The case against Google's claims of "quantum supremacy"
gilkalai.wordpress.com
gilkalai.wordpress.com
To answer your question on why the hyped fantastic claim, as you must know, the people who provide the funds for quantum computing research almost certainly do not understand the research they are funding, and need as feedback a steady stream of fantastic “breakthroughs” to justify writing the checks.
This has made QC research ripe for applied physicists who are skilled in the art of bullshitting about Hilbert Spaces. While I don’t doubt the integrity of a plurality of the scientists involved, I can say with certainty that approximately all of the people working on Quantum Computing Research would not take me up on my bet of $2048 that RSA2048 will not be factored by 2048 —- and would happily accept $204,800,000 to make arrays of quantum-related artifacts. Investors require breakthroughs or the physicists will lose their budget for liquid gases — certainly exceeding $2048.
While there might be interesting science discovered along the way, I think of QC a little like alchemy: the promise of unlimited gold attracted both bullshitters and serious physicists (Newton included) for centuries, but the physical laws eventually emerged that it is not scalable to turn lead into gold. Similarly it would be useful to determine the scaling laws for Quantum Computers. How big of an RSA key is needed before even a QC exceeds the total number of particles in the universe to factor it in reasonable time? Is 2048 good enough that we can shelf all the peripheral number-theory research in post-quantum-cryptography? Let’s not forget the mathematicians too!
I'm an ML/AI researcher. I get similar surveys regularly. I don't reply, neither do my colleagues. The people who reply are a self selected group who are heavily biased toward thinking that AGI will happen soon. Or they have a financial interest in creating hype.
Most of the experts from that report have a direct financial benefit from claiming that this will happen really soon now.
Rigetti $RGTI is up 400% this month. 135% this week alone. He’s right.
First, it says if you can't do accurate rotations then you can't factor. But the premise is false. Quantum error correction allows you to do arbitrarily accurate rotations. Specifically, you can achieve arbitrarily accurate 45 degree rotations around the X and Z axes by using state distillation and gate teleportation [1], and all rotations can be approximated by a sequence of these 45 degree rotations with the error decreasing exponentially versus the length of the sequence [2]. The paper's justification for saying you can't do accurate rotations is that they think quantum mechanics will end up being wrong (see section 4).
Second, even if you assume for the sake of argument that rotations are inherently noisy, the conclusion doesn't actually follow. The mistake the paper makes is to assume the algorithm will use the textbook QFT circuit [3], which uses n^2/2 rotations for an n-qubit QFT, allowing large amounts of error to accumulate. But in practice, because this QFT is at the end of a circuit, you would use the qubit-recycling QFT which only performs n rotations [4]. Alternatively, if rotations were really such a problem, you could perform ~30 rotations to prepare what's known as a phase gradient state. You can then achieve later rotations via the phase kickback of adding a constant into the phase gradient controlled by the qubit you want to rotate [5]. In other words, the paper asserts you need millions of rotations to factor a 2048 bit number when actually you only need dozens. Everything else can be done with Clifford+Toffoli gates.
So yeah, this paper is not at all a convincing takedown of quantum factoring. The formal part focuses on the wrong cost and the informal justification is that they think quantum mechanics is wrong.
[1]: https://en.wikipedia.org/wiki/Magic_state_distillation
[2]: https://www.mathstat.dal.ca/~selinger/newsynth/
[3]: https://en.wikipedia.org/wiki/Quantum_Fourier_transform#/med...
[4]: figure 3 of https://arxiv.org/pdf/1706.07884#page=5 is an example
“Gil’s problem is that the 2019 experiment was long ago superseded anyway: besides the new and more inarguable Google result, IBM, Quantinuum, QuEra, and USTC have now all also reported Random Circuit Sampling experiments with good results.”
We studied the Google 2019 claims and on the way we also developed tools that can be applied for further work and we identified methodological problems that could be relevant in other cases (or better could be avoided in newer experiments). Of course, other researchers can study other papers.
I don't see in what sense the new results by Google, Quantinuum, QuEra, and USTC are more inarguable and I don't know what experiment by IBM Scott refers to. And also I don't see why it matters regarding our study.
Actually in our fourth paper there is a section about quantum circuits experiments that deserves to be scrutinized (that can now be supplemented with a few more), and I think we relate to all examples given by Scott (except IBM) and more. (Correction: we mention IBM's 127 qubit experiment, I forgot.)
