[Cryptography] Why full-fledged quantum computers might always be five years away

[Peter Gutmann]

Jon Callas <jon at callas.org> writes:

>One of the things I like about the New Scientist reporting is the snark in
>their headlines. Over the last few months, they've been explicitly saying
>"useful" quantum computers, or some other oblique slur.

Cryptographers have been doing that for awhile, although I don't think they
realise what they've been doing.  The term there is CRQC or cryptographically
relevant quantum computer, which would imply that everything so far is
irrelevant.  It's the sort of term I'd expect from a critic of quantum
folderol, not it's supporters.

I see "CRQC" more as a True Scotsman Quantum Computer.  "Sure, we've had zero
results after 25 years of effort with Quantum Computers but just wait until
the True Scotsman Quantum Computer comes along, then you'll be sorry":

  In this ungracious move a brash generalization, such as no Quantum Computer
  has performed legitimate cryptanalysis, when faced with falsifying facts, is
  transformed while you wait into an impotent tautology: if ostensible Quantum
  Computers cannot perform legitimate cryptanalysis, then this is by itself
  sufficient to prove them not Cryptographically Relevant Quantum Computers.

    -- Antony Flew, if he'd written "God & Philosophy" today.

Peter.