Further reading matter.

[zotz]

Yesterday there was a Guardian article about the philosopher (and sculptor, and jazz pianist, and sailor, and cidermaker . . .) Dan Dennett, who it turns out is a man too talented to be allowed to live. I have recommended his work to some of you before - especially Consciousness Explained and Darwin's dangerous Idea.

Comments

[steer.livejournal.com]

Thanks for pointing that out. Dennett totally rocks. Have you read "The Minds I" that he edited with Hofstadter? That's an amazing book.

[ex-lark-asc.livejournal.com]

I've had Consciousness Explained recommended to me before after I had one of my rants about the nature of existence.. should get round to reading it really..

[ex-lark-asc.livejournal.com]

I'll wait till this $#!@ing cold's worn off first then..

[zotz]

Indeed. Currently on loan to karmicnull, I believe.

I thought the best bit was the Turing reprint, where he summarises Penrose's position on consciousness in one sentence and then dismisses it in the next. Written in the early fifties, of course.

[zotz]

It's very good, but isn't always light reading.

[ex-lark-asc.livejournal.com]

That was what she said. You're talking in unison with your girlfriend, you know..

[steer.livejournal.com]

Hmmm.... is that somewhere in "Computing machinery and intelligence"

http://www.abelard.org/turpap/turpap.htm#contrary_views_to_the_main_question

I'm afraid that Penrose's view is much closer to my own than Dennett or Hofstadter (though I think the latter two are better writers). Penrose was widely (and unfairly I thought) criticised for providing arguments in principle not proofs. Fundamentally I have much sympathy with his main thrust, that nothing we currently know about in physics can genuinely explain consciousness.

[zotz]

His argument is more that current physics forbids consciousness, therefore the only major hole in it - gravity - must therefore take a particular form, and consciousness must depend on gravity. The fact that this all follows from a reading of Gödel that pretty much everyone else thinks is hilariously wrong doesn't exactly promote the view to me, regardless of his excellence in other areas.

[zotz]

At least we don't forget which of us is which the way her computer does . . .

No, we dont. Really we don't.

[steer.livejournal.com]

His argument is more that current physics forbids consciousness

Yes, or better phrased, that nothing in current physics can explain consciousness. To me this is self-evident. I would have willingly accepted that as an axiom.

therefore the only major hole in it - gravity - must therefore take a particular form

I wish my copy of emperor wasn't in my office (and my copy of shadows is missing presumed lent)... that is certainly not what I took away from it. Certainly he wished to argue for a certain form of those physical laws - he's something of an expert there. But again, I recall this differently - that that was an optional part of his argument about minds... "Having done all that, I wish to convince you of that the following is one possible way forward..." I would say that his argument did not in the slightest rely on any particular form of theories on gravitation.... he just wished to champion one.

The fact that this all follows from a reading of Gödel that pretty much everyone else thinks is hilariously wrong

That was part of what I meant about the condemnation. I wouldn't say his reading of it was hilariously wrong but I would agree it is wrong. But his position was not that everything followed from that... his position remained tenable without that part of the argument and that is how I took his position. I did not believe his argument via Goedel was correct but having removed it totally the rest of his position was still intellectually sufficiently coherent for me.

[What is interesting, is that while everyone agrees that his interpretation of Godel is wrong there's a considerable amount of disagreement about WHY it is wrong -- which is in itself intriguing. If the experts in a field are all totally in agreement that someone is wrong in mutually contradictory ways then that at least tells you things aren't as simple as they might appear.]

[steer.livejournal.com]

You must enjoy this.

http://psyche.cs.monash.edu.au/v2/psyche-2-06-moravec.html

[zoo-music-girl.livejournal.com]

Yes, or better phrased, that nothing in current physics can explain consciousness

No, that implies that consciousness is something to be explained in terms of physics. It's not the same thing at all.

that is certainly not what I took away from it. Certainly he wished to argue for a certain form of those physical laws - he's something of an expert there.