“Gil Kalai #23: So we’re perfectly clear, from my perspective your position has become like that of Saddam Hussein’s information minister, who repeatedly went on TV to explain how Iraq was winning the war even as American tanks rolled into Baghdad. I.e., you are writing to us from an increasingly remote parallel universe. The smooth exponential falloff of circuit fidelity with the number of gates has by now been seen in separate experiments from Google, IBM, Quantinuum, QuEra, USTC, and probably others I’m forgetting right now. Yes, IBM’s gate fidelity is a little lower than Google’s, but the exponential falloff pattern is the same. And, far from being “statistically unreasonable,” this exponential falloff is precisely what the simplest model of the situation (i.e., independent depolarizing noise on each qubit) would predict. You didn’t predict it, because you started from the axiom that quantum error-correction had to fail somehow—but the rest of us, who didn’t start from that axiom, did predict it!”
Ouch.
It is disappointing that you endorse Scott's uncalled for and a little juvenile analogy. I think it is a wrong analogy weather I am right or wrong (both on the general question of quantum computation and on the specific question of my evaluation of the Google supremacy efforts).
In any case here is my response to Scott's comment:
"Hi everybody,
1) I found the analogy in #39 offensive and inappropriate.
2) As I said many times, I don’t take it as axiomatic that scalable quantum computing is impossible. Rather, I take the question of the possibility of scalable quantum computing as one of the greatest scientific problems of our time.
3) The question today is if Google’s current fantastic claim of “septillion years beyond classic” advances us in our quest for a scientific answer. Of course, we need to wait for the paper and data but based on our five-year study of the 2019 Google experiment I see serious reasons to doubt it.
4) Regarding our claim that the fitness of the digital prediction (Formula (77)) and the fidelity estimations are unreasonable, Scott wrote: “And, far from being “statistically unreasonable,” this exponential falloff is precisely what the simplest model of the situation (i.e., independent depolarizing noise on each qubit) would predict. You didn’t predict it, because you started from the axiom that quantum error-correction had to fail somehow—but the rest of us, who didn’t start from that axiom, did predict it!”
Scott, Our concern is not with the exponential falloff. It is with the actual deviations of Formula (77)’s predictions (the “digital prediction”) from the reported fidelities. These deviations are statistically unreasonable (too small). The Google team provided a statistical explanation for this agreement based on three premises. These premises are unreasonable as well and they contradict various other experimental findings. My post gets into a few more details and our papers get into it with much further details. I will gladly explain and discuss the technical statistical reasons for why the deviations are statistically unreasonable.
5) “Yes, IBM’s gate fidelity is a little lower than Google’s, but the exponential falloff pattern is the same”
Scott, do you have a reference or link to this claim that the exponential falloff pattern is the same? Of course, one way (that I always suggested) to study the concern regarding the “too good to be true” a priori prediction in Google’s experiment is to compare with IBM quantum computers."
In Kitaev's construction of the high purity approximation to a magic state, he starts with the assumption that we start with a state which can be represented as the tensor product of n mixed states which are "close enough". I don't understand where this separability property comes from. My (very) naive assumption would be that there is some big joint state which you have a piece of, and the information that I have about this piece are n of its partial traces, which are indeed n copies of the "poor man's" magic state.
Can I know more than that? There's lots of stuff in the preimage of these partial traces. Why am I allowed to assert that I have the nicest one?
I recommend just simulating the specific case you're worried about. It's only a 15 qubit circuit; not at all expensive to check. You'll either see it working and stop worrying, or have an amazing concrete counter example to publish.
I think it's clear.
(I am inclined to ignore the claims about quantum supremacy, especially when they're based on random circuit sampling which as you pointed out made assertions that were orders of magnitude off because nobody cares about this problem classically, and so there is not much research effort into finding better classical algorithms. And of course, there's a problem with efficient verification, as Aaronson mentions in his recent post.)
I've seen a few comments of yours where you mentioned that this is indeed a nice result (predicated on the assumption that it's true) [0, 1]. I worry a bit that you're moving the goalposts with this blog post, even as I can't fault any of your skepticism.
I work at Google, but not anywhere close to quantum computing, and I don't know any of the authors or anyone who works on this. But I'm in a space where I feel impacts of the push for post-quantum crypto (e.g. bloat in TLS handshakes) and have historically pooh-poohed the "store now, decrypt later" threat model that Google has adopted -- I have assumed that any realistic attacks are at a minimum decades away (if they ever come to fruition), and very little (if any) of the user data we process today will be relevant to a nation-state actor in, say, 30 years.
