Probably. The Chinese will have the Windows source code anyways though. I know of several organizations that have it and I know people that have done security evaluations on it. The chinese will without doubt qualify for access and there are enough of organizations where they could get it from via spying otherwise. The only way to make it secure against Chinese attacks is not to put in the spyware functionality (like Linux). But that would cost MS money, so all that crap stays in.

There is no entanglement in quantum links. It is just polarization. But reading that polarization has a only a ~ 50% probability of returning it. Hence if you send a new one onwards, the receiver gets the polarization only in ~ 37.5% of the cases (since you get a random result if you read the polarization wrong) and thereby can detect a listener together with the sender. This is grossly simplified, of course. Quantum entanglement is a entirely different thing and cannot be used for information transfer. This happens to be a very fundamental limit, i.e. most of Quantum Theory collapses if it turns out to be untrue.

No. You do not understand what the "signal" sent looks like. It is _single_ photions. They either arrive unchanged with regards to polarization and energy state at a point in the middle or not. If you attempt to read the polarization, you destroy the photon and this attack is reliably detected. Incidentally, there is no way to do "amplification" for a single photon. You cannot accelerate or slow it down. You cannot change its wavelength. If it does not arrive, you can do nothing. This is not a light-pulse that consists of many photons that you could add a few more to, as an optical amplifier does.

The _very_ security of a quantum link depends of it being impossible to copy the sigle photons going through it. Hence no repeater possible. Your statement about an "abstract device" is just nonsense and the idea does not apply to quantum links at all. It is all very concrete Quantum Physics.

The security of a quantum link centrally and critically depends on the photons transferred being impossible to copy, and hence impossible to "amplify". So no, the concept of amplification does not apply. If you can amplify, security is gone. The problem is that there is no way to "amplify" a single photon without making copies of it. The photon either arrives at an intermediate "amplification" station or not. If not, it is gone and cannot be "amplified" by any means. If it arrives, it is in its original energy state and "amplification" would mean making copies of it, which would mean destroying all security properties.

I think you do not quite understand how a quantum link gets its security.

Greed, stupidity and a fundamental non-understanding of security. All things MS is great at.

You are thinking amplifiers for conventional optical links. On quantum-level, they actually copy. Optical impulses have large numbers of photons in every pulse, so it makes sense modelling them as amplifiers. But they are actually not when you get down to quantum-level. Which you have to do for quantum signals.

The thing is that conventional optical "amplifiers" do not work for quantum signals at all, since they do not copy polarization. Incidentally, if they worked, quantum-links would not be secure. So what you would need to do on quantum-level, is to do a quantum handshake in both directions and recover a conventional signal in between. That means you lose end-to-end security. Hence you _cannot_ add repeaters/amplifiers in a quantum link. The original photons have to make it though all to the other side or security is lost. That should make it amply clear this is not an "Internet" technology and can at most be used for specific dedicated links.

Actually, they do.

Indeed. Hence you need to do a whole, second network that works _without_ repeaters! This is just a bad idea that needs to die.

The whole thing is neither "near" nor useful in any way. Ans since you need to actually encrypt conventionally after exchanging a key with quantum technology, it does not even boost security.

No? Not interested then. Porn is basically the only "art form", where aesthetics really is the main thing that matters and story, plot, etc. are irrelevant. Hence while this tech may be able to make good porn, it will suck for anything else.

To everybody else, this is just what is expected. Unfortunately, the deniers will kill everybody else with them. We need to stop tolerating these cretins.

And obviously record and get more training data? Great, more reasons not to use their crap!

The original deal was an utter disgrace. Fortunately, Boeing was incapable of adhering to ist conditions.

VmWare is dead, do not fall into that trap.

Actually, no. LLM-type AI is limited by training material. It likely has already peaked and will never get as good again as it is now.