I challenge you: your best argument for materialism

[marplots]

The computer may, somehow feel nausea.

But if the claim is that this world is what you would expect a material world to be like and also claim that sensations are computations - then you must be able to demonstrate that some computation must produce the feeling of nausea.

I don't see why, when my claim is that my own thinking is the result of some computation as well. It is perfectly reasonable to say a calculation cannot necessarily derive itself. Didn't Godel demonstrate some such?

In fact, isn't it the case that, despite our best efforts, the details of the quantum realm cannot be calculated, even in principle (at some level)? If uncertainty is a feature of the material world, it cannot be decomposed entirely to binary, step-wise recipes.

However, all is not lost. We can do a kind of reductive anti-calculation. We can take something that does feel nausea and start removing parts to see if the nausea goes away. This, I believe has been done. So, while we may depend on a complex universe to create the feeling of nausea, we can degrade the machine doing the calculation in a manner that corrupts the sensation. And we do this by way of material removal from a biological system (as in blocking nerve impulses) - the "machine" doing the calculation.

 

[RussDill]

In fact, isn't it the case that, despite our best efforts, the details of the quantum realm cannot be calculated, even in principle (at some level)? If uncertainty is a feature of the material world, it cannot be decomposed entirely to binary, step-wise recipes.

If a computer couldn't be used to simulate quantum interactions, the theories wouldn't be very useful. It'd be as if in a physics equation there was an operator just labeled "magic happens". That something would necessarily be more powerful than a Turing machine and would be in computing jargon, an oracle.

No, there isn't a problem with quantum simulations. And as far as randomness, you can just pick one outcome based on calculated probabilities, or calculate each possible outcome.

But as far as calculating beforehand the outcome of a quantum event with certainty, like when a particle will decay is another issue entirely and I think actually depends on your interpretation of quantum mechanics. With MM, by calculating each outcome, you properly predicted the quantum event.

 

[marplots]

But as far as calculating beforehand the outcome of a quantum event with certainty, like when a particle will decay is another issue entirely and I think actually depends on your interpretation of quantum mechanics. With MM, by calculating each outcome, you properly predicted the quantum event.

Ah, but then I can ask you to calculate which of the many worlds we will be in and get the same failure. So long as we agree there is a true random element, then we must agree that that element cannot be calculated - unless we are using a different idea of what "random" means.

My guess is we are saying the same thing in different ways here.

The argument with respect to materialism vs idealism is that the demand for experiences to be "calculable" is unsupported.

 

[RussDill]

Ah, but then I can ask you to calculate which of the many worlds we will be in and get the same failure. So long as we agree there is a true random element, then we must agree that that element cannot be calculated - unless we are using a different idea of what "random" means.

Again, it can be calculated and it was. In MM we actually then go and measure each one of the results. Its just we can't meet up with all those versions of ourselves and compare them. I think there are experiments that do something similar to this though, where the measurement becomes entangled with the measurement.

In any case, it isn't something that presents us with any difficulty in simulating the behavior of a hypothetical quantum system.

 

[marplots]

Again, it can be calculated and it was. In MM we actually then go and measure each one of the results. Its just we can't meet up with all those versions of ourselves and compare them. I think there are experiments that do something similar to this though, where the measurement becomes entangled with the measurement.

In any case, it isn't something that presents us with any difficulty in simulating the behavior of a hypothetical quantum system.

OK, here's a hypothetical system. It consists of one atom of Caesium-137. Calculate when it will decay.

 

[Brian-M]

But something that can only compute cannot reference all of a smooth interval, for example the number of values in the real interval [0,1] is countably infinite, but the number of values in that interval is uncountably unfinite. So a computer can only reference a vanishingly tiny proportion of the values on a smooth interval.

Okay, I see what you mean. But with enough decimal places you get to a point where the quantization error doesn't matter any more.

But it is also believed that the universe isn't continuous in that manner either.

The smallest unit of distance is apparently 1.6162×10⁻³⁵ meters. (Planck Length.)

 

[RussDill]

OK, here's a hypothetical system. It consists of one atom of Caesium-137. Calculate when it will decay.

Too easy, any answer that I give will be a 100% accurate simulation of a hypothetical quantum system.

 

[marplots]

Too easy, any answer that I give will be a 100% accurate simulation of a hypothetical quantum system.

I see. We have a function which can take on any value. A veritable universal function. And, a calculation under that function must, at some as yet unknown point, be correct.

I had rather a different idea of what "to calculate" actually meant.

On the other hand, I was arguing earlier that the only way to do such a calculation was to let the Caesium atom (as part of the material universe) do the calculation for us. So my answer isn't any better: "Wait and see..."