If I take the Willow announcement at face value (in particular, the QEC claims), should I update my priors? In particular, how much further progress would need to be made for you to abandon your previously-stated skepticism about the general ability of QEC to continue to scale exponentially? I see a mention of one-in-a-thousand error rates on distance-7 codes which seems tantalizingly close to what's claimed by Willow, but I would like to hear your take.
[0] https://gilkalai.wordpress.com/2024/08/21/five-perspectives-...
[1] https://quantumcomputing.stackexchange.com/questions/30197/#...
Considering that Google's 2019 claim of quantum supremacy was, at the very least, severely overestimated (https://doi.org/10.48550/arXiv.1910.09534) I would wait a little bit before making any decisions based on the Willow announcement.
30 year old skeletons in people’s closets can be great blackmail to gain leverage with.
edit: As I understand it this is a popular way for state actors to "flip" people. Threaten them with blackmail unless they provide confidential information or do some actions.
That is an extreme example but high value information is often stored and secured in ways that are very resistant to theft. Using less secure and/or historical data to gain leverage over those with access to that data is exactly how spies have been doing things for centuries.
b) Astronomy has (had?) the same conundrum: gathering acres of data, most of which will never be examined. (Don't call attention to yourself and hopefully law enforcement will ignore you.) Alas, now we're creating tools for chewing thru big data(s), to spot patterns and anomalies. For better or worse.
The NSA is most likely interested in all data let's be honest. At a bare minimum, in all foreign actor data.
On the other hand, if they're tapping even a 100 Gbps link that's run at 50% average utilization, over the course of a year, that's more than 300 EiB of data. This is a frankly stupid amount of storage for such a tiny cross-section of our actual traffic. And I'm supposed to believe that they actually want to do that for years, storing zettabytes (or even yottabytes, depending on the scale of such a collection effort) of traffic in aggregate, on the off-chance that they have a quantum computer in 30 years? Tape storage might be cheap (on the order of single-digit dollars per TiB), but even at that price, just 1 ZiB is billions of dollars.
Sure, maybe you could reduce those numbers by performing targeted wiretaps, but it's also way easier to just subpoena Google and ask them to provide search history for individuals on certain dates...
Are there good write-ups on the random, sampling problem that would help implement its get started?
What are the top classical algorithms, esp working implementations, for this problem?
Have the classical implementations been similarly peer reviewed to assess their performance?
Google claims to have proved its supremacy with new quantum computer (256 points, 1 year ago, 229 comments) https://news.ycombinator.com/item?id=36567839
Quantum computers: amazing progress, but probably false supremacy claims (126 points, 5 years ago, 73 comments) https://news.ycombinator.com/item?id=21167368
Google Achieves Quantum Supremacy. Is Encryption Safe? (38 points, 5 years ago, 21 comments) https://news.ycombinator.com/item?id=21100983
Google claims to have reached quantum supremacy (114 points, 5 years ago, 21 comments) https://news.ycombinator.com/item?id=21029598
Google Engineers Think This 72-Qubit Processor Can Achieve Quantum Supremacy (91 points, 7 years ago, 42 comments) https://news.ycombinator.com/item?id=16543876
Google plans to reach a Quantum Computing milestone before the year is out (147 points, 8 years ago, 49 comments) https://news.ycombinator.com/item?id=14171992
In 2024 Google used 67 quantum bits to solve the same/similar random circuit sampling benchmark, that they used 53 bits in 2019. The discussion from 2019 on what is the relevance (if any) of solving random circuit simulation problems, is equally valid today.
Maybe in 2030 Google or someone will use 200 or 400 bits to solve an even bigger instance of the random circuit simulation benchmark problem, and then we get to have this same discussion once again.
Still nothing really new here. I'm doubtful anything will convince the author short of shor's algorithm actually factoring large numbers.
As pointed out in [57], there has never been a genuine implementation of Shor’s algorithm. The only numbers ever to have been factored by that type of algorithm are 15 and 21, and those factorizations used a simplified version of Shor’s algorithm that requires one to know the factorization in advance.It is completely in line of the "The Case Against..."
