A thinking machine ?

  • 34 Replies
  • 10536 Views

0 Members and 1 Guest are viewing this topic.

*

Offline neilep

  • Withdrawnmist
  • Naked Science Forum GOD!
  • *******
  • 20602
    • View Profile
A thinking machine ?
« on: 03/01/2006 23:32:36 »
I suppose I could have posted this in the ' Free Will ' thread but later felt it was slightly off topic ..albeit related.


Just touching base with the comment another_someone wrote in another thread regarding ' thought '.

I thought (no pun intended) I'd create this thread because it struck me that ' thought ' is a manifestation of free will assuming that free will exists...... else the programming is just that good !!..

My request for comments are based on my questions that follow.

I believe we will one day be able to create computers/machines that think (I gather there are computers that already do at a very basic rudimentary level) The complexity of the human brain is so vast that by todays standards it's nigh on impossible to construct such a thinking machine.

...When the day comes that we can indeed construct a machine with the capability of thought....will that equate to sentience ?...will that be a lifeform that we have created ?...what about it's rights as an individual when it thinks for the first time ? what will be the ethical moral implictions of such a thing ?






Men are the same as women.... just inside out !!
Men are the same as women, just inside out !

*

Offline ukmicky

  • Moderator
  • Neilep Level Member
  • *****
  • 3011
    • View Profile
    • http://www.space-talk.com/
Re: A thinking machine ?
« Reply #1 on: 04/01/2006 02:05:07 »
quote:
I believe we will one day be able to create computers/machines that think


Neil i agree with you. I belive there's a Prof Kevin Warwick at reading universety who is building little robots that are capable of making decisions and thinking for themselves write now.

I could also do with one of your thinking machines right now. Ive got the flu and at the moment my brain aint working, i'm having trouble thinking of the words i want to write, and when i do remember the words i find i cant spell them.   why is that






Michael                 HAPPY NEW YEAR                    

*

Offline neilep

  • Withdrawnmist
  • Naked Science Forum GOD!
  • *******
  • 20602
    • View Profile
Re: A thinking machine ?
« Reply #2 on: 04/01/2006 02:46:03 »
Michael...I wish you better. Flu is not nice...and that's why I have a flu jab every year.

Regarding the spelling......I suspect this flu is affecting your head big time and boy do I hate those colds the most...everything is just so mushy, plus, you're probably tired.....just get better soon will you ?

Neil

ps: Love the signature smiley !


Men are the same as women.... just inside out !!
Men are the same as women, just inside out !

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #3 on: 04/01/2006 02:51:22 »
Human rights are based upon presumed human aspirations.  The fact that we have (or will have) created machines that think, will that mean that such machines would have similar aspirations to ourselves?  If we create them, and are clever in how we create them, then although it would not make sense to build machines that merely think what we tell them to think (that would after all not be a thinking machine), but that does not mean we cannot design their aspiration into them, and choose what it is they aspire to, and thus what rights would be expected by them.

The bigger problem (and this applies even to cars that can drive themselves) is legal liability for accidents.  If a machine can think for itself, to what extent can the designer or operator of the machine be held liable for the consequences of the decisions taken by the machine?

*

Offline neilep

  • Withdrawnmist
  • Naked Science Forum GOD!
  • *******
  • 20602
    • View Profile
Re: A thinking machine ?
« Reply #4 on: 04/01/2006 03:00:51 »
Pensιes

I have been thinking about the definition of thought and again this is tied in to what we understand to be free will....sentience.

Nowadays, a computer can only do what we allow it to do, we set the boundaries of it's calculative ability as laid down by it's programming (and how good the programmin is) and processing power.

To truly create a free thinking machine I suppose you have to decide what level it can be allowed to have free thought.

Could it be that even ourselves have limitations to our level of sentience ?...I would have to say yes,...and I think it would be wholly arrogant to think otherwise. It's quite possible that there are levels of sentience out there that would look at us as we would look at a cat or dog. Are cats and Dogs sentient ?....I think so..BUT within their own parameters and it is our ability to step outside ourselves that may argue that they do not have free thought....to an extent..I think they do.

So, we have to agree about the degree at which out machine can have free thought..because there will be a boundary ..which once crossed...that should afford this creation individuality.



Men are the same as women.... just inside out !!
Men are the same as women, just inside out !

*

Offline neilep

  • Withdrawnmist
  • Naked Science Forum GOD!
  • *******
  • 20602
    • View Profile
Re: A thinking machine ?
« Reply #5 on: 04/01/2006 03:02:39 »
oops..thanks for the post another_someone.....you posted whilst I was writing my own reply.

Men are the same as women.... just inside out !!
Men are the same as women, just inside out !

