Other Directory Sites: SeekWonder | Directory Owners Forum
The following archive was created by hippie-mail 7.98617-22 on Sat Apr 10 1999 - 01:18:26
jei@zor.hut.fi
Fri, 5 Feb 1999 20:55:36 +0200 (EET)
On Fri, 5 Feb 1999, Bruce Schneier wrote:
> At 01:04 AM 2/5/99 -0600, John Kelsey wrote:
> >>At 10:45 AM 2/1/99 -0700, staym@accessdata.com wrote:
> >>>Suppose someone discovers a way to solve NP-complete
> >>>problems with a quantum computer; should he publish?
> >>>Granted, the quantum computers aren't big enough yet, but
> >>>the prospects look bright for larger ones in the near
> >>>future. It would break all classical cryptography.
> >
> >>If he's a Good Guy, yes. It not only would revolutionize
> >>cryptography (sigh - back to the couriers with briefcases
> >>handcuffed to their arms) but would also revolutionize whole
> >>areas of mathematical practice - there are a _lot_ of
> >>NP-hard problems with real-world applications.
> >
> >Yeah, I was thinking this, too. Does anyone know how large
> >the impact of this would be? Like, would the costs of Fed
> >Ex, UPS, etc., go down substantially, because the way they
> >flew their delivery routes became so much more efficient?
>
> Not really. In most real-world situations, you don't need to
> actually solve the NP-complete problem. Reasonable analysis
> techniques quickly get you within a few percent of optimal,
> and that's good enough for FedEx, etc. Phone company
> routing may improve, but probably not by that much either.
IMO, you are underestimating the effect of QCs severely. Everyday
problems like phone and FedEX would not be where the benefits would
be visible. Instead, there are several areas of computational physics
and mathematics that would benefit immensely. The effect of QCs would
be similar to the effect that the introduction of computers had to
science. Or the introduction of writing, paper and pen. Medicine,
molecular physics, processor design, etc, all would benefit.
If I had a working QC-design, I would give it to the whole world.
This way, one would probably reap the most benefits. In a few years
time, you would see lots of new medicines, new materials, much more
effective non-QC computers, artificial intelligence and whatnot. I
think, instead of measuring benefits in dollars you get, you should
think about the quality and value of things available to you for the
dollars that you have. You could probably make a very decent living
as the inventor, even if you gave it away. Seeing what Bill Gates
and most other billionaires are doing with all of their billions,
they're still working, because they want to. The number of Dollars
Gates has, isn't making him linearly more happy, or giving him that
much better quality of life or expected number years of life before
he croaks. After a few million $s, the amount of money you have
just won't matter to most people's lives.
For a practical comparison, just think of the effect the Linux OS is
having on the scientific world and where we would be if everyone had
to use Windows. QCs would have a very similar effect wherever heavy
computation is used in designing things. They are the Holy Grail for
computational physicists. Mathematicians would still be able to find
hard to solve problems and new cryptographical methods that would
replace the existing ones. And it isn't to be expected that in every
respect a perfect QC is achievable. Some limitations to it's use and
performance are very likely, and just like now, cryptography would be
designed to be one step ahead of the bad guys and their computational
capabilities and the new crypto-algorithms would take the performance
into consideration, or simply use one-time pad.
The point I am making: The effect of widely available QCs on
crypto-world would be an immense increase in development and researh
for a good few years, after which we would probably be pretty much in
the same situation as we are now. The effect on life-quality, design
and research in general, however, would be immense.
P.S. Note that this is just my oppinions on how QCs would effect
things.
++ J
The following archive was created by hippie-mail 7.98617-22 on Sat Apr 10 1999 - 01:18:26