Paper here (pdf & technical warning). And you can read about the protocol here (pdf & technical warning). Big claim. Not sure I believe it, but at first glance seems plausible. Edit: actually just...
Paper here (pdf & technical warning). And you can read about the protocol here (pdf & technical warning).
Big claim. Not sure I believe it, but at first glance seems plausible.
Edit: actually just as a brief note, this is a bit artificial. No-one's breaking your SSH keys anytime soon.
I want to push back a bit against your edit. It's true that this demonstration (if it checks out) doesn't mean that standard encryption protocols are vulnerable any time soon. But up until now it...
I want to push back a bit against your edit. It's true that this demonstration (if it checks out) doesn't mean that standard encryption protocols are vulnerable any time soon. But up until now it has been extremely difficult to conclusively show there is any task that a quantum computer can do much better than a classical one ever could. Even well known examples like Shor's algorithim for factoring numbers is only better than our best known classical algorithm - it has not been shown that it is better in principle than any other classical algorithm.
Google demonstrating a concrete example of a task that is not feasible classically is a big deal and I wouldn't really characterize it as artificial - it may not immediately address any of the things quantum computing is "known for", but its very important in its own right. Whether or not their claims hold up we will have to see. The paper you linked to looks pretty convincing from my reading but I am not in quantum information so that judgement shouldn't be super highly valued.
Sure that's fair. I agree that this paper looks convincing and is significant; I look forward to seeing more realistic protocols (e.g. for a spin chain or something) thrown at this chip. I just...
Sure that's fair. I agree that this paper looks convincing and is significant; I look forward to seeing more realistic protocols (e.g. for a spin chain or something) thrown at this chip. I just think that it's important to point out that 53 qubits isn't a lot and smashing crypto is still distant.
Having not explored the space, I wonder what the odds are, if a commercially exploitable breakthrough is achieved by Google researchers, that they shroud the whole thing in secrecy and we end up...
Having not explored the space, I wonder what the odds are, if a commercially exploitable breakthrough is achieved by Google researchers, that they shroud the whole thing in secrecy and we end up with severely hindered progress as a result.
This is a big problem in many areas of research as big corporations pull scientists away from public research institutions. Government national security research is another place where...
This is a big problem in many areas of research as big corporations pull scientists away from public research institutions. Government national security research is another place where advancements in quantum computing (and AI) are almost certainly not making it into the general scientific community.
I actually think it's pretty likely the results will be publicized. Efforts in quantum computing heavily intersect with security, and security is a field where everyone (except the NSA sometimes,...
I actually think it's pretty likely the results will be publicized. Efforts in quantum computing heavily intersect with security, and security is a field where everyone (except the NSA sometimes, and even they do it after the blowbacks from the dual_ec backdoor) shares.
After all, it's to Google's detriment as well if they miss something and have major issues with their crypto.
I doubt we'd see something like Volvo giving the 3-point seatbelt for free when they created it because it was such an improvement. It would be sad if this is behind a wall, but I can also...
I doubt we'd see something like Volvo giving the 3-point seatbelt for free when they created it because it was such an improvement.
It would be sad if this is behind a wall, but I can also understand that it probably cost a pretty penny to achieve, so I doubt Alphabet would consider giving that away for nothing for us to enjoy.
Quantum Supremacy simply means that a quantum computer can solve a problem that traditional computers cannot. It sounds a lot more grandiose than it is.
Quantum Supremacy simply means that a quantum computer can solve a problem that traditional computers cannot. It sounds a lot more grandiose than it is.
Paper here (pdf & technical warning). And you can read about the protocol here (pdf & technical warning).
Big claim. Not sure I believe it, but at first glance seems plausible.
Edit: actually just as a brief note, this is a bit artificial. No-one's breaking your SSH keys anytime soon.
I want to push back a bit against your edit. It's true that this demonstration (if it checks out) doesn't mean that standard encryption protocols are vulnerable any time soon. But up until now it has been extremely difficult to conclusively show there is any task that a quantum computer can do much better than a classical one ever could. Even well known examples like Shor's algorithim for factoring numbers is only better than our best known classical algorithm - it has not been shown that it is better in principle than any other classical algorithm.
Google demonstrating a concrete example of a task that is not feasible classically is a big deal and I wouldn't really characterize it as artificial - it may not immediately address any of the things quantum computing is "known for", but its very important in its own right. Whether or not their claims hold up we will have to see. The paper you linked to looks pretty convincing from my reading but I am not in quantum information so that judgement shouldn't be super highly valued.
Sure that's fair. I agree that this paper looks convincing and is significant; I look forward to seeing more realistic protocols (e.g. for a spin chain or something) thrown at this chip. I just think that it's important to point out that 53 qubits isn't a lot and smashing crypto is still distant.
Having not explored the space, I wonder what the odds are, if a commercially exploitable breakthrough is achieved by Google researchers, that they shroud the whole thing in secrecy and we end up with severely hindered progress as a result.
This is a big problem in many areas of research as big corporations pull scientists away from public research institutions. Government national security research is another place where advancements in quantum computing (and AI) are almost certainly not making it into the general scientific community.
I actually think it's pretty likely the results will be publicized. Efforts in quantum computing heavily intersect with security, and security is a field where everyone (except the NSA sometimes, and even they do it after the blowbacks from the dual_ec backdoor) shares.
After all, it's to Google's detriment as well if they miss something and have major issues with their crypto.
I doubt we'd see something like Volvo giving the 3-point seatbelt for free when they created it because it was such an improvement.
It would be sad if this is behind a wall, but I can also understand that it probably cost a pretty penny to achieve, so I doubt Alphabet would consider giving that away for nothing for us to enjoy.
This being Google news gave it a somewhat foul aftertaste but you made me realize what, exactly, it is that bothers me.
Not believing it until I see a Scott Aaronson blog post confirming it.
I mean confirmation of the claims made in the paper, not that a paper exists.
He posted this today (might be worth a separate submission): Scott’s Supreme Quantum Supremacy FAQ!
Edit: posted to ~science.formal here
That was a good read, and cleared up a lot of confusion for me.
Now I can at least pretend to know what the buzz is about.
That's both an ambitious and probably BS claim. Probably.
But to tell the truth, "Quantum Supremacy" sounds like some kind of DARPA black ops program.
Quantum Supremacy simply means that a quantum computer can solve a problem that traditional computers cannot. It sounds a lot more grandiose than it is.