$1M a shot seems like a lot until you look at what goes into each one, how much development costs, and the environment it is designed for. You don’t need corruption to make these expensive. Also, the are based on 30 year old technology. Rather than being cheaper than new, they tend to be more expensive due to requiring miniature mechanisms which were cutting edge in 199x. And there are no other uses for them in their hardened configuration so you don’t get economy of scale of something like a commercial ship or board. Plus there’s a lot of paperwork and you can’t sell a single piece unless the government approves it. This is so far from a consumer item there isn’t even a way to compare costs.

FWIW, corruption would cost roughly the same price but you’d receive a missile with shoddy components and forged paperwork. Just look at Russia’s army to see the difference in munition reliability.

Cheapest city with moderately decent public transit is probably Washington DC. With an average home price comparable to the one I live in without public transit of about $600,000 more than my current home. Even if I didn’t own my truck outright (8 years old, 58k miles) and the price of gasoline doubled, my payback period for 100% free public transit is greater than infinity with a 5% cost of money calculated in.

It’s a bit like solar. I’ve run the numbers, and had others run the numbers, and the conclusion is that it would require replacing solar panels twice before I made back my investment, even with a 0% loan for the panels and install.

I’d love to be part of it. I’d love to have European-style public transit. Even in the few places where viable public transit exists in the US, it’s not affordable to move to those places. shrug

If only I could afford to live near transit…

It’s not code. It’s a matrix of associative conditions. And, specifically, it’s not a fixed set of associations but a sort of n-dimensional surface of probabilities. Your prompt is a starting vector that intersects that n-dimensional surface with a complex path which can then be altered by the data it intersects. It’s like trying to predict or undo the rainbow of colors created by an oil film on water, but in thousands or millions of directions more in complexity.

The complexity isn’t in understanding it, it’s in the inherent randomness of association. Because the “code” can interact and change based on this quasi-randomness (essentially random for a large enough learned library) there is no 1:1 output to input. It’s been trained somewhat how humans learn. You can take two humans with the same base level of knowledge and get two slightly different answers to identical questions. In fact, for most humans, you’ll never get exactly the same answer to anything from a single human more than simplest of questions. Now realize that this fake human has been trained not just on Rembrandt and Banksy, Jane Austin and Isaac Asimov, but PoopyButtLice on 4chan and the Daily Record and you can see how it’s not possible to wrangle some sort of input:output logic as if it were “code”.

I traveled from the east coast last fall (13 hour time change) and spent a week moving my day/night cycle by an hour or two a day. My wife thought I was crazy but I got a good 8+ hours of jet lag out of the way before leaving. The first day was rough, but after that it was manageable.