5 threads in 1

Joseph R. Kiniry (kiniry@cs.caltech.edu)
Tue, 17 Jun 1997 10:00:33 -0700 (PDT)


Robert Harley writes:
> JoeK wrote:
> Subject: Re: Physics of computing
> >how many forkers subscribe to the theory of quantum cryptography
>
> Not me. It has recently been shown to be fundamentally flawed.
> Supposedly two parties could send particles back and forth in states
> which are unknown in a strong sense: you know they are unknown because
> if your adversary observed a particle before sending it that would
> collapse its wave-function and you could tell that that particle was
> compromised. A generic attack is to create pairs of particles with
> correlated states (but don't observe them), send one of each pair and
> keep the other in a cardboard box under your bed. Everything goes
> according to plan but after the protocol is finished you can open your
> box and observe the states of the particles you kept thereby
> discovering those of the ones you sent (which were demonstrably
> unknown to you when you sent them and you were not supposed to be able
> to find out anything about them after having sent them).
>
> >"Even if experiments cannot yet tackle the measurement problem fully,
> >they have much to contribute to a very hot field: quantum computing.
> [...]
> >how does a machine "process" with bits that are both 0 _and_ 1?
>
> I've got a computer in my pocket. Most people call it a coin. When I
> want to solve a problem with an n-bit answer I flip the coin n times
> and each time write down 0 for heads, 1 for tails. The chance of
> getting the answer is exponentially small in n, but if n is less than
> 15 or so and I'm very patient, it works OK. Usually my Alpha is faster.

sure, but you don't know when you have the _right_ answer, you only
know that for a given coin-flip series, you have a 1/n chance of that
particular one being right.

oh yes, and your alpha is usually faster than most any computing
device on the planet. that's like saying "all my kids are bald" or
"all good micro$oft software is well-written".

> A quantum computer is a (non-human) device that flips a coin and
> records the answer in exactly the same way. As long as I don't
> observe the output, there is a superposition of all possible outputs
> including the correct one. The only problem is that the correct one
> is an exponentially weak component of the wave-function.
>
> The only thing quantum-computation guys haven't quite figured out yet
> is how to extract that signal reliably for n more than about 15.

especially when that signal is recorded in the spin state of some M
atoms of a cup of hot tea.

> In other words, quantum computation is currently a crock of shit.
>
> That's "on the record".

got it. thanks man. that's "on the record".

> Rohit wrote:
> Subject: Re: I'm collecting spam
> >This, which was only marginally related to begin with, has the added Rifkin
> >trademark of being WRONG. What part of the gTLD-MOU do you NOT understand,
> >Zippy? There are no gTLDs for sale. [...]
>
> Depends on how many zeros are on the check. Then again there is always
> hostile takeover of the Internet... Adam in commando gear leads an
> insurrection at ISOC, quickly followed by a coup d'e'tat at IETF...
> piece of cake! Hey, now we know why he's been working out so much!

adam needs to work out to bitch-slap those snapper-heads? i think he
should just run through the office naked screaming "the redmonds are
coming! the redmonds are coming!". that'd be enough to make me jump
out the window.

>
> -- Rob.

joe