The physics is just a hobby for me.
I think they could use a string physicist here though to straighten out their crazy road naming system in Hickory. Obviously designed to get people lost on purpose. 24 different 8th streets in the same city, Main St is a dirt road, street names like 12 St NE Dr CT Ave, different from 12 St SE Ave Dr, streets with two names neither of which agrees with the maps, and major roads that curve all over the place but don't go any place useful. Nothing intersects at right angles either.
They even have at least two of every major chain store because no outsiders can figure out how to get across town. No street numbers on any places either. People drive into this town and are not heard from for days. The natives drive like Jeff Gordon too. Nobody leaves a stop light here without squealing tires. And boy do they have a lot of trashed cars in the junkyards around here too. I don't understand how everyone affords a new car here, but it must be from necessity.
Did I mention Conover? I was looking for a Dollar General store over there, and found a shopping center there that if you drive counter- clockwise around it you can't get into it, but clockwise it does have an entrance. OK, the city planner is going to shoot me if he reads this, enough said.
Just came down here to relax for the winter,... really. Still deciding where I want (and can afford) to move permanently. Maybe Idaho, saw a news article about some old gold miner renting out caves for $25 a month there along some river. Very scenic too. Just looking for a cheap place to retire, maybe build some novel tube amps, work on a physics hobby project - the final technology. Nothing out of the ordinary. Get rigged up for solar power and a vegetable garden before the oil peak hits. Did I say I was an optimist? Well, keepin busy anyway.
Don
I think they could use a string physicist here though to straighten out their crazy road naming system in Hickory. Obviously designed to get people lost on purpose. 24 different 8th streets in the same city, Main St is a dirt road, street names like 12 St NE Dr CT Ave, different from 12 St SE Ave Dr, streets with two names neither of which agrees with the maps, and major roads that curve all over the place but don't go any place useful. Nothing intersects at right angles either.
They even have at least two of every major chain store because no outsiders can figure out how to get across town. No street numbers on any places either. People drive into this town and are not heard from for days. The natives drive like Jeff Gordon too. Nobody leaves a stop light here without squealing tires. And boy do they have a lot of trashed cars in the junkyards around here too. I don't understand how everyone affords a new car here, but it must be from necessity.
Did I mention Conover? I was looking for a Dollar General store over there, and found a shopping center there that if you drive counter- clockwise around it you can't get into it, but clockwise it does have an entrance. OK, the city planner is going to shoot me if he reads this, enough said.
Just came down here to relax for the winter,... really. Still deciding where I want (and can afford) to move permanently. Maybe Idaho, saw a news article about some old gold miner renting out caves for $25 a month there along some river. Very scenic too. Just looking for a cheap place to retire, maybe build some novel tube amps, work on a physics hobby project - the final technology. Nothing out of the ordinary. Get rigged up for solar power and a vegetable garden before the oil peak hits. Did I say I was an optimist? Well, keepin busy anyway.
Don
How about: there is no such thing as causality, just correlation, and the implied direction of the word causality is simply due to our view of time.
One can say the same about the future. I don't see where there's a physical distinction, just a mental one (see my next comment).
First your mention of memory. For some years already the link between entropy and information has been well established. Higher entropy directly corresponds to more information.
Now, we perceive time as flowing in a specific direction, the direction where we have higher entropy in the brain (more memories). I don't see what else there is to it.
