[extropy-chat] Bluff and the Darwin award

[Russell Wallace]

On 5/17/06, KAZ <kazvorpal at yahoo.com> wrote:
>
> Actually, this is untrue. It is akin to saying "if you're going to count
> to 100, you need 100 items to do the counting on". No, for a computer to
> count to 255 it only needs 8 bits. 24 bits, and it's counting into the
> millions. Likewise, though more subtlely, you can track any level of
> complexity with less than the actual components therein.


As I understand it, this isn't the case with ab initio quantum mechanics
calculations - that with QM, the calculation scales not as the number of
particles, or even as the number of pairwise interactions, but as the number
of possible configurations of the system. (I didn't keep links to any online
references, but I remember Drexler explains this in 'Nanosystems'. However,
I'm open to correction if there are any physicists in the audience.)

And, of course, you're assuming the simulation has to be purely
> deterministic. But you can also fudge the data.


Absolutely. That in fact is what molecular modeling systems do; they use
polynomial-time heuristics - educated guesses - about how atoms typically
behave. Sometimes these give the right answer, sometimes they don't. So yes,
you can do this - but the result is, as I said, a complement for laboratory
work. Not a substitute.

But, in reality, there is usually some other information, allowing a more
> rational analysis to be carried out.


And that's exactly what I've been trying to do.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20060517/a98d8999/attachment.html>