Quantum computing has already been demonstrated, but on a small scale. The challenge in achieving all of what we know quantum computing can achieve (as well as the things we have yet to discover) is in scaling it up, for which error correction is one of the key missing ingredients.
There are various applications. Some that you hear about most might be a little overblown (but still maybe not). For example, quantum simulation is an application within the physics field that already makes these systems worthwhile to pursue, without any mention of breaking encryption and the other things mentioned. Quantum logic, i.e. quantum computing on the smallest scale, has been employed to make cutting-edge atomic clocks (which are undeniably a worthwhile pursuit, with countless applications in the real world).
Putting "scaling up quantum computing, the pieces of which have already been demonstrated" in the same sentence as perpetual motion and free energy (known to be in violation of laws of physics) is completely unjustified.
Quantum computers aren't that powerful, in my understanding there should be no reason that a large quantum computer couldn't exist. [1] is a fun paper that discusses if efficiently solving NP-complete problems should be considered impossible in physics. In it it is also discussed why this does not exclude quantum computers.
QC is a direct consequence of the known laws of physics. To discover QC doesn't work would be proof the QM is fundamentally wrong - not that there is some more fundamental theory, but that we have been accidentally getting the most accurate predicitons of any theory in the history of science.
That possibility alone is worth all of the investment in this field.
In contrast, String Theory is just an extension of QM, motivated by nothing except some cool looking math. It doesn't even make any definite predictions (supersimmetry is possible in string theory, but not required, at least not at any particular energy level).
I agree with everything you said in the first two paragraphs. But string theory is not “motivated by nothing but cool looking math”. String theory is the generalization of quantum mechanics to higher dimensional fundamental objects, i.e. beyond particles, but in many senses the string theories that are well understood are just quantum mechanics in a different suit and hat.
It is already known that many of the most physically important quantum field theories are actually string theories if you rewrite them in different variables (this is called AdS/CFT duality), especially for understanding their large-coupling behavior string theory is the only tool available to analytically understand the theory. So string theory being completely wrong would be as surprising as quantum computing being impossible, for the same reason as you gave for QC. It would indicate something profoundly wrong about something we think we understand well. Its not impossible but the case is much, much more robust than the general internet understands.
My point about string theory is exactly what you're saying: string theory is a different matehmatical formulation of QM, that happens to permit additional phenomena that have not been observed. It being a generalization to higher-dimensional spaces is exactly my point: it is mathematically motivated, not motivated by observation.
And yes, to the extent that it's the same formulae as QM, finding counter-example would be fascinating. But (higher-dimensional) string theory could be wrong with no impact on QM/QFT.
Well my main point is that AdS/CFT duality is asserting an exact equivalence between quantum field theories and string theories. So for example N=4 SYM, a quantum field theory viewed from one perspective, is exactly equal to a string theory from another perspective. That’s why I say it would be a spectacularly weird thing if string theory turned out to be unphysical but SYM was, because it would turn out then that they weren’t equal, and we really strongly believe they are.
In other words I’m not saying that string theory is (only) a generalization of known correct physics, but that many string theories are equal to (and therefore by definition as physically valid as) physically important quantum field theories. Then there are other string theories that are truly interesting and novel, too.
Looking into this a little, AdS/CFT correspondence is only a conjecture, not a proven fact. More importantly, there are no AdS/CFT solutions that actually work with 4-dimensional curved space time, the working models are all in known non-physical geometries.
It’s a very strongly supported conjecture, but it has not been proven rigorously. As I said it would be extremely surprising if It were not true because of a frankly absurd amount of evidence. The second claim you make is false though— 3D holographic CFTs would be dual to a 4D quantum gravity theory.
Additionally, this conjecture is actually experimentally verifiable—you can set up strongly coupled systems in the lab and study the string theory description, and compare the predictions.
I respectfully disagree with your disapproval of my comment here, I agree that my opinion is quite strong but I don't think it is offensive.
Given that Quantum Computing has yet to deliver on its promise after decades of work and huge investment, I think it is safe to compare the field to other fringe science that once occupied some of the best minds on the planet (like Newton and his obsession with Alchemy).
Quantum computing isn't fringe science and almost no one in the physics community questions the validity of the underlying theories. It is a large engineering challenge so it's unclear whether a working large-scale quantum computer will be realized anytime soon, but that doesn't make it fringe science.