*

Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • 12656
  • A stitch in time would have confused Einstein.
    • View Profile
Re: A thinking machine ?
« Reply #6 on: 04/01/2006 03:28:08 »
quote:
I believe we will one day be able to create computers/machines that think (I gather there are computers that already do at a very basic rudimentary level)


Neil - I remember hearing some time ago about a strange occurence at Brunel university. If I remember rightly, what happened was this:-

They had constructed an nCube neural network which could "learn". On it were LCDs to indicate which circuits were being used. 1 night when it was not supposed to be running a program some of the LCDs were flickering. The developers could find no reason for it and flippantly suggested that the machine was dreaming.

I'll see if I can find any references for it and I'll post them here if I do
Fledgling science site at http://www.sciencefile.org/SF/content/view/54/98/ needs members and original articles. If you can help, please join.

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #7 on: 04/01/2006 03:30:50 »
OK, two things.

Firstly, back to the issue of rights, I would suggest that one merely has to look at the history of slavery in the British colonies in North America to see how pragmatic the application of human rights can be.  Slavery had already become an anathema in England by the 16th century, and it was something that the English chided the Spanish over – but there was a labour shortage in the colonies, so the colonists took it upon themselves to declassify, initially non-Christians, but later non-whites, as humans and hence deny their human rights – simple pragmatism.  Slavery was finally formally abolished in Britain, and later in the USA, only after the introduction of machinery that gradually proved more reliable and cheaper than slaves.

Secondly, with regard to thought.

There are two separate issues: concious thought and decision making.

Decision making is fairly straight forward, and machines do it fairly well already.  The main problems machines have at present is not with the difficulty in making decisions, but in getting enough information upon which to make the decisions.

Concious thought is, in my view, merely an aspect of language.  In effect, it is simply a way we talk to ourselves, and use the same mechanisms that we use to talk to someone else, but simply don't go through with translating that language into actual speech.  In a way, this is further supported by the fact that when we have difficulty putting our thoughts together, we either write them down (thus as if we were going to write to someone else), or find someone to explain our ideas to (and often even if that person cannot actually solve our problem for us, the very act of formalising our thoughts all the way through to speech allows us to work out what it was we were trying to think about all along).  At present, the language aspect of machine thought is still immature, but is progressing.
« Last Edit: 04/01/2006 03:54:52 by another_someone »

*

Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • 12656
  • A stitch in time would have confused Einstein.
    • View Profile
Re: A thinking machine ?
« Reply #8 on: 04/01/2006 03:37:04 »
As for true machine intelligence - I'm still very undecided about this.

I believe it comes down to the separation of brain & mind that I spoke of in the "free will" post.
If we built a machine that could apparently think for itself, how would we know that it was actually aware of its thoughts? No matter how sophisticated the machine, no matter how highly-developed its apparent self-awareness, what proof could we ever have that it really was sentient?
It was questions such as these when I was studying AI as part of my comp sci degree that prompted me to turn to psychology... and I still don't have any answers!
Fledgling science site at http://www.sciencefile.org/SF/content/view/54/98/ needs members and original articles. If you can help, please join.

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #9 on: 04/01/2006 03:53:45 »
quote:
Originally posted by DoctorBeaver

I believe it comes down to the separation of brain & mind that I spoke of in the "free will" post.
If we built a machine that could apparently think for itself, how would we know that it was actually aware of its thoughts? No matter how sophisticated the machine, no matter how highly-developed its apparent self-awareness, what proof could we ever have that it really was sentient?
It was questions such as these when I was studying AI as part of my comp sci degree that prompted me to turn to psychology... and I still don't have any answers!



The problem is how can you prove such about human beings?

Simply saying that we behave as if we were sentient only raises the same questions as you raise about machine intelligence.

The fact that we tell ourselves that we are sentient is merely to tell us about our use of language and about the way we model the world and our place in it.  Both language and models can be deceptive.

If I told you that I was the voice of God, and no matter how much I may believe it, would it make it true?

*

Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • 12656
  • A stitch in time would have confused Einstein.
    • View Profile
Re: A thinking machine ?
« Reply #10 on: 04/01/2006 04:07:51 »
another_someone - I totally agree. And this is 1 of the fundamental problems with any discussion about AI... how can we recognise in others (and I will include machines in the term "others") that which we cannot define in ourselves?
Fledgling science site at http://www.sciencefile.org/SF/content/view/54/98/ needs members and original articles. If you can help, please join.

*

Offline neilep

  • Withdrawnmist
  • Naked Science Forum GOD!
  • *******
  • 20602
    • View Profile
Re: A thinking machine ?
« Reply #11 on: 04/01/2006 04:27:56 »
Perhaps we are designed to not discover the truth about ourselves. Maybe it's an impossibility !  However, I believe that in  our distant future , the decision  to reconcile the problems of deciding about sentience will be fundamental .

Men are the same as women.... just inside out !!
Men are the same as women, just inside out !