 

[H'ethetheth]

Robin, what you are asking is: If I take my dog for a walk, and then shoot it with a disintegration beam and reduce it to atoms, where did the walk go?

Do you understand why this is not a useful question?

A less funny but in my opinion more apt comparison would be taking a look at a freeze-frame of someone taking his dog for a walk, and wondering why the person isn't moving, or wondering where the walkiness went.

 

[RussDill]

I see. We have a function which can take on any value. A veritable universal function. And, a calculation under that function must, at some as yet unknown point, be correct.

I had rather a different idea of what "to calculate" actually meant.

On the other hand, I was arguing earlier that the only way to do such a calculation was to let the Caesium atom (as part of the material universe) do the calculation for us. So my answer isn't any better: "Wait and see..."

Ok, if you are convinced the atom is doing a calculation. Tell me, in what way can I use the time of the decay of a single Caesium atom as a part of a larger calculation.

I think you are very confused as to what "to calculate" means.

 

[marplots]

Ok, if you are convinced the atom is doing a calculation. Tell me, in what way can I use the time of the decay of a single Caesium atom as a part of a larger calculation.

I'm not sure. But I don't understand why that's important.

I think you are very confused as to what "to calculate" means.

I was using it to mean mathematical symbol manipulation to derive some answer. In this case, an answer calculated before hand would be checked by what actually happens. Unless, as I propose, the "what actually happens" is all you can do to get an answer.

The context of the argument was one in which it is proposed that to count as a valid materialist viewpoint, a process would have to be calculable and hence capable of being mimicked by a computer program. My counterexample was meant to show there are parts of the accepted material world which are not so calculable - in other words, cannot be modeled specifically by a computation.

As I understand it, you can know everything it is possible to know about the current quantum state of caesium-137 and still not be able to model the decay as a discrete event. (By event, I mean something which happens in our universe.)

 

[Brian-M]

A pulse coded, global biased, switched network (for whatever definition you are using for "global biased" in this case) may or may not be a computer depending on how it is set up.

Pretty much irrelevant because if it is a computer then there will be an equivalent algorithm which runs on a register machine.

Unless it's (somehow) doing something non-computable (such as generating true-randomness), there will be an equivalent algorithm which runs on a register machine, regardless of whether or not the switched network actually is set-up to function as a computer.

 

[Belz...]

Do you mean when I say that if reality is continuous then it cannot be computed?

I didn't think that would even be controversial.

What the hell do you mean by "continuous" ?

Yes, which is what I said.

No, it's the opposite of what you said.

I don't think you understand the argument very well.

 

[Brian-M]

What the hell do you mean by "continuous" ?

He means that physical properties are without a smallest possible unit, or non-quantum, so that you'd need infinite digits to represent the values digitally with the same precision which exists in reality, so any digital representation would be inaccurate.

(Not that any simulation would need to be infinitely precise to be functionally indistinguishable from reality even if this were true.)

 

[Belz...]

So, nonsense, then.

 

[marplots]

He means that physical properties are without a smallest possible unit, or non-quantum, so that you'd need infinite digits to represent the values digitally with the same precision which exists in reality, so any digital representation would be inaccurate.

(Not that any simulation would need to be infinitely precise to be functionally indistinguishable from reality even if this were true.)

There's something to this, at least for inherently chaotic processes. Weather is the usual example, but any system sensitive to initial conditions would do. (As well as systems linked to those systems.) Fidelity is a real problem.

 

[Pup]

Have you ever felt nauseous? That.

Okay. Seems a not-very-useful definition for someone who wants to explore the question in depth, but...

I certainly have no reason to believe that the machine has felt nauseous as I feel nauseous.

Okay. But how does that relate to your definition? If we examine the brain of a human who claims to feel nauseous, that person's brain will exhibit a fixed pattern like the machine, at any given instant. Do you mean that one can only tell if a person, creature or machine is feeling nausea is to ask it or have it spontaneously express that it feels nausea?

You didn't answer my other question: how would you tell whether a locked in person, who was paralyzed and unable to communicate but was conscious, was feeling nausea, based on your definition above?

 

[Darat]

Have you ever felt nauseous? That.

...snip...

Have you ever felt a run?

 

[Myriad]

Have you ever felt that horrible sensation when your pages in RAM have gotten totally fragmented, tangling your pending process threads all to hell, and your stack is nearly full and the interrupts are still coming in fast and furious, and you know if you don't reboot soon you're just going to crash?

 

[Belz...]

Ugh. Story of my life, man.