I think the spectrum between where we are now and “existentially threatened wartime economy” is nonlinear: if there’s not an existential threat, but there is instead a popular foreign war (e.g. post 9/11), the fraction of defense economic capacity that gets unlocked is very small—still pretty close to what we have now. Which is a lot, but it’s far from efficient.

OTOH, if the attacker wants a specific objective (control infrastructure, seize a resource, gain political power), that equation changes a ton in favor of the defender.

Lots of modern military strategy seems to be occupied with finding the appropriate balance between those tactics.

Once sufficiently effective long distance military tools (bombs, missiles) could be produced at large scales, the defense equation was fucked. That happened before widespread nuclear weapon availability. If your peer adversary had the ability to spend debt on missiles, it was always an unaffordable, asymmetric proposition to counter that with a spend on missile defense. Sure, you spent some money on defense, but eventually the reality of the balance sheet became clear: damage at a scale that previously required your adversary to mobilize big chunks of their population in armed forces now just … cost money for missiles. And it would cost you much, much more than they spent to even semi-effectively defend against that. The opportunity cost of offensive action got massively reduced, IOW.

Nuclear weapons and other mutual-deterrence strategies emerged from that essential reality of modern mechanized warfare, they didn’t cause it. Once countries could no longer affordably defend against common attacks, they inevitably started looking for ways to deter attacks via incentive. The Cold War was the result.

Yes, but how do you know what the missile is going to hit? Or if it’s going to hit? Or when? Or how many casualties it might cause? Might overspending on defense encourage bellicosity in rivals? Might underspending on defense risk empowering them via vacuum?

I think that is, in a nutshell, why defense budgeting is so hard. Because that’s not evaluating cost vs. cost; it’s evaluating cost vs. the probability of an event with a cost (a probability whose causal influences are deeply entangled with the military and economic decisions of the defender).

In that comparison mode it is very hard for even experts to pick the right spending thresholds, to say nothing of the public. People are not good at thinking about probabilities, and are especially bad at evaluating probabilities compared to concretes.

I am not GP. GP said:

> I say "no thanks, they're going to get hacked" and then six months later... they get hacked.

I interpreted that to mean "they (the passwords) were compromised" not necessarily "they (the password vault provider) were compromised".

> you should be isolating your passwords across the company so that one user doesn’t compromise every login, no?

Correct.

On a very-few/very-skilled developer project, this might work. But on projects with large or changing teams, that downside bites hard.

The risk of bugs goes up when lots of people (with different timelines/skill levels/expectations) are working on an enterprise codebase. Ergo, the risk of "entire backend vanished" crashes goes up over time. A huge benefit of HLLs is that they have a vastly reduced risk of whole-program-aborting bugs--even Erlang, with it's "let it crash" philosophy, is referring to letting the boundary of affected code within a larger program crash, not aborting every transaction/reinitializing all resources and restarting from scratch.

Things like CGI mitigate this a little bit by containing the blast radius to one request's handler (at a pretty severe complexity/performance/compute cost to things like shared connection pools/file handles/startup time), but they don't solve it. The same is true for things like autorestart services/orchestrators.

Sure, you can decline to abort in the event of a segfault, but that's .... ballsy, shall we say.

- User account with access to the SaaS password manager was compromised.

- User account with access to the SaaS password manager was compromised.

- User account with access to the SaaS password manager was compromised.

- Password safe file (keepass) was compromised/leaked and its (medium-strength) master password was presumably brute-forced.

EDIT forgot one:

- Password manager was used by a user-facing web backend, which unlocked it with an unscoped read-everything master password. Web backend was hacked, predictable consequences ensued.

It’s easier to roughly copy, sure. As you say, examples of that exist. But outright cloning a game with only streaming/realtime content available? No decompilations, 3D asset parsing, or ability to run lots of simultaneous game loops? That would take ages and/or break the bank.

I’m not saying AIs are technically incapable of doing this. Rather, I’m saying that Stadia-as-DRM makes the act of copying technically infeasible (costly, slow).

The assumption that “stupids” are going to be more receptive to long-form/from-first-principles arguments is wrong, I think. So is the assumption that that’s why rationalists write that way (it’s more likely because of bubble culture and/or because people like hearing themselves talk).

If the latter, then it’ll only contend on simultaneous writes, which mutex anyway.

Genuine question; I do not know much about this area.

That means there are some (very poor) calorie-sufficient diets that don’t provide enough protein. But I agree that it’s still rarer than people think.

Possible, but doubtful. Small and large landlords in rent-stabilized NYC manage to be numerous. I also don't think that the inequality between small and large landlords is what's important here--inequality between resident single homeowners and non-homeowners is more impactful (ethically and for national stability and social progress).

I really do think that complexity and UI obtuseness was implemented as commentary, not as a mistake.