akoboldfrying

This logic doesn't work here, because it relies on the premise that you alone potentially have an unreleased zero-day exploit, which is not the case.

If prompting an AI could exploit this for less than $500 in token spend, then you should immediately do so, since if you don't, in a few minutes' time someone else will.

They aren't at all the same thing, but at a high level they both reduce the amount of fragile/tedious setup-specific futzing you need to do to get a few basic services up.

> What modern production calls “correction,” and mistakes for care, is really a form of epistemic violence

I'm so tired of people reaching for the emotional impact of the word "violence" in contexts that are far removed from its actual meaning -- and this context is about the furthest removed I've seen to date.

> If we cannot hear and feel coordinated human time, we cannot learn the forms of felt life that make mutual recognition possible.

I actually laughed out loud.

Remember folks, "punching in" is Okay. Gridifying is Not Okay. Because ________, obviously.

I often find it helpful to type comments (starting with a # character) directly in a shell session, to remind myself where I'm up to, what I need to do next, or what I've just done. Especially handy at the end of the day or when switching context to something else.

> Deregulation, smaller government, and reducing bureaucracy aren't conservative values.

They're fiscally conservative values, as you likely already know. Social and fiscal conservatism are correlated, as you likely also already know.

In order for you to make this statement, you had to take a word ("conservative") that can be interpreted in one of two well-known ways ("social conservative", "fiscal conservative"), and pretend that when I used that word, I must have meant the way that doesn't make sense, instead of the way that obviously does.

> the AfD doesn't even pretend to care about normal people.

It explicitly cares about people who identify as traditional Germans, who in their own eyes are "the normal people".

On the economic side, Gemini tells me that they favour deregulation, smaller government, and reducing bureaucracy for small and medium-sized enterprises (Mittelstand). They also want to increase family support payments to incentivise having children. These are all economic policies that make sense to conservative-minded "normal people".

I agree with this actually. All of these things were good for some and bad for some, and these groups overlapped. All were good for (most) humans-as-consumers, and bad for (some) humans-as-producers, and the IR was also bad for horses.

It's "worse for humans" in exactly the same way that first the Industrial Revolution and later computer programming made things "worse for humans", but the post you're referring to only notices that it's "worse" when they are the victim.

AI has already made mathematics better in the sense that it has greatly advanced what we know to be true (theorems).

The only way that it hasn't made mathematics better is in the sense of human mathematicians' career prospects -- and in this respect, it is identical to other forms of automation that the world, and especially computer programmers like us, have participated in in the past without complaint.

Interesting. LLMs seem to now handle other kinds of state changes with ease, so I'm not sure why they would trip over this.

Do you mean stack-based source code languages like Forth or PostScript, or directly writing machine instructions for a stack-based ISA/VM like the JVM?

I can see how thinness is a benefit for induction heating, but how does it benefit a regular countertop?

The system in TFA can only handle 2 "categories". Note that there could be many distinct types in each category; the important thing about categories is that everything in the same category should share an allocation/deallocation timeline. If this holds separately for each category, there will be no fragmentation.

For this 2-category case, the double-headed stack is strictly better than maintaining 2 separate half-size stacks, since it permits every sequence of allocations and deallocations that the latter permits, while there are sequences that it permits that the latter does not permit (namely, any sequence in which more than half the memory is temporarily needed by one of the categories, and the other category is temporarily not using it).

I think the idea of "software architecture" may become irrelevant.

Architecture, in the sense of applying/enforcing organising principles like modularity and levels of abstraction, is something we currently do largely to manage complexity -- so that we can fit as much of the system into our limited brains as we can. For a large enough system, this is necessary in order to work on it efficiently. But if you can "keep the whole thing in your head" without using these "crutches", then there's actually no need for them. They become overhead that can be done away with. We do this already for sufficiently small things -- no one plans a software architecture for a "Hello world" script.

It's possible that future (maybe even current?) AIs will be able to "keep the whole codebase in their heads" just fine, in which case the need for architecture goes away.

No.