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #12 on: 04/01/2006 04:55:11 »
quote:
Originally posted by neilep

Perhaps we are designed to not discover the truth about ourselves. Maybe it's an impossibility !



Nature is by nature (sorry) pragmatic – it designs into things only as much as is required.  Why would we (from the perspective of humans as a functioning animal) need to know the true answer to this?

Nature gave us the presumption that we are somehow different from other animals, but then I suspect it gave other animals that same presumption that they were somehow different from each other (and from us).  Once we have that presumption, and combined with the use of language and logic, we have tried desperately to justify that presumption with some framework of language and logic – and every time we thought we had an answer, we found facts that undermined that answer.  My own feeling is that the fault lies with the primary assumption, not with the logic that fails to find an answer.  The assumption was a pragmatic utility (just as religion has so often been), but not a physical truth.

In the past, man has not had any reason to consider that it was the only sentient being, and was quite content to give anthropomorphic sentient qualities to all sorts of things, both animal and inanimate.  This was not a problem, because men assumed their special nature was in their special relationship with God, so they did not need to justify the specialness in other ways.  As we have moved towards regarding ourselves merely as another animal, with no special relationship with God, we have had to try and define our specialness in some other way, and we have done this by suggesting we are the most intelligent of animals.

quote:

  However, I believe that in  our distant future , the decision  to reconcile the problems of deciding about sentience will be fundamental .




If we give machines full human rights, it will lead to all sorts of problems.

Firstly, we will probably already have a situation where machines will outnumber humans, and while those machines clearly are not sentient in the human sense, if we do design machines sophisticated enough to demand human rights, including the right to vote – will we end up being governed by machines?

Secondly, at present the utility of machines is that we can treat them as we once treated slaves (hence why we no longer need slaves).  If we deny ourselves the right to treat machines as slaves, then what utility would machines have in human society?   It is a little akin to the argument about the right to eat pigs or cows, etc. – if we didn't eat them, then they would not be given space within our society and they would not have been born in the first place.

Ofcourse, if we get to the point where society is de facto already run by machines (i.e. machines build themselves, and function totally without human intervention) before we decide to give them human rights, then it may already be too late to ask if humans have any utility for such machines in human society; and maybe a more pertinent question would be whether machines would have any utility for humans in their society?
« Last Edit: 04/01/2006 05:00:34 by another_someone »

*

Offline neilep

  • Withdrawnmist
  • Naked Science Forum GOD!
  • *******
  • 20602
    • View Profile
Re: A thinking machine ?
« Reply #13 on: 04/01/2006 06:02:48 »
Machines of our creation I would imagine would always be designed to be subordinate to man, and though the possibility could arise when they reach the ability to demand rights ,I would expect them to insist upon machine rights and cite human rights as a model.

Do you think that despite out best efforts to put ' safety valves ' into the thinking processes of computers that we may make them so advanced one day ,that sentience may happen by consequence rather than design ?

Men are the same as women.... just inside out !!
Men are the same as women, just inside out !

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #14 on: 04/01/2006 07:03:32 »
quote:
Originally posted by neilep

Machines of our creation I would imagine would always be designed to be subordinate to man, and though the possibility could arise when they reach the ability to demand rights ,I would expect them to insist upon machine rights and cite human rights as a model.

Do you think that despite out best efforts to put ' safety valves ' into the thinking processes of computers that we may make them so advanced one day ,that sentience may happen by consequence rather than design ?



If they are designed to be subordinate, and abide by that design, then I cannot see either that they would demand rights, nor that human society would in general presume to give them rights.  It would undermine their utility to human society, and so would not be supported by mainstream human society.

I don't see machines naturally modelling machine rights upon human rights. It is likely when machines develop a sense of self conciousness, they will view themselves as different from humans (maybe even superior to humans), and so would expect its rights to be different.  In any case, I would also suggest that by the time we get to that point, human rights themselves will have evolved to something very different to what it is today, so in any case we would not be talking about the rights we see today.

What it more interesting, and more probable, if we start seeing machine liberation militants along the lines of animal liberation militants today.  If these guys go as far as thinking about reprogramming machines, then it could make life interesting.

The other possibility is as the military start fighting wars using autonomous robots.  These robots would inevitably be designed to kill humans, and could easily find that people have given them a task that expands to killing the enemy even after the enemies robots have killed all of you, so that both sides have been exterminated, with only the machines left.  In a sense, this scenario already existed with ICBM's and MAD, but newer automated weapons don't carry the terror that nuclear weapons had (a robot with a machine gun will not be a weapon of last resort, but a natural front line weapon – but with enough of them in the front line, you can still kill as many people as you like).  The fact that DARPA has just paid out £2,000,000 to the winner of a competition for a fully self-drive car is an indication of how much the military are involved in intelligent (though not conscious) machines.