Serious question for you. Have you read any recent papers? Some quantum computing applications have outperformed classical computers already[0] and will get better with more qubits and less noise. This stuff is in its infancy and making small baby steps. The phone in your pocket didn't get here without punch card programming or computers the size of rooms. This is that and you're around to witness it. Have some patience!
I share the sentiment that quantum computer folks hasn't delivered but I dont feel its a fringe science. Its just hard to deal with noise at that level.
To me, string theory feels like a fringe science; and irritating one at best.
It's not that it was offensive, it's that it was too generic to be interesting. The problem with generic comments (like, e.g., "quantum computing is fringe science") is that they don't contain any specific information—they just make grand, empty claims. These nearly always cause discussion to go in a further generic direction (e.g. "no it isn't", "yes it is"). This is basically the arrow of entropy applied to forum threads.
There's another aspect too: the more generic comments are, the more predictable they are. Since they don't contain much actual information, there's nothing really new to say, which means people repeat the same things over and over. Often they get angry as they do it, since indignation usually fills the vacuum when curiosity is absent.
All of this is the opposite of what we're looking for on HN. What we want: specific, unpredictable, curiosity-gratifying conversation.
Where? I've been following this field for almost 30 years now, and I have yet to find a single publication where the produced result is useful in any way.
More work is needed to turn quantum enhanced optimization into a practical technology. The design of next- generation annealers must facilitate the embedding of problems of practical relevance.
Like fusion, current progress is promising, but practical results lie in the future?
"Not working as well as alternatives we already have" is a lot better than things which do nothing at all. There is no promising progress for cold fusion or alchemy.
Unlike perpetual motion or free energy, the math and physics behind quantum computing is rock solid - the possibility is baked into the framework of quantum mechanics at an absolutely fundamental level. Even if it turned out that all of quantum field theory was completely bogus, I would still expect the basic framework of quantum mechanics (i.e. unitary evolution, the Born rule for converting amplitudes to probabilities, density matrices describing mixed states) to be correct.
If quantum computing was not possible, that would mean that everything we thought we knew about physics was hopelessly wrong. Physics would have to be much stranger - we'd have to give up on either locality (i.e. causation being limited by the speed of light), reversibility (closely related to conservation of energy), or nondeterminism in order to make the math work out. There are several papers in the literature that derive quantum mechanics from these or related assumptions (often together with some extra technical detail, that differs between approaches):
I think that quantum mechanics is a useful model and there is a lot of truth in it, but it is not the truth.
Using rock-solid math and physics I think you can do time travel and speculate about parallel universes, but I don't think we should blindly believe the models.
We should never confuse the map and the territory.
Then how do you explain the extraordinary predictive power of QM, and why do you believe that the same maths that gives us measurements accurate to 23 decimals is completely inaccurate in its predicition of QC?
30 years is nothing given how complex the engineering problem is of actually building a working QC. Fusion is much simpler and very well understood and yet it has been more than 50 years since research into fusion power plants has started, and the most optimistic estimates say we'll have the first ever research power plant 30 years from now, if everything goes well (EUfusion's DEMO plant).
There is no solid math or physics which supports time travel. There are a lot of cases where people claim to have found some loophole or another in the impossibility proofs for time travel, sketching out a model that looks plausible at first glance, but they all end up postulating some blatantly non-physical entity somewhere along the way (i.e. cosmic strings, negative energy density, ... - I've gotten my hopes up quite a few times).
Speculating about parallel universes is a different story - even if they exist, we can't travel to them, so there is no sense in claiming that we can somehow harness their power. It's a philosophical question, not a physical one. In contrast, the rules of quantum mechanics (when phrased in terms of density matrices, at least) only describe things we can actually measure and observe, and people actually do test every single aspect of the predictions quantum mechanics makes.
As a nitpick, negative energy densities are indeed uncommon but they aren’t unphysical. In fact, every quantum field theory admits at least one state with negative energy density—its a theorem. But even in quantum field theory there isn’t enough negative energy density to actually form acausal (faster than light) shortcuts through the spacetime.
Sure, experiments will always be the ultimate arbiter of truth. But if you don’t “visit the territory”, you’ll never know that your map is wrong. The map has been challenged in many many ways that seemed more impossible than large scale QC, and the map turned out to be right.
Backwards time travel isn’t possible in general relativity, though, making very mild and physically plausible assumptions about the spacetime. Same in semiclassical quantum gravity.
How sure are we that quantum computing will ever work? I think this boils down to whether the wave function collapses as expected(?)
