Really? Schrödinger's cat again?

[DaveE]

I have a new Physics hypothesis (ok, really Journalism, I guess): Whenever you read anything the features that damn cat, it will be at least misleading, maybe wrong. The worst way to learn about QM is to start by listening to someone talking about quantum cats, it's practically a guarantee that they don't know what they are talking about.

The research is cool, for reasons that are completely missed by the journalists.

https://www.sciencealert.com/physicists-think-they-ve-figured-out-a-way-to-save-schroedinger-s-cat

 

[EPR]

Quantum jumps and radioactive decay are 2 different things. I do not yet know what this experiment has in common with the cat. They should show how they undecay an atom.

 

[DaveE]

Correct me if I'm mistaken, since y'all know more QM than I. Plus, I didn't read the paper. What I think they show is:

1) They have a method of detecting that a jump has occurred, or is very likely to occur, sooner that the way other people normally detect it. I suspect the jump has already happened (started?) in both cases, they just know sooner.

2) If you know the initial state of your system (QuBit, in this case). You can build a machine to reinitialize it when you want to. Really creating a new initialized system. Alternatively, if you expand your system definition from the QuBit to the whole experiment, then they've built a really complex system that exhibits the quantum jump in ways other than the before/after state of the QuBit, like the big pulse of radiation to reset things.

So the good part of the research is they have advanced the ability to monitor the state of their QuBit. No cats required.

 

[DaveE]

So my question is, in a real computer, how do you know when to stop the decoherence? Which jumps are noise and which jumps are signal. Is this just a classical memory bit that has a known state and only changes when you make it change?

 

[jedishrfu]

I think this is more of a statistical issue. Normally when running Quantum Computer algorithms, you set up the machine state, execute the algorithm steps and then read the result.

You don't do this once you do it many times and the answer pops out statistically as the most likely value.

Basically, you have to throw out any notion of classical thinking here qubits are not bits. Over some time period, they statistically behave like bits but not so at any given instant.

 

[Nugatory]

I have a new Physics hypothesis (ok, really Journalism, I guess): Whenever you read anything the features that damn cat, it will be at least misleading, maybe wrong.

Hypothesis? I think you have just identified another law of nature.

 

[vanhees71]

By the way the scientific paper and in all details from the PhD thesis underlying the paper, from which one can understand what really has been achieved, can be found here:

https://arxiv.org/abs/1803.00545
https://arxiv.org/abs/1902.10355