(From my reading/understanding of it, for a while there's been little point in trying to make a quantum computer big enough to do such work, because the individual parts would not work well enough for it to have any chance of success, while this result primarily is showing that they're now at the cusp of the predicted tipping point where the qubits have a low enough error rate that building a larger system out of them has a hope of working. That's the big news in google's recent announcement, not them pushing up the numbers in this somewhat contrived benchmark)
To remind us of the large difference between the public image vs the reality of quantum computing.
Expectation: From the news, people are getting a feeling that quantum computing revolution is almost behind the corner, and current cryptosystems will soon become irrelevant. People are (here, in Hacker News) asking "is our society ready for this", "what could I do if I had this chip at home", "existing cryptography technology is in danger", "is it time to 100x key length on browsers".
Reality: Using Shor's algorithm to factor 15 = 5×3 is still far beyond the reach of current quantum computers. We can factor 15 = 5×3 and even 21 = 7×3 if we cheat by eliminating those branches from the calculation that are not on the happy path the correct answer.
You mean like Skunkworks when they designed the SR-71?
Or DARPA with Atlas?
Or the DoD with the global GPS network?
Or the NSA's ANT catalogue?
Yeah, the government has never had any issues outpacing the private sector. You only ever find out about it when they want you to. Usually that happens 5 to 30 years after the market has already caught up.
Airlines don't want high altitude supersonic spy planes, for example.
And GPS is a collective action problem: no one wants to pay for it, but we do all benefit from it being ambiently around.
The great quantum computing nightmare, from the NSA point of view, is someone sidestepping out of nowhere with a viable machine that works in an unexpected and easy to reproduce way.
Edit to add: see also https://en.m.wikipedia.org/wiki/DNA_computing which while bounded in the same sense as conventional machines would still be a game changer.
Like other alternate forms of computing, the systems we build on CPUs today are truly hard to beat, partly because people are trained on those systems, partly becaus the high performance libraries are there, and partly because the vendors got good at make stupid codes run stupid fast.
At this point I cannot see any specific QC that could be used repeatedly for productive work (material simulations, protein design) that would be more useful than a collection of entirely conventional computing (IE, a cluster with 20K CPUs, 10K GPUs, 100PB of storage, and a fast interconnect). Every time I see one of these "BMW is using a quantum computer to optimize logistics" and look more closely, it's obvious it's a PR toy problem, not something giving them a business edge.
It’s sad to imagine the amount of smart nerds out there whose only actual experience with women is from the fking honeypots that Langley or ft mead et al use to make sure that some prospective AI talent on discord isn’t about to release a bioweapon. Clearly a lot of AI talent overlaps with incels and adjacent communities (see civit.ai as an example of this), It’s common knowledge that field agents skew female since they’re less suspected by patriarchal idiotic targets (they tout this in recruiting for DEI reasons). It’s probably smart for AI startups to tell their male coworkers to be on the lookout for random attractive women trying to talk to you. The US military and foreign service et al specifically warns its members about this and it is a clear and present danger.
And FYI, if the glowies aren’t doing this, they’re not doing their jobs, since the risk of some crazy open source AI person deciding to lone wolf society is rather high, at least according to the less wrong folks (that community I bet is also crawling with spooks). AI is so full of industrial and business espionage that I get scared just being in the space.
I know this is happening too because the private version of it, expert networks, are extraordinarily lucrative and rely on basically laundering of material non public information with plausible deniability. The “experts” on an “expert network” are basically private business spooks, ackin to a private investigator targeting a business.
Possible, but I think zero chance they have anything more practical than Google Willow, which is itself completely impractical for anything except quantum computing research.
One of their big needs for production operations is having 24/7 support through christmas break, which favors companies like IBM.
Very few labs around the world are tasked with testing results instead of trying to produce new science, and in this specific case, only google has access to the device being built and the computation tested by them impossible to verify with classic computers, unlike Shor's algorithm which is trivial to test with known primes.
The point is to ensure that researchers are actually doing what they think they are doing and not deluding themselves. Its not meant to prevent outright fraud.
Well, from one possible attack vector, anyway. They can still point lasers at your windows.
Hmm... What exactly are you transmitting in such a scenario? What's the physical layer protocol? I.e. what are the wires made of and what flows through them?
oh and all other passwords/pass phrases/secrets might get broken too, if they're not yet based on quantum save algos.