Perhaps reality is actually only approximated by the wave function but reality is something else under the surface which means quantum computers may need to be reconsidered.
The only barriers to them working are engineering challenges which may prove insurmountable, may not. There are not any theoretical considerations that would make them not work.
There are certainly possible reasons why quantum computing could be impossible. To me the most compelling of these would be that the error rate of physical qubits might fundamentally always exceed the threshold for correction with any real ECC as the number of logical qubits increases.
I highly recommend reading [1] for an overview of the various skeptical viewpoints on quantum computing. Note that, while Aaronson argues most of the initial arguments are trivially wrong, he admits that whether or not error correction is fundamentally impossible is an open question.
Not really, the Alcubierre drive depends on things we have no reason to think exist in the universe (negative energy). It's also likely that the Alcubierre drive doesn't actually allow accelerating from slower than light speeds to higher than light speeds.
QC in contrast must exist or else QM is fundamentally wrong - not incomplete, but wrong.
We know its possible to have a fusion reactor, we would just have to dump enough mass together to build a sun. Hopefully we can make it more cost effective.
There isn't a natural phenomena that solves a problem by encoding it and (ab)using quantum physics.
2. Everything after the first sentence doesn't make sense. Any system that implements a restricted class of unitaries and features local measurement as described by Born's rule can efficiently solve problems in BQP. We have no reason this should apply to 10 qubits and not 1000.
The theory of quantum mechanics has been completely predictive in materials and has resulted in the most accurate predictions in nature (Anomalous magnetic dipole moment to a part in a trillion.) It doesn't matter if the theory is "Macroscopic" or whatever, it's predictive and within the constraints of its predictions you have known computational power up to problems in BQP. Whether BQP=P is an open question, but as is P=NP.
Known problems are things like wiring, materials science/sources of noise -- eventually including things like cosmic rays. It should work and we have no reason to think it shouldn't due to fundamental limitations of the natural world. It is, however, one of the hardest scientific things humans have ever tried to do so far. To say it'll definitely work is hubristic.
This is not quite true. There are spontaneous collapse models [1] that are consistent with the data, like the GRW model [2]. Personally I give long odds against these being true, but you can't rule them out on the basis of current data. And if they are true then they would present a fundamental limit on how big you can make a quantum computer before it fails.
There have been superpositions maintained on the scale of a meter using these fountain experiments. I think the chip traps, SC qubit chips, dot chips, etc… will be much smaller than this.
Not sure what you're referring to here. Maintaining a superposition is not hard. The light from distant galaxies has been in superposition across literally billions of (light) years before we detect it. What is harder to maintain over large distances are entanglements, but even those have been maintained across hundreds of miles.
When I talked about a "fundamental limit on how big you can make a quantum computer" I wasn't talking about it's physical size, I was talking about the number of mutually entangled qbits it contains, and the amount of time it is able to maintain those entanglements in a coherent state.
As you can see I don't really know much about this stuff.
There's apparently noise in the system - how do we know that the reality is different to the model, and what we are putting down to noise is actually reality being different to the model? Kind like Einstein's correction of Newton.
This is part of the reason I like that the first phase of Google's current quantum computer roadmap is "physics derisking" [1]. There's no known reason that it should be impossible to make a thousand qubits dance together in a highly controlled way and get the expected results, but maybe there are some unexpected or unknown obstacles.
That being said, if physics derisking failed, it seems really unlikely to me that it would be because quantum mechanics itself was wrong as opposed to there just being some unforseen problematic error mechanism that was consistent with quantum mechanics. Finding a flaw in quantum mechanics itself would be a scientific revolution, which would be awesome, but those are also pretty rare.
I also think it's a little telling that people keep thinking it's the philosophically unsatisfying parts of quantum mechanics that are going to break, when entering a new regime. It smells of confirmation bias.
> How sure are we that quantum computing will ever work?
Up to 12 qubits or so.
> Perhaps reality is actually only approximated by the wave function but reality is something else under the surface which means quantum computers may need to be reconsidered.
Possible, that reality is only approximated, and that approximation breaks down at more qubits. However, that would be the more exiting outcome, since then we have an experiment that gives us a window into a post quantum theory.
