Can a machine be conscious?

Please compose all posts in Emacs.

Moderators: phlip, Moderators General, Prelates

User avatar
Argency
Posts: 203
Joined: Wed May 19, 2010 12:43 am UTC
Location: Brisbane, Australia

Re: Can a machine be conscious?

Postby Argency » Tue Jun 29, 2010 2:01 am UTC

You've misread me. I wasn't just talking about survival against Darwinian natural selection. The reason the computer in your example's survival is ensured is because it is optimised for survival against artficial selection: it does exactly what humans want it to do and therefore they will preserve it and replicate it.

And no, you can't just refer to my definition as survival optimisation, because that wasn't my definition. You've cleverly managed to miss the most important bit, which was the computational bit. Also, I never said anything about optimisation. I said fitness enhancing computational ability. That doesn't mean it has to be optimally fitness enhancing. So all that stuff you said about survival optimisation is talking across my argument.

Conflating mechanism and result makes your definition weaker, not stronger.
Unless you're suggesting that there's some aspect of consciousness other than computation (in which case a purely computational machine like a computer could never be conscious) then in this case the mechanism is the result. There's no conflation there. Consciousness is nothing more than a very complex computational process. That's why I chose my definition - because this was a debate about whether or not a machine could be conscious. I'm not really interested in arguing on a broader field. But either way - I've already agreed to accept your definition and move on, as long as we keep the mechanism of consciousness in mind. Why are you arguing about it when it's not important anymore?

The problem with this is that it is not a solution because it leaves you in an incorrect state at the end. All groupings of sand are not "nominally heaps" because they do not satisfy the definition of a heap. There is an alternative formulation of the paradox which might make this more apparent:

Is 1 grain of sand a heap of sand? (No, by definition of a heap)
If we add one, are 2 grains of sand a heap? No.
If we add one, are 3 grains of sand a heap? No.
.... (ad infinitum)
Therefore no matter how many grains of sand we add, we will never have a heap. Therefore, heaps don't exist.

You're begging the question. By nominally declaring all groupings of sand to be heaps, we change the definition of a heap. Go look up "nominally" in the dictionary. So to argue that groups of sand aren't nominally heaps because they don't satisfy the definition of a heap is to say that my definition of a heap is wrong because it isn't the correct definition of a heap. You're assuming your conclusion in your premises.

Your "solution" is functionally equivalent to agreeing that "heaps don't exist". But of course heaps exist so it is a non-solution. I understand what you're saying. But you're basically sacrificing correctness for convenience.
That's not functionally equivalent to my solution and you know it. It's functionally opposite to my solution. With my solution we end up with heaps existing. With that conclusion, we end up with no heaps existing. And, again, by arguing that I'm sacrificing correctness, you presuppose that you're right and I'm wrong, which begs the question.

Finally, the analogy with water molecules isn't bogus. What's bogus is your choice of predicates. You should have chosen "is a heap" and "has a temperature". They're both binary contingencies and my argument is that both of them are true right down to collections of one item (in the case of heaps, one grain of sand; in the case of temperature, one molecule of water).

This argument we're having used to be a definitional debate, which means I couldn't have been wrong purely on the basis of my definition. My definition was the thing you're saying I was wrong about. If you reread my last post, you'll find I've accepted your definition: that is, I'm willing to concede that yours is more useful in this (conversational) theatre and that therefore I was wrong. We now seem to be arguing about whether I was wrong because I'm an idiot or whether I had some reasonable reason for using the wrong definition. That's not really an argument I'm interested in having.
Gonna be a blank slate, gonna wear a white cape.

User avatar
WarDaft
Posts: 1583
Joined: Thu Jul 30, 2009 3:16 pm UTC

Re: Can a machine be conscious?

Postby WarDaft » Tue Jun 29, 2010 9:37 am UTC

Yakk wrote:Why are you messing up your model with R0, R1, L0, L1? It seems to me you needlessly multiplied states. So either you are confused (and are talking about a needlessly complex TM), or I'm confused (and those states actually do something interesting).

Well, if we cut it down to just L0, L1, it's kind-of hard to argue that the machine is computing anything at all. If there's a smaller instruction set that will necessarily contain the computation of every Turing machine, that works too, but I would have had to look that up, as opposed to just typing 6 more characters to ensure it.

Xanthir wrote:Actually computing one of those numbers would indeed be equivalent to simulating the universe. Is it necessary to even use the real numbers, though? If the universe simulations eventually loop, we can cut them off there. Then we have a finite-depth tree of turing-machine histories, which map to the integers, not the reals.
A bounded size volume eventually loops, and people can unquestionably live just fine in a bounded volume. The universe we live in however, expands, and does not necessarily loop by the proof I offered.
All Shadow priest spells that deal Fire damage now appear green.
Big freaky cereal boxes of death.

