The Outer Space Treaty declares outer space "international waters", which means that the orbital data centre is by no means extraterritorial, it has to be registered under a flag. Legally, per the OST, it's no different from having it in the middle of the ocean.

So that's not it, chief.

Obviously it is not cost, as you have pointed out.

Cost is not the objective, jurisdiction is not the objective, which leaves: Military applications.

And we have known that data mining impacts people's lives with rockets since the Snowden revelations.

Unless you know who did it already, you have no way of knowing if they are in China or Korea.

Machines are generally more secure today than 20 years ago.

Software firewalls were a big thing back then, and offered a convenient remote attack vector:
All ports open, just to report if someone pings your box. Running with administrator rights, of course.

Nowadays the common way is either to privilege escalate from JavaScript, or to go through Bluetooth.

Virulently infecting the anti-virus gives you a free rootkit, and nobody is surprised if that roots through their box.

The C&C aspect just helps mask any worms in the infection.

That is how anti-virus works.

It is a root kit with remote C&C that you expect to read all of your data.

I will include more context when quoting in the future, just so your chatbot will not be misled by judicious elision.

Even though the context is still present on this very page for those who actually care about it.

And netiquette advises to quote the salient parts, not the full context, for the simple reason that the context is readily available for those who actually care about it.

Netiquette also advises against verbosity.

Sounds like motivated reasoning to me.

That you can reason yourself into feeling certain ways does not make cognition a requirement for emotion.

What you actually said:

Suggesting that the word "drone" means "mindless bee."

The word "drone" can refer to male hive insects. Or it can refer to remote-controlled airplanes. Completely different things, but neither really describes a mindless unthinking being. If I had to guess, I would say you got that from bad fan fiction.

I am not. Nothing I said suggests disruption. As you said yourself:

However, what you are choosing to ignore is that, once a bee has been accepted into a different hive, it becomes part of that hive. It does not return to its original hive. It has forgotten.

You yourself gave a better explanation than I could of how that is not true.

Drones are specifically the male bees. Most bees are not drones.

And of course they all act individually, how else would individual bees act.

The behaviour of the hive is different from the behaviour of the individuals: Individual bees are communicative, hives are not.

And an individual bee has limited memory. They even forget which hive they are from after a while if they don't return to it.

Deja-vu!

Didn't OpenAI say the exact same thing five years ago?

10 years ago, Eliza was not AI yet.

It's what AlphaFold is doing, but AlphaFold is not an LLM, it's something completely different.

There are experiments that have shown insects to not modify their behaviour in certain situations when any mammal would lose patience and try a better way. That is probably the reason why "people" "assume" that insects are "reflex-based machines."

Of course it is also known that bees can learn. They learn of the locations of sources of nectar, for example, and communicate that to their hive through dance. But it has also been shown that individual bees retain that information only for half an hour. The hive can retain it for much longer.

That's not trial-and-error learning.

I don't think anybody questions that insects have emotions. Emotions are hormones. Insects have those. Plants do, too.
And insects have pain receptors, so they are able to feel pain, but the effect is not the same as in mammals. (Or in plants, for that matter.)

Brains are central organs. Insects have nerve nodules that serve the same purpose, but they are not as centralised. In that sense, insects do not have brains, but they have something similar that might as well be called a brain. (And there are fungi that are specialised in manipulating those, physically, to elicit specific behaviour in ants, which works only because their "brain" structure is not as flexible as actual brains.)

That bees can move obstacles and climb on things does not in itself prove that they plan those solutions through abstract thinking.
So this seems like another paper that has been published for the readers to have a good laugh.
Even if it were evidence of abstract thinking, that has no implications for emotions or pain.

With the separation of L1 cache into data cache and instruction cache, every modern CPU is effectively Harvard architecture.

The Minix in the Intel CPU was put there by Intel, not by any government. (I suspect that Intel use it to spy on the NSA.)

