3 ms·

I think a really good example of this is boson gasses versus fully connected neural networks (phase only). A cold enough boson gas will undergo a phase transiti

by jayajay 10y ago

I think a really good example of this is boson gasses versus fully connected neural networks (phase only). A cold enough boson gas will undergo a phase transition where each particle falls into the ground state. A fully-connected strongly coupled random phase network will lead to each neuron having the same phase (i.e. same state). You would readily call a neuron matrix "self-organizing" or "swarmy", since it leads the EEG signals we measure, but how is that any more self-organizing or swarmy than the Bose-Einstein condensate if the math is the same? The premise is that it isn't. That's why it seems silly to claim that this paper's statistical system is not self-organizingy or swarmy enough, and that "real" self-healing systems are "more" self-organizingy and swarmy.

slacka 10y ago
The way biological neurons self-organize into a neural network could correctly be described as swarm behavior. In fact there is quite a bit of research into this and it's amazing how during development you can move a neuron from one part of the brain into another area, and it will self-correct and take up the proper role in the new region. There are complex emergent properties here, which your Bose-Einstein condensate lacks. Which is also why it has not practical use beyond theoretical research. I subscribe to the cellular automaton interpretation of quantum mechanics, that our most fundamental laws work like the game of life. A basic part of their program is to increase entropy. This results in homogenous and lack of complexity. These is where your counter point falls apart. Your Bose-Einstein condensate is less complex. Unless we could change the laws of our universe, we have no way to program its behavior and is therefore not useful. So if you must, I will quality my original assertion that it will be a programmable swarm-like behavior. Although whether it be from computer code or DNA, this is a characteristic than any traditional swarm behavior possesses.
- jayajay 10y ago
  > There are complex emergent properties here, which your Bose-Einstein condensate lacks. Which is also why it has not practical use beyond theoretical research. I'm sure someone said that "doping insulators has no practical use beyond theoretical research", too. > Unless we could change the laws of our universe, we have no way to program its behavior and is therefore not useful. Something is useful if and only if we can "program" its behavior. Do you think this assertion is true? Would you consider an iron hammer useful? What about bone tools?
- jayajay 10y ago
  Honestly mate, I appreciate the conversation and your response, but this is starting to reek of quantum spirituality, which is fine, but you're presenting ideas which I can't argue against. Regarding interpretations of quantum mechanics, the interpretation is unimportant -- the math is the same is the same is the same. If you want to think about particles as being controlled by fairies who are locked up in hidden dimensions, that's kinky and totally fine -- as long as we use the same math, I don't really care. If your fairies end up predicting new phenomenon that exist, which are not predicted by the Born/Copenhagen interpretations, sign me up -- I'll go to wonderland. Until then, let's leave interpretations out of our conversation. My point in bringing up boson gasses and neural networks was to show that two vastly different systems at different scales can both exhibit self-organization described by similar mathematics. I am not saying those systems are equivalent, but I am saying that self-organization is a very versatile term and that throwing it around with no context warrants more content.
  - slacka 10y ago
    > reek of quantum spirituality... particles as being controlled by fairies who are locked up in hidden dimensions With the cellular automaton interpretation everything could be calculated and is deterministic. More importantly, there are no silly paradoxes like cat's being both alive and dead at the same time until it is observed. The universe at its most fundamental level are bits of information, or automaton. The speed of light is the speed of causality. It's the clock-rate of our reality. No need for fairies, multiple worlds, or collapsing probability waves that lead to paradoxes. > that self-organization is a very versatile term and that throwing it around with no context warrants more content. No argument that you could "cheat" by going low enough to claim anything is a result of self-organization. But it's not cheating if the actual mechanism is. For example, saying that cells in a neural network are self-organizing is not cheating. Say that a neural network is self-organizing because of some fundamental laws of physics that’s many layers down is cheating. > how that two vastly different systems at different scales can both exhibit self-organization described by similar mathematics. That is the unreasonable beauty of mathematics Finally, yes a hammer is not very useful if you have no way to control its behavior. While your magical hammer might provide a plethora of data for people who study it. You need to control what the hammer hits to use it do useful work. There might be some use from this self-healing tech or any of the past claims. It's the hype that I object to. Then again, I also object to people calling Telsa's cars "self-driving".

by jayajay 10y ago

slacka 10y ago

The way biological neurons self-organize into a neural network could correctly be described as swarm behavior. In fact there is quite a bit of research into this and it's amazing how during development you can move a neuron from one part of the brain into another area, and it will self-correct and take up the proper role in the new region. There are complex emergent properties here, which your Bose-Einstein condensate lacks. Which is also why it has not practical use beyond theoretical research. I subscribe to the cellular automaton interpretation of quantum mechanics, that our most fundamental laws work like the game of life. A basic part of their program is to increase entropy. This results in homogenous and lack of complexity. These is where your counter point falls apart. Your Bose-Einstein condensate is less complex. Unless we could change the laws of our universe, we have no way to program its behavior and is therefore not useful. So if you must, I will quality my original assertion that it will be a programmable swarm-like behavior. Although whether it be from computer code or DNA, this is a characteristic than any traditional swarm behavior possesses.

jayajay 10y ago
> There are complex emergent properties here, which your Bose-Einstein condensate lacks. Which is also why it has not practical use beyond theoretical research. I'm sure someone said that "doping insulators has no practical use beyond theoretical research", too. > Unless we could change the laws of our universe, we have no way to program its behavior and is therefore not useful. Something is useful if and only if we can "program" its behavior. Do you think this assertion is true? Would you consider an iron hammer useful? What about bone tools?
jayajay 10y ago
Honestly mate, I appreciate the conversation and your response, but this is starting to reek of quantum spirituality, which is fine, but you're presenting ideas which I can't argue against. Regarding interpretations of quantum mechanics, the interpretation is unimportant -- the math is the same is the same is the same. If you want to think about particles as being controlled by fairies who are locked up in hidden dimensions, that's kinky and totally fine -- as long as we use the same math, I don't really care. If your fairies end up predicting new phenomenon that exist, which are not predicted by the Born/Copenhagen interpretations, sign me up -- I'll go to wonderland. Until then, let's leave interpretations out of our conversation. My point in bringing up boson gasses and neural networks was to show that two vastly different systems at different scales can both exhibit self-organization described by similar mathematics. I am not saying those systems are equivalent, but I am saying that self-organization is a very versatile term and that throwing it around with no context warrants more content.
- slacka 10y ago
  > reek of quantum spirituality... particles as being controlled by fairies who are locked up in hidden dimensions With the cellular automaton interpretation everything could be calculated and is deterministic. More importantly, there are no silly paradoxes like cat's being both alive and dead at the same time until it is observed. The universe at its most fundamental level are bits of information, or automaton. The speed of light is the speed of causality. It's the clock-rate of our reality. No need for fairies, multiple worlds, or collapsing probability waves that lead to paradoxes. > that self-organization is a very versatile term and that throwing it around with no context warrants more content. No argument that you could "cheat" by going low enough to claim anything is a result of self-organization. But it's not cheating if the actual mechanism is. For example, saying that cells in a neural network are self-organizing is not cheating. Say that a neural network is self-organizing because of some fundamental laws of physics that’s many layers down is cheating. > how that two vastly different systems at different scales can both exhibit self-organization described by similar mathematics. That is the unreasonable beauty of mathematics Finally, yes a hammer is not very useful if you have no way to control its behavior. While your magical hammer might provide a plethora of data for people who study it. You need to control what the hammer hits to use it do useful work. There might be some use from this self-healing tech or any of the past claims. It's the hype that I object to. Then again, I also object to people calling Telsa's cars "self-driving".