One should consider the possibility that Gil Kalai is a similar sort of skeptic making well founded objections to weak arguments, but that nonetheless in the long run the extraordinary claims will turn out to be more or less correct. It's true that those involved in plate tectonics didn't have Bitcoin to sell you, but they were looking for oil.
He's very much claiming this, e.g. https://www.math.ucdavis.edu/~deloera/TEACHING/VIDEOS/Kalai-...
It's a shame we cast their role in the history of science in such a negative light, as if science was a game to be won or lost instead of a collective process where a clear path forward only appears in hindsight (only to be proven wrong again sometime later) .
Oh yeah, and that anecdote you told about tectonic plates screams survivorship bias; that's the problem with anecdotes.
Very good point. For every tectonic plates theory or heliocentric system or H. pylori causing ulcers there are thousands of claims that are plain wrong. Statistically speaking, knowledgeable critics acting in good faith (eg, not having strong conflict of interests) are correct with the overwhelming probability.
You get this all the time with perpetual motion machines. The near certainty of the claim being false leads to confident dismissals that go 'blah, blah, laws of physics, blah blah thermodynamics, therefore can't happen'
The real question to be asking about a claim of a perpetual motion machine is 'Where does the new energy come into being?'.
Citing laws of physics won't help you because any claim to have made a perpetual motion machine is implicitly claiming to be a proof by counterexample that one of those laws is wrong.
Citing the laws of physics in this case is the shorthand way to point the overwhelming number of proofs by example that the laws are correct.
If your law is all liquids flow off a ducks back. Water off a ducks back does not prove it. Acid off a ducks back disproves it. https://i.imgflip.com/7waajp.png
I don't think it is a matter of a shorthand. I think it is because humans have a tendency to express a strong opinion when they intend to express that a weaker opinion is strongly held. Citing laws of physics does not say "Your perpetual motion machine won't work" but rather "I am confident that your perpetual motion machine will be shown to not work".
A single device, made in some garage, that appears to disprove it is simply not rigorous enough to prove anything and isn’t worth third party investigation until the creator has shown they’ve ruled out possible explanations.
Galileo's initial results could not predict many things that the old geo-centric models could predict for centuries.
This is almost inevitable with a groundbreaking new framework, but the skeptics aren't wrong to point it out, it's up to the supporters to show it can do what old model can do and more, which Galileo was never quite able to show in his lifetime, if I recall correctly.
Again, it's easy to look back and talk about how these "irrelevant nitpicks" of tectonic plates and helio-centrism were "wrong" or "irrelevant", but that's just not how science works, you don't get to skip over the details when you present a theory that undermines everything we know, that's just crackpot behavior.
When i was an undergrad student 20 years ago I did hear that "soon, quantum computing will change the world", and yet here we are, every year someone builds a new machine but no one has yet to factorize that 21 = 7x3 in a general way.
Whether we'll be able to replicate it profitably at small scale is a question.
I'll keep this short.
- Google’s Willow quantum chip significantly outpaces current supercomputers, solving tasks in minutes that would otherwise take billions of years.
- Hypothesis: Accelerating advancements in tech and AI could lead to quantum supremacy arriving sooner than the 2030s, contrary to expert predictions.
- Legacy banking systems, being centralized, could transition faster to post-quantum-safe encryption by freezing transfers, re-checking processes, and migrating to new protocols in a controlled manner.
- Decentralized cryptocurrencies face bigger challenges:Hard forks are difficult to coordinate across a decentralized network.
- Transitioning to quantum-safe algorithms could lead to longer transaction signatures and significantly higher fees, eroding trust in the system.
- If quantum computers compromise current cryptography, tangible assets (e.g., real estate, stock indices) may retain more value compared to digital assets like crypto.
Thoughts?
This is the point that was disputed in the article, and you're instead taking it for granted.
About the other points, quantum computers are massively different from classic ones, so much that there are very few algorithms for them. For example, GPUs are faster at matrix multiplication because it can be implemented independent parallel threads, but they suck at other problemas. A quantum computer is good for running quantum algorithms [1], of which there are very few at the moment, and most of them are useful for simulating quantum physics. It is not a "faster" classic computer in any way.
From what I know about banking world, though second hand, having been working in payment processing systems I can say with confidence that it’s not the compute that’s holding them back.
Shor's is a completely different matter entirely: the difference between exponential and linear time is so huge that even a comparatively tiny QC (only a few million qubits) would significantly outpace the largest classical supercomputers put together on this specific problem.