To the authors of such blurbs about scientific breakthroughs: please cite the exact paper you refer to within the first paragraph, including a hyperlink. Using a generic phrase like: "... a team has demonstrated a way to detect errors in the setting of a quantum bit..." is not enough information to unambiguously understand what the author of this piece is talking about. In paragraph 7, the author of the blurb mentioned the name of the relevant scientists for the first time, after the blurb quoted a number of other scientists, included a link to a previous paper, and the reader still doesn't know which paper this blurb is referring to... I guess that even an AI-based blurb generator could learn some simple rules about being specific in crediting the source that inspired their writing.
It's actually from last year, and other groups have done better experiments since then (in particular, experiments doing multiple rounds of correction instead of just immediately smashing the system and analyzing the pieces). It's quite funny to see a big deal being made of this paper now rather than a year ago.
51 comments
[ 96.0 ms ] story [ 1888 ms ] threadIt is, in my opinion, not much more credible than perpetual motion or free energy, and less credible than cold fusion...
There are various applications. Some that you hear about most might be a little overblown (but still maybe not). For example, quantum simulation is an application within the physics field that already makes these systems worthwhile to pursue, without any mention of breaking encryption and the other things mentioned. Quantum logic, i.e. quantum computing on the smallest scale, has been employed to make cutting-edge atomic clocks (which are undeniably a worthwhile pursuit, with countless applications in the real world).
Putting "scaling up quantum computing, the pieces of which have already been demonstrated" in the same sentence as perpetual motion and free energy (known to be in violation of laws of physics) is completely unjustified.
My intuition is that large scale quantum computing is also in violation of laws of physics, but we don't have good models/theory about those yet.
What I am calling for is a reality check on this field.
Pretty much like string theory I am ready to bet it won't go anywhere.
[1] https://www.scottaaronson.com/papers/npcomplete.pdf
That possibility alone is worth all of the investment in this field.
In contrast, String Theory is just an extension of QM, motivated by nothing except some cool looking math. It doesn't even make any definite predictions (supersimmetry is possible in string theory, but not required, at least not at any particular energy level).
It is already known that many of the most physically important quantum field theories are actually string theories if you rewrite them in different variables (this is called AdS/CFT duality), especially for understanding their large-coupling behavior string theory is the only tool available to analytically understand the theory. So string theory being completely wrong would be as surprising as quantum computing being impossible, for the same reason as you gave for QC. It would indicate something profoundly wrong about something we think we understand well. Its not impossible but the case is much, much more robust than the general internet understands.
And yes, to the extent that it's the same formulae as QM, finding counter-example would be fascinating. But (higher-dimensional) string theory could be wrong with no impact on QM/QFT.
In other words I’m not saying that string theory is (only) a generalization of known correct physics, but that many string theories are equal to (and therefore by definition as physically valid as) physically important quantum field theories. Then there are other string theories that are truly interesting and novel, too.
Additionally, this conjecture is actually experimentally verifiable—you can set up strongly coupled systems in the lab and study the string theory description, and compare the predictions.
https://news.ycombinator.com/newsguidelines.html
Given that Quantum Computing has yet to deliver on its promise after decades of work and huge investment, I think it is safe to compare the field to other fringe science that once occupied some of the best minds on the planet (like Newton and his obsession with Alchemy).
[0] http://arxiv.org/abs/1905.02860
To me, string theory feels like a fringe science; and irritating one at best.
There's another aspect too: the more generic comments are, the more predictable they are. Since they don't contain much actual information, there's nothing really new to say, which means people repeat the same things over and over. Often they get angry as they do it, since indignation usually fills the vacuum when curiosity is absent.
All of this is the opposite of what we're looking for on HN. What we want: specific, unpredictable, curiosity-gratifying conversation.
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...
https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...
https://news.ycombinator.com/newsguidelines.html
Why do we think it is more valid than Alchemy?
More work is needed to turn quantum enhanced optimization into a practical technology. The design of next- generation annealers must facilitate the embedding of problems of practical relevance.
Like fusion, current progress is promising, but practical results lie in the future?
If quantum computing was not possible, that would mean that everything we thought we knew about physics was hopelessly wrong. Physics would have to be much stranger - we'd have to give up on either locality (i.e. causation being limited by the speed of light), reversibility (closely related to conservation of energy), or nondeterminism in order to make the math work out. There are several papers in the literature that derive quantum mechanics from these or related assumptions (often together with some extra technical detail, that differs between approaches):
https://arxiv.org/abs/quant-ph/0101012 https://arxiv.org/abs/1011.6451 https://arxiv.org/abs/0911.0695
Using rock-solid math and physics I think you can do time travel and speculate about parallel universes, but I don't think we should blindly believe the models.
