Protecting data now as the quantum era approaches

Startup QuSecure will this week introduce a service aimed at addressing how to safeguard cybersecurity once quantum computing renders current public key encryption technologies vulnerable. It's unclear when quantum computers will easily crack classical crypto – estimates range from three to five years to never – but …

[Anonymous Coward]

Quantum has a simple flaw

Is Schroedinger correct?

Does the atom exist in all states until it is measured, at which point it is in *one* known/measured state?

Does the act of measuring it, set its state? And all the states of everything that it interacted with (and thus must have a known component... when you measure atom1 the whole universe changed to fix up the result due to atoms1 infinite previous interactions), even propagating those changes faster than the speed of light and backwards in time? (i.e. entanglement)

Does this magic happen?

Because Public/Private key encryption requires that Schroedinger be correct. That the magical entanglement thing happens.

Solvers don't work for these problems, you cannot run a solver to decrypt a message, something like this:

DecryptedDelta = Delta(Decrypt(Msg, EncryptionFunction(PrivateKey(A,B,C,D,E,E,F,G,....))). realMsg)

Solve DecryptedDelta=0 by changing A,B,C,D.... to find the private key.

Being NEAR the private key, does not result in the decrypted message being NEARER the decrypted text. The decrypted text is garbage till you have the *actual* real key. Solvers do not work and never will. Its not that you narrow the key down, and as you do that, so the decrypted text becomes more and more readable! The quantum computers of today are solvers.

When you see a Chinese researcher say "its 10000 times faster", thats garbage, because it would need to be infinity faster if it actually worked.

Literally, a Quantum Computer would really have to be in *every* state at the same time, constrained so that the only valid state is the correct result, then you measure that "correct" result. Bingo you have the private key.

What you measure is the net effect of the detector and atom. The atom's state is unknown only because you do not (yet) know the state of the net(detector, atom). It's the detector that's the unknown! Those "successful entanglement" filters you run on your experiments, they are filtering for the detectors. Effectively to ensure that net(detect1, atom) = net(detector2, atom)... then, as if by magic, you find lots of the parameters you measured with detector1(atom) are the same as detector2(atom).....

Is that magic? Are you comedians? Tell me you cannot understand what I just said there. Tell me it's not staring you in the face!?

At some point science has to clean house.

These quantum computers don't work and they never can, because Schroedinger's model was an approximation of the system, not the system itself.

He did not understand what he was looking at, so he modelled a statistical approximation to it.

As long as you throw money at Quantum Research there will be lots of companies who will be there to take that money. This thing of huge value (decoding all encryption) is apparently not worth them spending their OWN money, but gee, they'll take the billions in research you thrown at it.

Underlying that is the unpleasant truth there.... even these huge companies with lots of their own money can see the problem with the physics here.

They see you coming.

[DS999]

Re: Quantum has a simple flaw

When you see a Chinese researcher say "its 10000 times faster", thats garbage, because it would need to be infinity faster if it actually worked

You clearly don't understand how this works. They aren't worried about the encryption being compromised (i.e. AES or whatever) they are worried about the key exchange being compromised. If you are using a key exchange method depending on prime factors, a working quantum computer will make short work of factoring those numbers.

Even if it can't handle enough bits to factor it instantly, if it can handle 10 bits worth of simultaneous states that's 1024x faster factorization than today (assuming it can test the same number of factors per second, which may be a problematic assumption, but leave that aside for now)

So yes quantum computers can speed things up, assuming we can get them to work. Now whether that will ever happen is not known. You are assuming you understand how quantum mechanics works, when no physicist in the world will make that claim. Schroedinger's model might be how things actually work, no one knows for now, including you. It certainly has not been disproven.

So we can either decide to take your view and assume quantum computers will never work, and if they do they will never be able to factor large numbers thus compromising key exchange. Or we can decide to play it safe and use something that would not be vulnerable to quantum computers if/when they appear.

If we play it safe, and are wrong, our data is still encrypted so we lose nothing other than the cost of updating systems (or for most people, simply letting the old systems/software age out be replaced by new ones using quantum resistant cryptography)

If we stick your head in the sand like you advocate, and are wrong, then when quantum computers arrive not only are we left scrambling trying to fix the problem at the time, we also have to be concerned that if data exchanges in the past (years, decades even) were stored by an adversary, all that data can be decrypted. Most of us don't care, if someone can decrypt an ssh session of mine from 20 years ago and get the password I only lose if I'm dumb enough to still be using the same password. Governments and major corporations do have reason to care, and thus a reason to play it safe given that the cost is not all that high and the risk is massive.