Well, he needs a physical interaction that is only described by a non-computable description. We don't have any such thing, and the only obvious place for such a thing is in gravity. I'm not aware of any practical reason to suspect that any physical process would be like that. It's consequent on his reading of Gödel - Dennett asked him personally which way round the argument was suppose to work. He's not trying to argue the gravity line from his knowledge of physics.

I did not believe his argument via Goedel was correct but having removed it totally the rest of his position was still intellectually sufficiently coherent for me.

I read a TLS review of Emperor before I read the book itself. There was a summary of his argument and I counted points in it that I as a biologist thought were definitely or probably wrong, or entirely speculative. 27, as I recall. Without his argument from principle, it collapses like a stack of cards.

[ex-lark-asc.livejournal.com]

It could be worse. You could do what I do, and confuse the names of your two current love interests - but only when talking to the one who hates the other one, whose name you have of course just accidentally called him by.

Actually I haven't done that *yet*, but it's only a matter of time.

[ex-lark-asc.livejournal.com]

that implies that consciousness is something to be explained in terms of physics

Interesting - my layman's view is that physics can only (potentially) explain the neurological mechanisms which make consciousness possible in the brain, much as they can only explain the mechanics of a computer's hardware shooting electrons around, not the logic of the software that "brings it to life". I've found what little I've read about evolutionary psychology very interesting, seeing consciousness as the ultimate result of a massively and recursively self-monitoring system - put enough layers of feedback mechanisms into a system and eventually the system becomes aware that it must exist as an entity in its own right. Given the amount of time the software of the human brain has had in which to evolve, that doesn't seem implausible to me.

[zotz]

Bah. That was me again, of course.

[steer.livejournal.com]

No, that implies that consciousness is something to be explained in terms of physics. It's not the same thing at all.

To me the two are entirely equivalent. To say that "X is forbidden by current physics" (when it obviously occurs) is simply a rearrangment of the statement "current physics cannot explain X". I am not sure I have understood you here.

What does "X is forbidden by current physics" (for an event X which occurs) mean apart form "current physics cannot explain X"?

Well, he needs a physical interaction that is only described by a non-computable description.

Ah - I think you've misunderstood here - that comes from QM not gravitation. Remember, the waveform collapse part of QM is already non-computable [or according to best current theories]. The waveform evolution is computable but which way the event occurs is non computable -- a probabilistic process with no hidden variable.

Without his argument from principle, it collapses like a stack of cards.

I think we're just going to have to disagree on that one - I read his book, not believing the Goedel part of the argument, taking the biological part (all the stuff about microtubules etc) as unnecessary and speculative and still believed that the book was intellectually coherent. Those parts were not central to his argument as I saw it.

[zoo-music-girl.livejournal.com]

To say that "X is forbidden by current physics" (when it obviously occurs) is simply a rearrangment of the statement "current physics cannot explain X".

Physics can't explain, for instance, Pythagoras' Theorem, but it doesn't and can't forbid it. Neither does it explain the course of the English Civil War. I absolutely cannot see where the idea that consciousness has to have some special dispensation from the laws of physics to exist comes from. It strikes me as very literally nonsensical.

Remember, the waveform collapse part of QM is already non-computable [or according to best current theories]. The waveform evolution is computable but which way the event occurs is non computable -- a probabilistic process with no hidden variable.

Well, what I took from it is that these are indeed computable, because the probabilities of various outcomes can be computed very accurately. I can go back and have a look, but my understanding was that being non-deterministic isn't enough to count as non-computable.

I read his book, not believing the Goedel part of the argument, taking the biological part (all the stuff about microtubules etc) as unnecessary and speculative and still believed that the book was intellectually coherent.

It seems to me that you've just said that you don't accept either leg of his argument but still regard his conclusion as proven. What am I missing here?

