4 ms·

There are a lot of hints about how the Zones work, spread through both books. The overall picture seems to be this: * What we think of as the "laws of physics

There are a lot of hints about how the Zones work, spread through both books. The overall picture seems to be this: * What we think of as the "laws of physics" in any given Zone are actually a kind of VM layer, running on top of some (unknown) physics. * The implementation is done through extraordinarily tiny machines. What we see as quarks, leptons etc (and the laws they follow) are in reality complex machines made of far smaller constituents. * The implementation was probably done by some power-beyond-the-powers that lives in the center of the galaxy. It uses the Zones as a protective device - civilizations in the Slow Zone, the Beyond and the Transcend have only relatively limited abilities, unless they can break through the virtualization layer. (That, incidentally, is why the Blight is such a threat in AFUTD, and why the Zones change so much in response.) If you read both books with this set of ideas in mind, a whole lot of otherwise puzzling scenes and one-off observations in the books make a great deal more sense. There's no reason to call the Zones unscientific.

abecedarius 16y ago
Yeah, I agree except for the last parenthetical remark -- I don't recall any evidence for it in the text. Why would the powers-squared need us to invoke them for their own security? I suppose it was more about protecting the wildlife preserve, to the extent it makes any sense to us.
nikhilgk 16y ago
Can't remember where I read it, but another way to look at the Zones was that it breaks the idea and occurance of a technological singularity from a temporal to a spatial dimension.
endtime 16y ago
Given what we know in the realm of complexity theory, I see no justification for arbitrary thresholds in computing power based on space. Nanomachines can change matter but they can't change math, and the fact of the matter is we know a reasonable amount about what we can compute (and how long it takes) and what we can't. Either P = NP or it doesn't, for example; we don't know which, but we're pretty sure the answer doesn't change dependent on where you're standing. Leaving the galactic plane isn't going to enable me to solve the halting problem.
- michael_nielsen 16y ago
  If you change the underlying physics, computational power can change quite a bit. The best-known example of this is that quantum computers are widely believed to be more powerful than classical computers. An even more striking example, though less well known, is that certain non-linear variants of quantum mechanics allow NP-complete problems to be solved in polynomial time. This result was written up here: http://arxiv.org/abs/quant-ph/9801041 http://arxiv.org/abs/quant-ph/9801041 We could imagine, then, that what's going on is that the Universe really runs according to the laws of one of these non-linear variants of quantum mechanics, but the power-beyond-the-powers has arranged so that at the length scales that can be probed by the races in Vinge's story things look effectively like standard (linear) QM. The thing that bothers me a bit about the Zones: how is it that humans think? We are, after all, just machines. So if AI can't be built, then how is human consciousness possible? I'd be interested to know what (if any) response Vinge has to this. A possible-but-not-very-satisfactory response is that life on Earth is really an artifact from another Zone, and one that goes beyond ordinary Slow Zone capabilities. What prevents us from building AI is that we can't reproduce those capabilities. Seems a bit of a stretch.
  - endtime 16y ago
    I'm no expert on the field, but I believe the optimality of Grover's algorithm strongly suggests that the upper bound on the speedup from quantum computing is quadratic (and I apologize if I'm fudging something here). I have no idea about non-linear QM - really over my head - so I suppose I'll have to concede that point. I agree with you that having conscious carbon but not conscious silicon is arbitrary and silly. Once we understand the brain better we'll surely be able to build one. Again, not something that should depend on location, nanomachines or not.
    - michael_nielsen 16y ago
      The optimality of Grover's algorithm doesn't in any way suggest that the upper bound on speedup is quadratic. That result only applies to one very particular class of problems: unstructured (oracle) search problems. There are other oracle problems for which the speedup is known to be exponential. For non-oracle problems the situation isn't yet clear, although many people believe the speedup will be exponential. That will be true if factoring isn't in BPP, for instance.

abecedarius 16y ago

Yeah, I agree except for the last parenthetical remark -- I don't recall any evidence for it in the text. Why would the powers-squared need us to invoke them for their own security? I suppose it was more about protecting the wildlife preserve, to the extent it makes any sense to us.

nikhilgk 16y ago

Can't remember where I read it, but another way to look at the Zones was that it breaks the idea and occurance of a technological singularity from a temporal to a spatial dimension.

endtime 16y ago

Given what we know in the realm of complexity theory, I see no justification for arbitrary thresholds in computing power based on space. Nanomachines can change matter but they can't change math, and the fact of the matter is we know a reasonable amount about what we can compute (and how long it takes) and what we can't. Either P = NP or it doesn't, for example; we don't know which, but we're pretty sure the answer doesn't change dependent on where you're standing. Leaving the galactic plane isn't going to enable me to solve the halting problem.

michael_nielsen 16y ago
If you change the underlying physics, computational power can change quite a bit. The best-known example of this is that quantum computers are widely believed to be more powerful than classical computers. An even more striking example, though less well known, is that certain non-linear variants of quantum mechanics allow NP-complete problems to be solved in polynomial time. This result was written up here: http://arxiv.org/abs/quant-ph/9801041 http://arxiv.org/abs/quant-ph/9801041 We could imagine, then, that what's going on is that the Universe really runs according to the laws of one of these non-linear variants of quantum mechanics, but the power-beyond-the-powers has arranged so that at the length scales that can be probed by the races in Vinge's story things look effectively like standard (linear) QM. The thing that bothers me a bit about the Zones: how is it that humans think? We are, after all, just machines. So if AI can't be built, then how is human consciousness possible? I'd be interested to know what (if any) response Vinge has to this. A possible-but-not-very-satisfactory response is that life on Earth is really an artifact from another Zone, and one that goes beyond ordinary Slow Zone capabilities. What prevents us from building AI is that we can't reproduce those capabilities. Seems a bit of a stretch.
- endtime 16y ago
  I'm no expert on the field, but I believe the optimality of Grover's algorithm strongly suggests that the upper bound on the speedup from quantum computing is quadratic (and I apologize if I'm fudging something here). I have no idea about non-linear QM - really over my head - so I suppose I'll have to concede that point. I agree with you that having conscious carbon but not conscious silicon is arbitrary and silly. Once we understand the brain better we'll surely be able to build one. Again, not something that should depend on location, nanomachines or not.
  - michael_nielsen 16y ago
    The optimality of Grover's algorithm doesn't in any way suggest that the upper bound on speedup is quadratic. That result only applies to one very particular class of problems: unstructured (oracle) search problems. There are other oracle problems for which the speedup is known to be exponential. For non-oracle problems the situation isn't yet clear, although many people believe the speedup will be exponential. That will be true if factoring isn't in BPP, for instance.