. At that point, the satellite will either burn up completely on re-entry, or partially and the rest will fall to earth.
Either way, each of these satellites will be completely gone from orbit after a few years.
ULA is already a private company. I don't think the US government has done any of their own work to get to space since the shuttle.
If this actually did lead to faster matrix multiplication, then essentially anything that can be done on a GPU would benefit. That definitely could include games, and physics models, along with a bunch of other applications (and yes, also AI stuff).
I'm sure the papers authors know all of that, but somehow along the line the article just became"faster and better AI"
The above post is referencing/quoting a line from the show "It's always sunny in Philadelphia", which is why people up voting it
Well, yes and no.
Quantum computers will likely never beat classical computing on classical algorithms, for exactly the reasons you stated, classical just has too much of a head start.
But there are certain problems with quantum algorithms that are exponentially faster than the classical algorithms. Quantum computers will be better on those problems very quickly, but we are still working on building reliable QCs. Also, we currently don't know very many quantum algorithms with that degree of speedup, so as others have said there isn't many use cases for QCs yet.