Merging an app with the OS, that's what Microsoft tried with the Internet Explorer. It did not end well for them.
Effectively, half of Microsoft Office is baked into Windows, especially the parts that date from the MS-DOS days. (The "if Lotus will run, DOS is not done" days.)

All modern CPUs have had vector extensions that facilitate matrix multiplication since the last millennium. Nothing new there.

For a minimal real-time OS, there is plenty of choice. RTOS is a common one.
Plan 9 would have the advantage of supporting the fork-join process model from 1957 that Windows still struggles with.

That's also why apps are not a good idea: They are based on the batch process model, not taking advantage of the parallelism primitives offered by modern operating systems. Not only is that model slower, and uses more RAM, it also makes it more difficult for programmes to interact.

AI apps typically run on GPUs or similar specialised hardware, which Linux will never be ported to because, while the hardware is highly parallel, it has no native support for context switches between processes. The parallel hardware might make context switches superfluous anyway, but that would require a new OS written from the ground up. Instead, Microsoft are using Android here, which uses a Linux kernel (and ships with a specialised daemon for inter-process communication between apps).

Emulating an OS is a stupid idea: Any Turing-complete machine can emulate any other Turing-complete machine, but an OS abstracts the hardware. There is no point in abstracting the abstraction. What you want instead is a compatibility layer that implements alien OS primitives. (There's a reason why WINE Is Not an Emulator is faster than native Windows.)
Of course you lose all the advantages of a hardware-specific OS either way.

Memristors are another hardware architecture that is looking for applications. There was hope that they might be more suitable for AI, but that has never materialised. Currently, no OS exists that can run on memristors. There is also not C standard library port for them, and it might not make sense to create one. There is also no back-end for the GCC for any memristor architecture. (GPU and quantum computer code is usually generated by LLVM back-ends, not GCC; I expect that it will be no different for any other exotic hardware except maybe FPGAs which have their own proprietary compilers.)

What could go wrong?

You are missing the point. It's not about the violence.

In chess, you win the game by capturing the king.
All the other pieces are dispensable, it doesn't matter how many you lose. Pawns can be made rooks and knights and bishops and queens, but when the king is captured, it's game over. Everything hinges on protecting the king.

It's a single point of failure. The goal is to exploit it.

Eventually.
Because while the Mongols were busy capturing Nanjing, the Chinese were capturing Ulan Bator.

And the Manchu dynasty didn't have an "unbroken lineage" either, let alone to ancient times.
There were a lot of revolts, revolutions, palace revolutions, palace coups, changes of dynasty, historical revisions, wars, and civil wars.

That's not the point.

The point is: There is no king to capture in Go. No single point of failure to exploit.
While you are busy capturing an area, you opponent is putting your eyes out.
While you are busy putting eyes out, your opponent is busy encircling you.

That's why any would-be conqueror ends up speaking Mandarin.
Not because of the Heavenly Mandate. If you pay attention, China doesn't even have an emperor anymore.

While Britain and Russia were playing chess, China was playing Go.

It really doesn't.

Chess is the game of kings, mostly because it's all about the king, but also because it is a model of warfare, which used to be what kings concerned themselves with. But chess doesn't model logistics, it models battles.

Go is a much simpler strategy game. It has no pawns, rooks, knights, bishops, queens, or kings, all with their own move set. It only has two rules, everything else follows from that. Which makes it much harder to master.

Everything affects everything else. A wrong move early in the game can lose you the whole game down the line.

Poker isn't a strategy game at all. By rules alone, it is purely a game of chance.

There are other differences as well. It's not a board game, it's a card game. It's not one-on-one, any number can play. It's not a perfect information game, it involves bluffing and guessing.

You don't misdirect an opponent by hiding an attack by feigning another attack. You don't attack at all.

More to the point: There is no strategy. There is no planning.
The cards that you are dealt do not depend on what you did before.
It is more advantageous to lose often and early, and to win rarely but big. (And to let your opponents win often and small.)

So, while the USA were busy bluffing, everyone else was counting the cards.

It does explain some things in history.