That's about what I remember from reading a summary of this story (or
novel?) several years ago. I'd like to track it down.
Thanks!
Bob Hearn
Karl M. Syring
This was discussed, with an almost identical description, in late December.
The subject of the thread was "YASID: Easy armageddon."
Unfortunately, I don't think anyone ever actually identified the story :).
The later editions omitted the afterword, which
is available here:
http://faculty.washington.edu/jcramer/EBridge/EOS.html
I'm not sure this the same story, but aspects of it remind me of
Michael Burstein's "Broken Symmetry", in Analog, February 1997. Except
in that story they end up communicating with other parallel worlds in
which the accelarators are working. (And the accelerator is the SSC,
begun and then abandoned for budgetary reasons in our world.)
If you don't mind, what's your topic?
--
Nancy Lebovitz http://www.nancybuttons.com
http://livejournal.com/users/nancylebov
My two favorite colors are "Oooooh" and "SHINY!".
If his answer includes "Fools! I'll show them!" or "Muahahahahahaha!", I'd
recommend calling any square-jawed steely-eyed heros that you know.
-dms
Actually, it's the computational complexity of games and puzzles. But
there's a section on possible physics that would enable the existence
of perfect game players. There's an existing idea of "anthropic
computation", that's been explored a bit in the CS literature - the
idea is that to solve a problem that it's hard to guess the answer for,
but easy to test when you have the right answer (technically, an
NP-complete problem), you can do the following: flip some quantum bits
to generate a random answer. If the answer works, great. If it doesn't,
then kill yourself. Now, you're guaranteed to only experience timelines
in which you got lucky and guessed the right answer. (You also have to
not kill yourself, with some extremely tiny probability, to ensure that
if there's no answer you will experience not finding it and surviving.)
This works (in a theoretical sense!) only for NP-complete problems,
though. To solve harder problems, one needs a more generalized notion
of anthropic computation. One such would involve a similar scheme to
the above, but instead of killing yourself, you destroy the entire
universe. The problem with anthropic computation as described is that
it only works for you - those around you will almost certainly just see
you die. But if you can take the whole universe with you, then only
timelines that are good for everyone are allowed to exist. Technically,
this enables one to solve much harder problems, and in particular to
allow perfect games players, in a certain sense, to exist.
Bob
Aha! Go to Las Vegas, bet your life savings, and if you lose,
kill yourself! Or better... destroy the universe, of course.
Cut out all that "double your bet" nonsense.
Wayne Throop thr...@sheol.org http://sheol.org/throopw
> ...But
> if you can take the whole universe with you, then only timelines
> that are good for everyone are allowed to exist. Technically,
> this enables one to solve much harder problems, and in
> particular to allow perfect games players, in a certain sense,
> to exist.
Hmm... upside: perfect games players. Downside: destruction of
countless universes.
So, any square-jawed steely-eyed heros in the house?
Mike
--
Michael S. Schiffer, LHN, FCS
msch...@condor.depaul.edu
There's a Greg Egan story (possibly the beginning of _Quarantine_) with
a premise very like that except that the problem was some secret
agentish task.
>
>This works (in a theoretical sense!) only for NP-complete problems,
>though. To solve harder problems, one needs a more generalized notion
>of anthropic computation. One such would involve a similar scheme to
>the above, but instead of killing yourself, you destroy the entire
>universe. The problem with anthropic computation as described is that
>it only works for you - those around you will almost certainly just see
>you die. But if you can take the whole universe with you, then only
>timelines that are good for everyone are allowed to exist. Technically,
>this enables one to solve much harder problems, and in particular to
>allow perfect games players, in a certain sense, to exist.
Practicality would spoil the beauty of it?
Yes but in _Cosm_, the accelerator experiment **creates** a baby universe
rather than destroying ours. Unusual for having a female black physicist as
protagonist.
--
Mike Dworetsky
(Remove "pants" spamblock to send e-mail)
Yes, Quarantine is a bit like that. The amusing thing about Quarantine,
as I recall, was that there was some alien or cosmic species that
relied for their existence on the universal wavefunction not being
collapsed too often, and those damn humans were screwing things up by
making measurements and collapsing the wavefunction all over the place.
Or something like that. It was fun, but I didn't really buy that there
was a consistent physics behind it - in fact it seemed to point out the
absurdity of taking the Copenhagen interpretation too literally. I
liked _Quarantine_, but I prefer the other two of his "subjective
cosmology" book, _Permutation City_ and _Distress_.
Bob
Karl M. Syring
I'm 99% certain that you're talking about "The Doomsday Device" by John
Gribbin (Analog, February 1985). IIRC, the punchline explained (among
other things) why we haven't destroyed ourselves by nuclear winter...
Jeffs
Why not just send the information back in time that the computation
doesn't work--a reset of the universe to a different state as in Thrice
Upon a Time.
: "Gene Ward Smith" <genewa...@gmail.com>
: Why not just send the information back in time that the computation
: doesn't work--a reset of the universe to a different state as in
: Thrice Upon a Time.
Because everybody knows collapsing the false vacuum and multi-words
interpretation of QM is Hard Science Fiction, while time traveling
particles is airy mush-brained treknobabble speculation.
Though note, I'm expecting the Eschaton uses that computational strategy.
That's it!!! Thank you!!!
Wow, the story is only 4 pages long. And between the two of us we
pretty much described the entire content.
Bob
Yes, that's also been studied - CTC (closed timelike curves)
computation. That does let you solve problems harder than NP-complete.
It gets you up to PSPACE. I'm looking for some kind of physics that
will let you solve harder problems still. However, it's not clear that
the "doomsday computation" or "universal anthropic computation" I had
in mind based on this story would actually enable this.
Bob