If I'm a wage-paid employee in a large shop (or any organization), then I'm looking to a) not get fired b) get a raise c) get a promotion. The bigger and more bureaucratic the organization, the more personnel decisions tend to become disconnected from ground truths and the more it becomes possible to rise through politics, whether or not the shop gets better. The people at the top, e.g. the owners of Walmart, want their shops to be more efficient, but if line worker #529 doesn't think he'll be paid more for making his local shop 5% more efficient, then he won't- and if there are more reliable ways up, he'll take those.

When a government owns/runs the shop, even the people at the top don't care about it making money- it's not their money to lose! The shop is probably useful as a political tool/division point, though. Imagine left-right debates about whether the government store should stock condoms, or meat, or whatever.

The disagreement here seems to over whether Cloudflare is more like a restaurant, or more like a utility company. Given the ubiquity of DDOS attacks online, (and the fact that KF seemingly can't stay online without DDOS mitigation), I think there's a good argument to be made that DDOS mitigation might be more like a utility.

The 'bad guy' is whoever is the aggressor. Everyone has a right to life; if someone starts shooting at you, or swinging a bat at you, you have the right to defend yourself, and if you kill the aggressor in self-defence, so be it.

I've heard from an internet lawyer that it's possible in principle for two people attacking each other to both have grounds for self-defence, too, if both were in reasonable fear for their lives. Whether A kills B, or B kills A, either may be able to argue self-defence given the right (wrong) circumstances. But that's just some internet theorycrafting I heard.

Eventually, over 10^100 years of operation you get into interesting problems with the number of possible states a mind of a given size has being finite; but if you can continually add more matter via non-energy-conserving interactions, you could add more memory/processing up to fundamental physical limits: you can only fit so much information in a given volume before it collapses into a black hole; you can only communicate at c, so if you make your mind bigger, it has to parallelize or slow down; Given the expansion of space, at some scale, light signals from one end of your mind can't reach the other (the space inbetween expands faster than C), which would be the the fundamental limit... except that gravity counteracts the expansion of space, and I'm not sure what the expansion properties of a visible-universe-filled volume of just-barely-black-hole-subcritical computronium is. I imagine you're in a big crunch regime at that point, except... Isn't that the same thing as a black hole collapse? Nonetheless, there's presumably some balancing density that perfectly avoids both expansion and contraction, though it's presumably unstable. Ignoring the instability-difficulties, though, that means you can continue growing the amount of memory you have by slowing down your processing speed to one clock cycle every ~2*the diameter of your mind in light-years years. I'm not sure what the ultimate limiting factor is here, given infinite time; does the amount of computation of this system in infinite time go to infinity as your scale goes to infinity, or does it asymptotically approach some finite upper bound?