We should never confuse the map and the territory.
30 years is nothing given how complex the engineering problem is of actually building a working QC. Fusion is much simpler and very well understood and yet it has been more than 50 years since research into fusion power plants has started, and the most optimistic estimates say we'll have the first ever research power plant 30 years from now, if everything goes well (EUfusion's DEMO plant).
Speculating about parallel universes is a different story - even if they exist, we can't travel to them, so there is no sense in claiming that we can somehow harness their power. It's a philosophical question, not a physical one. In contrast, the rules of quantum mechanics (when phrased in terms of density matrices, at least) only describe things we can actually measure and observe, and people actually do test every single aspect of the predictions quantum mechanics makes.
This is false. Time travel to the future is possible and proven[0]. Perhaps you are referring to traveling to the past.
[0] https://en.m.wikipedia.org/wiki/Experimental_testing_of_time...
Backwards time travel isn’t possible in general relativity, though, making very mild and physically plausible assumptions about the spacetime. Same in semiclassical quantum gravity.
Perhaps reality is actually only approximated by the wave function but reality is something else under the surface which means quantum computers may need to be reconsidered.
I highly recommend reading [1] for an overview of the various skeptical viewpoints on quantum computing. Note that, while Aaronson argues most of the initial arguments are trivially wrong, he admits that whether or not error correction is fundamentally impossible is an open question.
[1] https://www.scottaaronson.com/democritus/lec14.html
We can keep solving the engineering challenges, but the universe might simply not allow us to get a return on investment with every additional qubit.
QC in contrast must exist or else QM is fundamentally wrong - not incomplete, but wrong.
We know its possible to have a fusion reactor, we would just have to dump enough mass together to build a sun. Hopefully we can make it more cost effective.
There isn't a natural phenomena that solves a problem by encoding it and (ab)using quantum physics.
https://arxiv.org/abs/nlin/0101014
2. Everything after the first sentence doesn't make sense. Any system that implements a restricted class of unitaries and features local measurement as described by Born's rule can efficiently solve problems in BQP. We have no reason this should apply to 10 qubits and not 1000.
The theory of quantum mechanics has been completely predictive in materials and has resulted in the most accurate predictions in nature (Anomalous magnetic dipole moment to a part in a trillion.) It doesn't matter if the theory is "Macroscopic" or whatever, it's predictive and within the constraints of its predictions you have known computational power up to problems in BQP. Whether BQP=P is an open question, but as is P=NP.
Known problems are things like wiring, materials science/sources of noise -- eventually including things like cosmic rays. It should work and we have no reason to think it shouldn't due to fundamental limitations of the natural world. It is, however, one of the hardest scientific things humans have ever tried to do so far. To say it'll definitely work is hubristic.
[1] https://plato.stanford.edu/entries/qm-collapse/#ContSponLoca...
[2] https://en.wikipedia.org/wiki/Ghirardi%E2%80%93Rimini%E2%80%...
When I talked about a "fundamental limit on how big you can make a quantum computer" I wasn't talking about it's physical size, I was talking about the number of mutually entangled qbits it contains, and the amount of time it is able to maintain those entanglements in a coherent state.
There's apparently noise in the system - how do we know that the reality is different to the model, and what we are putting down to noise is actually reality being different to the model? Kind like Einstein's correction of Newton.
That being said, if physics derisking failed, it seems really unlikely to me that it would be because quantum mechanics itself was wrong as opposed to there just being some unforseen problematic error mechanism that was consistent with quantum mechanics. Finding a flaw in quantum mechanics itself would be a scientific revolution, which would be awesome, but those are also pretty rare.
I also think it's a little telling that people keep thinking it's the philosophically unsatisfying parts of quantum mechanics that are going to break, when entering a new regime. It smells of confirmation bias.
Disclaimer: am on google quantum team.
1: https://youtu.be/VvHh6GoNhy8?t=311
Up to 12 qubits or so.
> Perhaps reality is actually only approximated by the wave function but reality is something else under the surface which means quantum computers may need to be reconsidered.
Possible, that reality is only approximated, and that approximation breaks down at more qubits. However, that would be the more exiting outcome, since then we have an experiment that gives us a window into a post quantum theory.
It's actually from last year, and other groups have done better experiments since then (in particular, experiments doing multiple rounds of correction instead of just immediately smashing the system and analyzing the pieces). It's quite funny to see a big deal being made of this paper now rather than a year ago.