Posted on 12/20/2017 11:05:42 AM PST by RoosterRedux
A company in California just proved that an exotic and potentially game-changing kind of computer can be used to perform a common form of machine learning.
The feat raises hopes that quantum computers, which exploit the logic-defying principles of quantum physics to perform certain types of calculations at ridiculous speeds, could have a big impact on the hottest area of the tech industry: artificial intelligence.
Researchers at Rigetti Computing, a company based in Berkeley, California, used one of its prototype quantum chipsa superconducting device housed within an elaborate super-chilled setupto run whats known as a clustering algorithm. Clustering is a machine-learning technique used to organize data into similar groups. Rigetti is also making the new quantum computerwhich can handle 19 quantum bits, or qubitsavailable through its cloud computing platform, called Forest, today.
The demonstration does not, however, mean quantum computers are poised to revolutionize AI. Quantum computers are so exotic that no one quite knows what the killer apps might be. Rigettis algorithm, for instance, isnt of any practical use, and it isnt entirely clear how useful it would be to perform clustering tasks on a quantum machine.
(Excerpt) Read more at technologyreview.com ...
Has anyone asked Schrödinger’s cat what he thinks of all this?
Isn’t he always?
Not if you ask him.
Will it be able to answer the ultimate question of life?
No. It does NOT.
Quantum "entanglement" is a slight misnomer. Because the "particles" in an entangled system are indistinguishable, neither of them are really present at either location in the sense they would be in classical physics (or in our intuitive understanding of particles.) Entangled "particles" are nothing more than indistinguishable states in a quantum system. Altering one projection of this multi-component state effects the projection of another component, and the state as a whole. Because it is never anything but a single state, it cannot be used to transmit information. If it could, there would exist Lorentz frames in which the state vector of the component measurable by the "receiver" would transition before the component was altered in the "sender." That could be used to violate causality. [Loosely: time-travel]
https://en.wikipedia.org/wiki/No-communication_theorem
https://en.wikipedia.org/wiki/EPR_paradox
I should mention there are some--actually reputable--physicists who believe the No Communication Theorem is not true. This considered a fringe position.
“It would imply instantaneous communications over any distance...”
That is a common misconception, but anything of that nature is specifically ruled out by the no communication theorem in quantum physics. Transmission of information using quantum methods is limited to the speed of light, just as it is using normal methods.
Public key encryption becomes useless. But there are still some encryption methods (one-time pads are the best and most straightforward example) that even in theory cannot be broken.
You mean that ultimate question of...what flavor of cat food should I buy and know that my cats will not just walk away with noses in the air?
I don't know if any human or human invention can answer that question.
*sneeze*
The Chinese are way ahead in this area - they have QCs which communicate from the ground to satellite and to other ground entities - communications which cannot be intercepted and cannot be broken. As far as clustering goes - we gave them the clustering algorithms for free, enabling them to make the fastest super computers in the world.
My wife figgered that out long ago!
He both approves and disapproves.
If it defies log, then aren’t its conclusions invalid?
CC
Logic.
CC
That something is a wave or a particle depending on which you test for is logic-defying to me!
The number of states is large, but finite. It is also clear that there is an indeterminacy price to be paid using qubits, and some that has to be given back in a loss of coding density; this further reduces the number of states dedicated to the problem (the lost states are essentially given up to error correction.) Quantum code breakers will be fast enough to crack all public key encryption in real time, but certainly not infinitely fast.
The experiments carried out so far are not much different from polarization experiments which verify Bell's Theorem or, if you prefer, the EPR Paradox, which is really a strange result, but no paradox at all.
Couple things to keep in mind here.
First, a lot of the literature on this subject is very confused, much like Schroedinger's Cat and EPR. People are simply not careful enough to draw the conclusions they've claimed. Second, Wheeler never went as far as to say there would be Lorentz frames in which a quantum measurement would violate causality. He came very close but never said so. Third, that eigenstate into which a state vector collapses upon measurement is a stationary state of an Hermitian operator. There actually is no operator corresponding to "looks like a wave" or "looks like a particle" and that's a GOOD THING, because the state vector (or wave function, if you prefer) of a quantum system is NOT (despite elementary analogies made in junior high to the contrary) "sometimes a wave" and "sometimes particle" it is always one thing: a state vector, which completely describes everything we can know about (say) a photon. That's all you get, and that's all there is, and it isn't sometimes one thing and sometimes another.
I think a lot of people are drawn into the "isn't causality violated by this" trap by a fundamental misunderstanding which arises from examining pointlike behavior, which is an illusion. Quantum systems do not "behave" as trajectories evolving from point-to-point in the absence of constraint, which behavior can sometimes lead to paradoxes when we change our methods for measuring them or when we consider entangled systems. This is falsely claimed as a unique feature of quantum mechanics. Actually, it isn't. Classical mechanics also has this feature. In his (in my opinion, highly overrated) graduate text on advanced Classical Mechanics Goldstein actually comments on this in the context of refraction.
Even in classical physics, light takes the path which involves the shortest physical distance between two points, which is why (for example) a ray of light bends toward the normal when a relatively higher refractive index is placed between the source and destination. How does the light ray "know" bending toward the normal will shorten its effective path? How does it "know" that it's time to stop doing this when the medium changes again?
In both classical physics and quantum mechanics, the answer is the principle of least, or "stationary" Action. And as Goldstein points out in his exposition of stationary Action in classical physics, "reflection on this often leads to pointless teleological arguments."
Just so.
What is happening is better understood via the Feynman Path Integral than via any other conceptual or theoretical tool. The Feynman Path Integral formulation works just great in classical physics, too, and it leads to "correct thinking" about these things. A physical particle's state vector literally travels every conceivable path through space time, and the trajectory which minimizes its Action is the one that ultimately is observed. New measuring instruments, silvered mirrors, half-silvered mirrors, microwaved atoms, or whatever else you want to throw into its possible paths will factor into that outcome, throughout all of spacetime.
So, why do these point-to-point interpretations seem to fail so miserably when we do thought experiments, when they are, in fact 100% equivalent to the Feynman Path Integral?
The answer is that constraints are cooked into the motion. As is the case with any good magical act, preparation is everything, and the rabbit goes into the hat long before people are watching. The failure of "local realism" is a failure of inattention on the part of the audience. A light ray bends toward the normal, following the path of least Action because a light ray is constrained by an equation (the Klein-Gordon Equation) which determines the behavior of light throughout ALL OF SPACETIME. And this is the part that people forget when they allow themselves to be tricked by these apparently "non causal" or "non locally realistic" theories.
These theories are NOT LOCAL TO BEGIN WITH!
[
This "surprising" outcome is well understood by mathematicians, but occasionally forgotten by physicists. The behavior of a twice differentiable function throughout any volume of space is entirely determined by what happens on its boundaries. Once you have set what happens on a boundary, nothing the function does anywhere else has any real freedom.
]
Same thing with electrons. In non-relativistic quantum mechanics, you have the Shroedinger Equation, which, again, permeates all of space for all time (non-relativistic, so space and time are not codependent.) In relativistic quantum mechanics, the Dirac Equation does the same thing.
Because in the Feynman perspective, you are integrating the Action throughout all of spacetime, you can't forget about the constraining equations, which always apply, and which rule out strict locality.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.