Now free will. You have to define the term first. I was surprised to find that many physicists seem to equate it with the inability to prognosticate. Even in a fully deterministic universe, you still cannot predict your future, because you would have to then predict yourself predicting yourself, ad infinitum. QM adds randomness, but you cannot equate that with free will, because there are many other manifestations of such randomness that have nothing to do with the mind. Moreover, from the point of view of information processing ability: from things like the Bekenstein bound and such, it appears we're subject to the types of limits that finite computational machines are (I've argued this here in a past life and I don't want to get into it again; there are tons of references). It is possible to form a correspondence between a human and an abstract finite state computational machine. An argument was put forward a few years ago that unlike the FSMs from theory, a human (as well as a real world computer) is interacting with the environment during information processing, and thus does not have the same limitations of the FSMs which only receive input in the beginning; however, this can be sidestepped by simply including the environment of the subject into the system. Unless there is faster than light communication, the relevant portion of the environment is limited to the subject's light cone (which is also finite in a timelike direction unless the subject is immortal, which is almost certainly impossible given the current cosmology of accelerating expansion, which eventually will move non-gravitationally bound clusterings of matter away from each other faster than the growth of local Hubble volumes, leaving only finite amount of matter within each Hubble volume forever). From these FSMs, a (universal) Turing Machine bounds the set to which one can reasonably equate the subject. Actually, a human is less than a TM because it cannot have infinite memory. Now what about QM nondeterminism? It has been proven long ago that a nondeterministic TM has the exact same limits as a deterministic one (the difference is basically that the nondeterministic one can be much more efficient, a fact that the study of quantum computation exploits).
The point is, there's really no such thing as free will, although perhaps it makes sense from an evolutionary psychology point of view for us to feel as if we did have it.
That statement is dangerously close to panpsychism. Plus, Penrose already tried mixing QM and consciousness in a collaboration with that wacko Hameroff, and made a fool of himself.
It is extreme arrogance, and ignorance of cognitive psychology and neuroscience, for a physicist to assume that those two sciences are insufficient to explain consciousness, and jump unwelcomed to the rescue. While ultimately the mind, like everything else, reduces to physics, no argument has stood up untouched making a reasonable claim that there's anything special from a physics point of view about the mind.
Indeed, the best evidence is that neuroscience, as one would expect, is the correct route. Neural correlates of consciousness are now being studied with interesting progress, and as an example of a neurological theory of consciousness, I want to bring up Damasio's work, as the evidence from brain physiology and pathology is backing his ideas quite nicely.
Back to the previous comments. My views on time, causality, etc., align with Mohrhoff's interpretation, which I find the most straightforward of all I've come across (look around the LANL preprint archive for more info). It belongs to the class of interpretations where QM is limited metaphysically to just being a model/probability algorithm, rather than being equated with 'reality'. Some very interesting comments on this class by Marchildon can be found here, and on the specific interpretation I mentioned, here.
~~~~~~~~
Actually on topic: any comments on the non-unity gain error correction reference I posted? Hawksford doesn't go into much detail, especially regarding implementaiton considerations.
First off, I better say that I AM enjoying the nice weather this winter in Hickory. And the mountains (Wilsons Creek esp.) are great. Just have to memorize the streets here (and buy a 500 HP car with 6 air bags!).
I just did a quick look thru the Hawksford paper and it is interesting indeed. As I suspected, it involves feedforward correction as well as error feedback. The feedforward allows the loop gain around the error correction and gain stage to be reduced below unity, with the feedforward filling in the remaining part to get unity correction, providing a big improvement in amplifier stability and insensitivity to gain settings.
For a gain stage with greater than unity gain, this just requires some fixup factors (equal to nominal stage gain or 1/nom. gain) on the pathes so the total correction ends up being unity.
(for a simple error feedback case without feedforward, this just requires attenuating the error feedback by the gain factor so it gets amplified back up to unity by the gain stage, ie so the error correction signal is just the size of the output error when it reaches the output)
The paper goes well beyond this by finding an optimum distribution between feedforward and error feedback for minimum sensitivity to gain parameters. I need some more time to read this more carefully.
Implementation wise, fig 3.2 seems to be a re-configuration into a design resembling a conventional feedback amplifier but requiring an attenuator on the output (1-b) and gain in the feedforward path.
Maybe it is more practical to stick with fig. 3.1, which is what the error corrected mirror circuit uses by using both outputs from the error detecting diffl. amp. Whatever part one injects into the amp somewhere for feedforward corr. has to be taken into account of by reducing the error feedback somewhat. I'll have to print this article out and read it more carefully.
