Thursday, October 30, 2008

Waking a Mind Node

Includes small remark on reality-simulation.

I had an intuition that a mind node would need to be run for a while to 'wake up' and become not simply non-deterministic, but actually conscious.

And here's why it's true.

The point of a mind node is to violate physics, specifically physical causality, so that I can invoke my proof that consciousness is not physical and propose that a non-physical 'spiffy' stuff closes this causal hole. To do this properly, the probability of a mind node must, unlike every other physical situation, diverge. Now, mathematically we can see immediately that the probability of any particular state rapidly approaches zero, and similarly that if we had a non-deterministic or spontaneous event it would look exactly like a mind node, but this isn't really good enough. We are not building a mind node out of math, but rather out of physics, and it needs a good physical mechanism whereby it can know its supposed to be a mind node in the present, without having to be able to see the future or the past.

This mechanism is by violating the No Infinities Principle. Specifically, in a present moment, when you run a mind node, it cannot tell whether it has been run for 'long enough' to sufficiently diverge, or if it was in fact just built directly into the state its in. (This is the generalized relativity principle.) Except.

Probability is conserved, like energy. It is a real physical quantity. Ergo, it must follow the NIP, as statements like "Q= 64% - lim x->∞ (1/x)" are not physically meaningful. Q=64% but by assumption some physical process subtracted an infinitesimal amount from Q, and because probability is conserved, it added an infinitesimal amount somewhere else. If this somewhere else starts at zero, the infinitesimal doesn't collapse and thus the total probability will not add up to 100%. Instead it's (64+36+1/∞) which using physics-math collapses, but we can't collapse it because we'd immediately have to add back the 1/∞ because that process still exists, but then it collapses....and there's a barber I'd like you to meet.

(Thinking about it, this is yet another proof of the NIP: allowing infinitesimals means we end up adding an infinite number of infinitesimals, which will add up to a finite number, which really badly violates conservation, plus usually the finite number it adds up to is poorly-defined.)

This specific singularity is shielded by the uncertainty principle. Probabilities only add up to 100% on average, actually, not in any particular moment. This unavoidable, fundamental stochastic physics swallows changes below a specific threshold, of which there is a formula to determine. We can consider any probability to be higher than this threshold* or else zero.

*(And probably a quantum multiple. Because probabilities vanish below a threshold, continuous calculation of the probability of a free electron would not add up to 100% - the line simply ends after a certain distance and you can no longer integrate over infinity. You can fix this by quantizing probability, and thus the line flatlines long before it ends, restoring the integral's integrity. Probably - if you have the necessary math you should try it.)

However, a mind node breaks this. After having run for a while, it is in a state that had a lower-than-threshold probabality in its initial condition. This happens because the mind node is continuously diverging states - at some point there are more states than the maximum number of probabilities, (100%/[quantum of probability]) Again, by generalized relativity, physics won't be able to pick which states are 'special' so as to discard them, and all states drop to zero probability. In other words, it becomes something that had a zero probability of existing. Thus, the mind node can't at any moment tell if its been running for some time or if it was just built, except when it becomes an impossible object. (You're awake because you got Echer-ized!) It is at this point the mind node wakes up and must resort to consciousness to continue existing at all. (Or to stop exploding the universe, like the LHC.)

(This may be solved by our future understanding of time relative to probability. It seems possible that this kind of treatment may resolve the arrow-of-time question, but such a solution would bork the above proof.)

This may explain why dreams are hazy - when the mind node first wakes up, it doesn't produce very high-quality consciousness, and thus dreams. It may also be during this time that the brain entangles the state of the interpreter to the state of its sensory organs.

Note that this means the universe as a whole has a zero probability of existing, because it includes mind nodes. This may have some very interesting ramifications. I suggest starting with the fact that before the first instant, there must have been nothing, and therefore, I think, a similarly zero chance of anything existing.

Two incidental remarks; this is a reason to think we're not living in a simulation - the first mind node would have crashed the thing;* and it may very well be that physics is organic and only solves problems when it encounters them, forming new physics on the fly, and thus consciousness and spiffy stuff did not exist before the first mind node broke existing physics. (This idea comes from the same kind of intuition that thought mind nodes must have to 'wake up.')

*(If true, this means the simulation idea is not metaphysics.)

Remember that relativity still plays a role, though, in that in any moment, the pentagon has only five possible states, and thus the mind node has only five possible next steps. (Non-abstract pentagon can have more states, which exponentially acclerates the waking process.) This tension between possible and impossible is what allows the mind node to work as a transphysical process - it mathematically violates physical laws but in any particular quantum of time, follow exactly physical laws. It is non-determinism implemented deterministically. It is consciousness that is physical. It is both, and probably violates the laws of either alone.

Any proposed solution to consciousness must have this same property - it has to be causally connected to physics and consciousness, to nifty stuff and spiffy stuff, and have some reason it can't simply collapse to one or the other, so that we can tell, on this side, whether a thing is conscious or not. There has to be that interaction we can observe.

Even if it turns out that someone does to me what I've done to the physicalists,* and we restore monism, consciousness still has to have this property of causal linkage. There has to be that interaction we can observe, or else it doesn't exist. (The most obvious problem is that epiphenomenalism violates Newton's Third Law.)

*(Which specifically is that I saw a solution to the problem that was completely outside their box.)

I would also like to clarify my remarks about the hole in my proof. Specifically, you can violate relativity and assume an underlying absolute coordinate-type reference system under math itself, and simply declare certain equations 'conscious' by fiat. This works, but it is extremely inelegant and forms a solid brick wall to progress - study of consciousness gets folded into the mathematical study of those equations declared conscious, and the mystery becomes permanent. Stuff is conscious just because it does.

Finally, because I forgot to specifically mention it in the original article, the mind node has the advantage of maintaining relativity - anything which acts conscious can be conscious, regardless of what it is made of, just the way life does not necessarily have to be squishy, cellular, or carbon-based. (The reason life is easy and consciousness isn't is because life's definition is arbitrary because it is not fundamental, while consciousness is very much real, objective, and fundamental.) This fact is contained in my remark that using this theory we can build consciousnesses, if I haven't made a mistake, but I wanted to say it explicitly.

No comments: