"That's not how previous wars worked" would be a more accurate take. Considering how few aircraft there are in modern militaries and how long it takes to manufacture modern fighter aircraft and missiles, one side will be out of high tech aircraft/missiles before the other and then it's game over.

This is one of those things I kind of wonder about the reality of. It's absolutely true that you're unlikely to be able to manufacture a lot of what you'll use during the war.

But simultaneously, it's not like the United States is really trying to fix this problem. The modern generation fighter aircraft are boondoggles with massively inefficient manufacturing designed to win senators and congresspeople votes - not act as the backbone of a hardened supply chain.

They're being built the way they are because the US doesn't seriously expect to face any enemy who can substantially deplete its forces, or is going to have any ability to strike the mainland.

My suspicion is that if there was an actual belief in the need to have reliable supply lines, it could be done and manufacturing could be streamlined a lot. You could definitely imagine a nearly fully automated factory rolling these things off very quickly.

> "My suspicion is that if there was an actual belief in the need to have reliable supply lines, it could be done and manufacturing could be streamlined a lot. You could definitely imagine a nearly fully automated factory rolling these things off very quickly."

Unlike previous wars, cruise missiles are a mature technology, making it much easier to precisely zap an opponent's manufacturing facility and infrastructure than before. High tech gizmos also have a lot more dependencies that go into them, making it easier to disrupt by knocking out a few or even just one of them.

In case of a prolonged conflict you need to consider the AI and it's programmers as a joint system. The AI by itself may be exploitable, but the programmers will adjust it to counter anti-ai strategies.

We don't program what are called AI's nowadays. We curate data, sometimes supervise the learning process, and do our best to detect when a model is bad.

If you think there's any intention or well defined predictability to these things, I've got a bridge to sell ya.