[zotz]

Thats roughly the Functionalist position, yes - consciousness is a property of certain systems, and depends on the nature of the system rather than what material it's made of. We've only seen one such type of system, but in principle there could be others which become conscious if they're suitably organised and suitably complex.

[zotz]

Me again. Never mind, I'm going home tomorrow. It won't happen again for a while.

[zotz]

It's certainly wonderfully sarcastic.

[steer.livejournal.com]

Physics can't explain, for instance, Pythagoras' Theorem, but it doesn't and can't forbid it. Neither does it explain the course of the English Civil War.

I think these are both slightly specious examples and I'm not sure if you are tongue in cheek when you suggest them. Pythagoras Theorem is part of a mathematical framework which is the basis of physics. You could try to explain the Civil War from the point of view of physics but it would be a tangled viewpoint -- we could go into the whole "the particles making up Oliver cromwell were imparted a certain trajectory" but it is the wrong level of detail to look at the problem. [Doesn't Penrose actually go through this sort of example in his book?] Pythagoras is really sitting at a meta-level (it is part of the framework used to explain physics) -- the english civil war is, at a certain level, explicable by physics -- but that is an absurd level at which to try to understand it.

I absolutely cannot see where the idea that consciousness has to have some special dispensation from the laws of physics to exist comes from.

That's a rather unusual way around of looking at things... the point is not that consciousness needs a dispensation from the laws of physics but that most people argue that consciousness is completely explicable within the current laws of physics (cf Dennett "consciousness explained" and Hofstadter's arguments about the Carrenium in GEB). The laws of physics, presumably govern everything which happens in nature -- if the laws of physics as currently known are incapable of explaining consciousness then either we accept:

(a) Consciousness is inexplicable -- as mystics argue.
or
(b) We need new laws of physics -- as Penrose argues.

It's hard to see an alternative to this apart from accepting that current physics can explain consciousness.

Well, what I took from it is that these are indeed computable, because the probabilities of various outcomes can be computed very accurately

Certainly the probabilities are computable but, for most interpretations of QM then the waveform collapse is not -- as with much in QM this is, of course, contraversial. This would make it a non-computable system as non-computable is normally understood -- see http://mathworld.wolfram.com/ComputableFunction.html. [Perhaps my mathematicians understanding of computable is insufficient -- but computable is usually taken to be algorithmic or turing equivalent -- this explicitly rules out anything which is probabilstic.]

It seems to me that you've just said that you don't accept either leg of his argument but still regard his conclusion as proven. What am I missing here?

Because those weren't the only reasons for his argument - indeed he very explicitly said taht the biological micro-tubule part was speculative and unnecessary to his argument. Neuro-biologists got very up in arms about that and said it was all highly speculative -- which Penrose himself said. That was not a necessary part of anything he was arguing. The Goedel part was only one of the reasons he gave for believing that current physics could not explain consciousness. It was (if I recall correctly) a single chapter in the original book (and understandably one of the most criticised). The basic, and most compelling point, is that nothing in physics connects at all to qualia, feeling or self-awareness no matter how Dennett or Hofstadter may talk of emergent processes.

[ex-lark-asc.livejournal.com]

Good lord, you mean I'm in agreement with a school of philosophical thought?

You understand I now have to commit ritual suicide..

[steer.livejournal.com]

(grin) Don't worry.

Oh - I should stress that Penrose was talking about theories being non-computable in a more interesting way than purely being random. But nonetheless something which is purely random is certainly non-comptable.

[zotz]

Well, obviously please accept my advice not to, but if you feel like going ahead anyway can I take pictures?

[zotz]

Indeed. This is why I said that as far as I know there was nothing known similar to what he suggested. He said that we can do something which isn't algorithmically simulable, so it must depend on an interaction, at some level, which isn't algorithimically simulable either (ie non-computable). To simplify it horrendously. Otherwise we oculd just simutate in silico (and ab initio) every electron, proton and neutron in a human being together with their interactions, which is philosophically an entirely reasonable approach. He needs to find a reason why even that wouldn't work.