Re: Quantum computing practically impossible

[Phillip Hallam-Baker <[e-mail]>]

On Thu, Nov 5, 2020 at 1:45 AM Viktor Dukhovni <[e-mail]>
wrote:

> On Wed, Nov 04, 2020 at 11:23:32PM -0500, Phillip Hallam-Baker wrote:
>
> > It is my understanding of the work on Quantum error correction that it is
> > correcting errors in the measurement of quantum states rather than trying
> > to compensate for decoherence so the base assumption of the paper seems
> to
> > be off.
>
> No.  That's not the case.  The scepticism around in principle
> scalability of QCs is in fact centered on whether environmental noise
> can be corrected in principle or not.  Much has been written about this
> by Gil Kalai, who has a long-running debata with Scott Aaronson on this
> issue.  Neither side has conceded.  The Google "quantum supremacy"
> announcement has given the pro-QC camp some reason for new optimism, but
> Gil Kalai is disputing the conclusions of that experiment.
>

I was trying to avoid wading into that fight. My bigger point was that
particular debate is irrelevant as far as the scope of QIRG goes. My minor
point being that is not the only reason for needing correction. Quantum
computers are interesting in that your results come back in the form of bar
graphs...

As far as I am concerned, Quantum Computing is really the name of a branch
of experimental physics looking to find out if quantum states decohere
independently or obey the Q-rule of 'one down, all down'. And that is an
interesting experiment but not one we know the result of, nor is it
something that we can possibly know through extrapolation of other results.
I am pretty sure we will find that the states do behave as you suggest but
only because every other observation in physics is compatible with the
observed universe being a simulation performed using lazy evaluation but
that is a butter-side-up mode of reasoning.



> Time will tell who's right.  The devil is in very difficult details of
> the noise models that the experts don't agree on.
>

Experiment will show who is right - if someone can work out the experiment
to build.