0xBADFEED
Posts: 687
Joined: Mon May 05, 2008 2:14 am UTC

Re: Can a machine be conscious?

Postby 0xBADFEED » Tue Jun 29, 2010 2:45 pm UTC

Argency wrote:You've misread me. I wasn't just talking about survival against Darwinian natural selection.

I wasn't suggesting that you were. I was talking about survival against an arbitrary selection process. In this case "survival" means beating the "selection process". In the case of living organisms that might be natural selection, but in other cases it might be completely different.
The reason the computer in your example's survival is ensured is because it is optimised for survival against artficial selection: it does exactly what humans want it to do and therefore they will preserve it and replicate it.

The point of the computer example was to try and remove the selection process entirely. I'm still wondering what you would say about such a system in the absence of any selection process.
And no, you can't just refer to my definition as survival optimisation, because that wasn't my definition. You've cleverly managed to miss the most important bit, which was the computational bit. Also, I never said anything about optimisation. I said fitness enhancing computational ability. That doesn't mean it has to be optimally fitness enhancing. So all that stuff you said about survival optimisation is talking across my argument.

An optimization process need not necessarily converge on the optimal solution, only move from less to more desirable configurations. I'm not trying to talk across your point. I'm trying to frame it in a way that I can really understand what you're talking about. Because the phrase, "computation enhancing fitness for selection" doesn't mean a whole lot on its own. It's so vague that without further clarification it could mean any number of things:
  • survival optimization (against an arbitrary selection process)
  • doing things to make itself better (for any arbitrary context of "better")
  • running a set program that performs well against a particular selection process
  • or any other myriad interpretations I can come up with that conceivably fit the meaning of the phrase
Can you just talk a bit more about why you think this idea is important and why it is the central property of consciousness so that I might better understand where you're coming from?
Unless you're suggesting that there's some aspect of consciousness other than computation (in which case a purely computational machine like a computer could never be conscious) then in this case the mechanism is the result. ....Why are you arguing about it when it's not important anymore?
I had no problem with computation bit it was the "aimed at fitness selection" bit that I thought loaded the definition. But you're right it's not important, I was just trying to explore some of the limitations I perceived from the phrasing.

You're begging the question. By nominally declaring all groupings of sand to be heaps, we change the definition of a heap. Go look up "nominally" in the dictionary. So to argue that groups of sand aren't nominally heaps because they don't satisfy the definition of a heap is to say that my definition of a heap is wrong because it isn't the correct definition of a heap. You're assuming your conclusion in your premises.

I am well aware of what "nominally" means and I can do without the snarkiness. I'm saying that the colloquial definition of a heap is a premise of the paradox and therefore invariant.
That's not functionally equivalent to my solution and you know it. It's functionally opposite to my solution. With my solution we end up with heaps existing. With that conclusion, we end up with no heaps existing. And, again, by arguing that I'm sacrificing correctness, you presuppose that you're right and I'm wrong, which begs the question.

I'm not making this stuff up. It's a well-known conundrum (for example) of the sorites paradox that if you accept one conclusion (all numbers of grains making a heap) you're also forced to accept the opposite version (heaps not existing). The chains of logic leading to the two conclusions are the exact same. If you accept one chain as a valid line of reasoning then you implicitly accept the other. I'm not presupposing I'm right. I'm saying that accepting the conclusion doesn't get you anywhere. It's convenient because it allows you to accept the line of logic as sound, but it doesn't solve the larger problems of inconsistency, with heaps both being any number of grains and not existing.
This talk about the heap is a complete aside anyway and I'm happy to just accept that we disagree on its interpretation.
This argument we're having used to be a definitional debate, which means I couldn't have been wrong purely on the basis of my definition. My definition was the thing you're saying I was wrong about. If you reread my last post, you'll find I've accepted your definition: that is, I'm willing to concede that yours is more useful in this (conversational) theatre and that therefore I was wrong. We now seem to be arguing about whether I was wrong because I'm an idiot or whether I had some reasonable reason for using the wrong definition. That's not really an argument I'm interested in having.

I was never saying your definition was wrong. I was only trying to point out implications of your specific definition that I found troubling, limiting, or confusing. Your definition is no more wrong than mine or anyone else's. There may be questions of utility but not correctness. You're clearly not an idiot, and I would never suggest or imply that you are one.


Return to “Religious Wars”

Who is online

Users browsing this forum: No registered users and 6 guests