From: Anders Sandberg (asa@nada.kth.se)
Date: Wed Aug 06 2003 - 13:00:11 MDT
On Wed, Aug 06, 2003 at 10:28:31AM -0700, Charles Hixsn wrote:
> Anders Sandberg wrote:
>
> >Huh? I get the feeling that you think entanglement links are some
> >kind of communications channels you can send information through?
> >
> I believed that was the proposition. I find it dubious, but
> conceivable.
There is no known such information transmission effect. But I hope we
can squeeze something out of quantum gravity (if spacetime on the Planck
scale is filled with wormholes, there ought to exist shortcuts for
communication - but I fear there might be probabilistic laws showing
that you cannot on average send information faster than light).
> If you were discussing using them as q-bits, ... well, I
> don't really know how to think about that. I REALLY don't. If that is
> ever made easy, it could be much more dramatic than thinking of them as
> communication links. The estimates that I've seen as to how capable
> they would be make me rank them as "Don't try to think past this until
> engineering contraints are more obvious." 32-qbits would be, O, 32
> atoms (plus, of course, a whole lot of support equipment,
> refridgerators, etc.). And large classes of problems would disolve into
> "trivially solved". Large classes. And this change could come *quickly*.
We are already moving towards larger and larger quantum computing
systems. In general N qubits are worth 2^N ordinary bits, *but it
depends on the problem class*. I don't think large classes of problems
will trivially fold; we have had quantum computer theory for several
years now, and Shor's factoring algorithm and Grover's search algorithm
still seem to be the most useful ones found. I doubt one could do (say)
a chess algorithm this way. But we don't need large classes of
algorithms to be speeded up, just some essential classes we can build
computers on top of. Imagine a quantum algorithm that can act as a
general but fast Turing machine - that would be revolutionary!
> >Yes, but this is based on FTL communication, not entanglement. If
> >we can make small wormholes this might work.
> >
> How does one get the negative energy density to stabilize them?
Beats me, but I would take a look at the Casimir effect. It is IMHO the
only macroscopic negative energy effect I know.
> I feel realtively confident that wormhole
> communicaitons will always be expensive. But it might have a very fast
> bit rate. (Would a high speed laser link cause problems with keeping up
> the needed negative engery gradient? I don't know the rules here.)
If you pass too much mass/energy across, you get problems. My
calculations (http://www.transhumanist.com/volume5/Brains2.pdf) said
that a one meter diameter wormhole has a max capacity of 1.56e87 bits/s,
and a nanometer wormhole 1.56e69 bits/s. Quite a good bandwidth.
> Many people understand a lot more about it than I do. I'm basically a
> programmer. And, I must admit, cryptography doesn't interest me. My
> understanding is more at the level of Scientific American and Science
> News. So I know the general outlines of current beliefs on, say,
> entanglement, but not the details or the certainty. I do, however,
> remember the absolute confidence that I heard given to sentences like
> "We'll never be able to see a single atom. Quantuum theory forbids
> it." Which is sort of true...but not really, as quantuum dots and the
> Scanning Electron Microscope showed. The math was solid, but the
> translation into physical models was flawed.
Sure. A lot of stuff gets said with absolute certainty that shouldn't be
said that way. On the other hand, much is said in a doubtful way when it
should be said certainly.
My personal rule is to avoid assuming stuff that would be better than
current physics allows, or when you do it clearly mark it as wishful
thinking.
-- ----------------------------------------------------------------------- Anders Sandberg Towards Ascension! asa@nada.kth.se http://www.nada.kth.se/~asa/ GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y
This archive was generated by hypermail 2.1.5 : Wed Aug 06 2003 - 13:05:46 MDT