On conciousness and free will I am admittedly in the dark. My main train of thought is simply generallizing Wheeler-Feynman theory to take into account of 2 time dimensions. Including the time retarded and time advanced solutions of electrodynamics (or QM for that matter) keeps the present compatible with the future as well as the past and in all time directions when generalized to 2T. So one could say the whole universe is already painted and we are just tape recordings playing out TuringM machinations. However, once one achieves practical time machine ability, one could be said to attain a certain limited free will by choice of path thru the cosmos/time. But again, its probably not much different from choosing ones path in traveling the 3D world either.
A speculation on my part: The statistical/uncertainty weakness of QuantMech may be due to our inability to take into account of the (normally) very weak interactions from the other T dimension. Once one has ALL the input data, it may actually be deterministic in theory at least. (I believe others have speculated on this) (the spooky action at a distance QM stuff already disappears with normal Wheeler-Feynman theory)
My best guess for avoiding closed time loops and all the problems they bring with consistency is that they are not possible period. One may be able to cross ones previous path but not "lock onto" it with identical time or space velocities. This I surmise from the fact that string theory is reduceable to a two dimensional theory, so all parts of the universe must be reachable by a 2 dim thread. Closed loops would subdivide the universe into isolated pieces. Of course this means that all electrons are the same etc, and all conciousness the same too. Hello Brother!
(I suggested an even crazier scheme once: everyone has to sleep because they wake up every day as someone different with a monotonous memory of always having been that person. Should give the philosophers some real headaches!)
Don
I just did a quick look thru the Hawksford paper and it is interesting indeed. As I suspected, it involves feedforward correction as well as error feedback. The feedforward allows the loop gain around the error correction and gain stage to be reduced below unity, with the feedforward filling in the remaining part to get unity correction, providing a big improvement in amplifier stability and insensitivity to gain settings.
For a gain stage with greater than unity gain, this just requires some fixup factors (equal to nominal stage gain or 1/nom. gain) on the pathes so the total correction ends up being unity.
(for a simple error feedback case without feedforward, this just requires attenuating the error feedback by the gain factor so it gets amplified back up to unity by the gain stage, ie so the error correction signal is just the size of the output error when it reaches the output)
The paper goes well beyond this by finding an optimum distribution between feedforward and error feedback for minimum sensitivity to gain parameters. I need some more time to read this more carefully.
Implementation wise, fig 3.2 seems to be a re-configuration into a design resembling a conventional feedback amplifier but requiring an attenuator on the output (1-b) and gain in the feedforward path.
Maybe it is more practical to stick with fig. 3.1, which is what the error corrected mirror circuit uses by using both outputs from the error detecting diffl. amp. Whatever part one injects into the amp somewhere for feedforward corr. has to be taken into account of by reducing the error feedback somewhat. I'll have to print this article out and read it more carefully.
On conciousness and free will I am admittedly in the dark. My main train of thought is simply generallizing Wheeler-Feynman theory to take into account of 2 time dimensions. Including the time retarded and time advanced solutions of electrodynamics (or QM for that matter) keeps the present compatible with the future as well as the past and in all time directions when generalized to 2T. So one could say the whole universe is already painted and we are just tape recordings playing out TuringM machinations. However, once one achieves practical time machine ability, one could be said to attain a certain limited free will by choice of path thru the cosmos/time. But again, its probably not much different from choosing ones path in traveling the 3D world either.
A speculation on my part: The statistical/uncertainty weakness of QuantMech may be due to our inability to take into account of the (normally) very weak interactions from the other T dimension. Once one has ALL the input data, it may actually be deterministic in theory at least. (I believe others have speculated on this) (the spooky action at a distance QM stuff already disappears with normal Wheeler-Feynman theory)
My best guess for avoiding closed time loops and all the problems they bring with consistency is that they are not possible period. One may be able to cross ones previous path but not "lock onto" it with identical time or space velocities. This I surmise from the fact that string theory is reduceable to a two dimensional theory, so all parts of the universe must be reachable by a 2 dim thread. Closed loops would subdivide the universe into isolated pieces. Of course this means that all electrons are the same etc, and all conciousness the same too. Hello Brother!
(I suggested an even crazier scheme once: everyone has to sleep because they wake up every day as someone different with a monotonous memory of always having been that person. Should give the philosophers some real headaches!)
Don
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.