[Doctor Stochastic]

In short, there are many processes that we must treat as random even if we know they are deterministic in nature.

This is surely true; often it's more useful to treat things as random even if we know the deterministic version. On the other hand, QM doesn't seem to be this way at all. I don't think the analogy of QM and strong PRNGs is all the good. (I have published about both.)

The problem is that with a cryptographic PRNG (such as DES or AES or Skipjack) is that the "True State" can be known by knowing the key. In QM, the "key" would also be subject to the QM laws. This has always been the problem for "hidden variable" (or cryptographic like) keys.

[tortoise]

The problem is that with a cryptographic PRNG (such as DES or AES or Skipjack) is that the "True State" can be known by knowing the key.

Yup, that is the problem with QM. We have no means currently with which to look inside the black box. For a strong PRNG, what can you do if the key is lost? As such, we can only build the kinds of models that we are all familiar with. I am under no illusions that we will ever be able to create a deterministic model of this process for the foreseeable future. My personal suspicion is that QM falls under the auspices of Fisher information and the general predictive limits of our universe i.e. it will never be decipherable as a deterministic process.

I don't think the analogy of QM and strong PRNGs is all the good. (I have published about both.)

Heh, well I've published on neither. That said, I knew the analogy was loose when I posted it. I am familiar with both and the computational aspects of both generally, but I don't claim any particular expertise. Computational chemistry and molecular modeling was my thing many years ago, but I've long since lost interest in it and generally become rusty.

The analogy was really about the information theoretic perspective of deciphering the internal state of a PRNG from the output with no prior knowledge of the internal state (e.g. the "key"). With QM, we approach it with the perspective of someone without the key, much like a third party would trying to cryptoanalyze something encoded with a PRNG. Theoretically, there are grave limits to the kinds of processes that can be deciphered without prior knowledge of the internal state, even very simple processes. Thus it is with QM, IMHO.