Is Penrose wrong about Strong AI?

  • 12 Replies
  • 4729 Views

0 Members and 1 Guest are viewing this topic.

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« on: April 24, 2013, 05:52:04 pm »
Quote from: sciborg2
Admittedly I'm not well read on the subject, but from a Wiki-trip it seems that Penrose is on to something when he notes that a computer couldn't apprehend something like Godel's theorems. (I haven't read Shadows of the Mind yet)

thanks,

Sci

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #1 on: April 24, 2013, 05:52:12 pm »
Quote from: Jorge
I've always boiled it down to something simple: can the main physical processes responsible for thought be emulated by a Turing machine?

Because I happen to think that the main physical processes required for human-like intellect, memory and creativity are mostly limited to cellular (action potential) and molecular (protein expression) mechanisms, then the answer is yes. A Turing machine should feasibly be capable of running something like Gerald M. Edelman's model of consciousness, so I really don't see why not.

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #2 on: April 24, 2013, 05:52:22 pm »
Quote from: Callan S.
At about this point I guess I'd question if anyone can actually comprehend these Godel's theorems.

If there's some metric for measure if they do indeed understand, then seemingly a machine could understand.

If there isn't a metric/can't be, that'd make me speculate such theorems are an agreement game (a little like table top roleplay!)

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #3 on: April 24, 2013, 05:52:33 pm »
Quote from: sciborg2
Jorge: Will have get back to you, know nothing about Gerald M. Edelman.

Callan: Well, I don't yet have the insight necessary to comprehend Godel, but I have my fingers crossed....for advanced nootropics that is. ;-)

Basically, it goes back to Scott's idea that skin spies can't apprehend paradox. Mind you I've yet to get a hold of the book, but my understanding is that Penrose basically points out several mathematical problems that you can't solve using algorithms. As such, he points out that no current understanding of computation can produce human consciousness.

From there he takes a bold tactic in that he asserts consciousness has something to do with quantum mechanics. But barring that he's saying the same thing as Chalmers, that consciousness is something *more*....but this is would require another thread.

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #4 on: April 24, 2013, 05:52:37 pm »
Quote from: Callan S.
Hmm. To speculate, I'd propose instead a multiple processors approach to the mind - the ability to 'comprehend' paradox is actual two different processors going 'WTF?' 'WTF!?' to each other, unable to produce/hand shake on one single 'world' on the matter between them, to exist in/take as the one truth (is the lier a bad lier? Is he not a lier - two worlds, no way of marrying them into one. A mind split).

I've no idea how a skin spies brain is set out, but maybe it doesn't have multiple processors/the legion. It can never be 'in two minds' on a matter.

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #5 on: April 24, 2013, 05:52:42 pm »
Quote from: sciborg2
If I understand Penrose's argument neither multiple processors nor a quantum computer will suffice to bridge the gap.

Note that another mathematician said he found multiple problems with Penrose's reasoning but concluded Penrose was still, essentially, correct in his assertion. There's possibly some other issues with how the brain works that may not be replicable via AI algorithms, but this is something I've only seen a cursory summary of.

It gets to a point where it's hard to follow these guys.

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #6 on: April 24, 2013, 05:52:50 pm »
Quote from: Callan S.
I guess I'd go back and question whether people actually 'apprehend paradox', either? Ie, what does this apprehending actually do? What's the next step this apprehending leads to?

It doesn't seem to go anywhere - so how is paradox apprehended? It does not seem to be understood at all - just people saying 'paradox', folding their arms and nodding sagely?

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #7 on: April 24, 2013, 05:52:57 pm »
Quote from: sciborg2
Here's some more info on the controversy:

http://www.quantum-mind.co.uk/mathematics-logic-and-godel-c334.html

As with all things, take it with a grain of salt. (I use the same salt shaker for Dennet.)

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #8 on: April 24, 2013, 05:53:05 pm »
Quote from: sciborg2
An interview with Penrose. He actually elucidates things rather well here:

http://simplycharly.com/godel/roger_penrose_godel_interview.html

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #9 on: April 24, 2013, 05:53:12 pm »
Quote from: Ser Scot A Ellison
Sciborg,

I'm going to read your links later.  I think it comes done to whether you believe "conciousness" exists or not.  In my opinion a materialist would have to assert that "conciousness" is nothing special that it is mearly a byproduct of something being alive, the ability to percive the fact that you are alive.  Otherwise, if conciousness is something unique to humans (at least of Earthbound species) then it is something special that truely does set us apart.

That said if other species posess conciousness I don't know if that means materialists are right and we are limited to this meat we inhabit.  ;)

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #10 on: April 24, 2013, 05:53:19 pm »
Quote from: Ser Scot A Ellison
I'm reading the Penrose interview and had a thought.  What if conciousness is the ability to think about things that are "non-computable" as Turing defines the term?

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #11 on: April 24, 2013, 05:53:44 pm »
Quote from: sciborg2
Scot: As to consciousness relating to the non-computable, I think that is partially Penrose's idea, and while I have no idea if his proofs are correct or not I sort of see it in the same way I think you are proposing.

IF he's right, then consciousness is a combination of "free won't", deliberation, and mathematical thinking. Though for him this is the collapse of the wave function, which leads to the Mind being fundamental and thus neither deterministic nor random.

Of course, for most things, I think you can compute them - someone on the Bakker blog mentioned AIs and goal states, and I do agree for the majority of things we sort out what will make us happy and then attempt to achieve those goals.

As a friend of mine said - "Having to choose everything by deliberation would lead to people being stressed and unhappy." Many of the most "meaningful" things, like loving a spouse or wanting to be a musician, aren't really things that we choose. We "look inside ourselves", aka query our happiness measurement systems, and then act.

What's interesting is that the weight we apply to certain things seems to something we can edit - therein comes in the ability to motivate ourselves. Most of our will seems to relate more to pushing ourselves and enforcing self-discipline - both things can fall into the "free won't" paradigm.

What Came Before

  • *
  • Administrator
  • Emwama
  • *****
  • Posts: 0
    • View Profile
    • First Second Apocalypse
« Reply #12 on: April 24, 2013, 05:53:55 pm »
Quote from: sciborg2
Scot: As to consciousness relating to the non-computable, I think that is partially Penrose's idea, and while I have no idea if his proofs are correct or not I sort of see it in the same way I think you are proposing.

IF he's right, then consciousness is a combination of "free won't", deliberation, and mathematical thinking. Though for him this is the collapse of the wave function, which leads to the Mind being fundamental and thus neither deterministic nor random.

Of course, for most things, I think you can compute them - someone on the Bakker blog mentioned AIs and goal states, and I do agree for the majority of things we sort out what will make us happy and then attempt to achieve those goals.

As a friend of mine said - "Having to choose everything by deliberation would lead to people being stressed and unhappy." Many of the most "meaningful" things, like loving a spouse or wanting to be a musician, aren't really things that we choose. We "look inside ourselves", aka query our happiness measurement systems, and then act.

What's interesting is that the weight we apply to certain things seems to something we can edit - therein comes in the ability to motivate ourselves. Most of our will seems to relate more to pushing ourselves and enforcing self-discipline - both things can fall into the "free won't" paradigm.