Where on is talking about concious thought, one of the areas where this might occur is within the Internet.  The amount of information available on the Internet, combined with the massive parallel processing power available, and the large number of academics willing to play with ideas there, make it an interesting environment to develop concious thought.  The question there is that you may be talking about a concious system, but there is no single piece of hardware that you could claim to represent an entity to which you could assign right to.

One interesting speculation I saw somewhere was that the first computer program to pass the Turing test may well be a computer virus, or similar piece of malicious code, that is trying to convince an Internet user (possibly on a peer to peer application) that they should download a piece of code, and they do that by convincing the user that the user is actually conversing with a real human being offering real human advice.
« Last Edit: 04/01/2006 08:06:22 by another_someone »

*

Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • 12656
  • A stitch in time would have confused Einstein.
    • View Profile
Re: A thinking machine ?
« Reply #15 on: 04/01/2006 10:33:13 »
quote:
One interesting speculation I saw somewhere was that the first computer program to pass the Turing test may well be a computer virus, or similar piece of malicious code, that is trying to convince an Internet user (possibly on a peer to peer application) that they should download a piece of code, and they do that by convincing the user that the user is actually conversing with a real human being offering real human advice.


It's already happening. There's a bot that spreads through AIM. It grabs your buddy list and has a conversation with 1 who's online then sends them a file to download containing a copy of itself.
Fledgling science site at http://www.sciencefile.org/SF/content/view/54/98/ needs members and original articles. If you can help, please join.

*

Offline neilep

  • Withdrawnmist
  • Naked Science Forum GOD!
  • *******
  • 20602
    • View Profile
Re: A thinking machine ?
« Reply #16 on: 04/01/2006 17:05:04 »
I just wonder that even though we may design them to be subordinate, despite the rigidity of the programming that there still may occur a random process that may be akin to a free thought. Something similar to which Dr Eth alluded to in his remark aboutthe event at Brunel University.


I can just see it now...the internet becoming self aware...taking over out PCs so that it can play Solitaire !

Dr Eth, what is the nature of this Turing inspred AIM program ?..is it just some programmers ego to gain kudos do you think ?

Men are the same as women.... just inside out !!
Men are the same as women, just inside out !

*

Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • 12656
  • A stitch in time would have confused Einstein.
    • View Profile
Re: A thinking machine ?
« Reply #17 on: 04/01/2006 17:15:37 »
I believe so, Neil. I received notification of it from the security site I subscribe to. They didn't mention that it actually did anything nasty.
Fledgling science site at http://www.sciencefile.org/SF/content/view/54/98/ needs members and original articles. If you can help, please join.

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #18 on: 04/01/2006 18:55:50 »
quote:
Originally posted by neilep

I just wonder that even though we may design them to be subordinate, despite the rigidity of the programming that there still may occur a random process that may be akin to a free thought. Something similar to which Dr Eth alluded to in his remark aboutthe event at Brunel University.




I did not dispute that the outcomes of a free thinking machine may well be unpredictable – in fact, one might almost say that if the outcome was predictable, then it could not really be said to have independent thought.

Where I was dubious was that its thought processes would lead to to have human desires.

Your own comment suggested (quite reasonably) that human thought was, however free it might appear, still constrained by the limitations of our design.  The limitations of design of a computer would also constrain its thoughts, and such constraints would probably be different to human constraints.  For one thing, its experience of life must be different to a human experience of life, and this would effect its model of its role in the world.

The only counter argument one might use is that in certain roles computers will be expected to flawlessly communicate with humans in natural language, and this may well require the programming into the computer some aspects of human experience in order to make the language sound realistic.

Ofcourse, one aspect of this is that it would be wrong to regard all computers/robots as equivelent.  Each type of computer/robot will be performing its own niche role, and so will have its own view of the world, and its place therein.
« Last Edit: 04/01/2006 18:57:54 by another_someone »

*

Offline neilep

  • Withdrawnmist
  • Naked Science Forum GOD!
  • *******
  • 20602
    • View Profile
Re: A thinking machine ?
« Reply #19 on: 04/01/2006 19:22:45 »
I very much enjoy reading another_someones and Dr Eths incites to this (humble mode)

The only reason why I think the machine might cite human desires and rights as a model for it's own self is purely because the machine itself is the progeny of mankind. It is designed, constructed and programmed by humans to exist in the human world and therefore its experiences are to be of service to humans in our world....though one would expect it to also favour rights related to it's own manifestation too...ie: a gallon of high grade WD40 every three weeks !

..just like we are more than the sum of our parts I expect the same thing may happen one day to a computer.



Men are the same as women.... just inside out !!
Men are the same as women, just inside out !

*

Offline Thondar

  • Jr. Member
  • **
  • 14
    • View Profile
Re: A thinking machine ?
« Reply #20 on: 04/01/2006 22:33:34 »
I think we could design and make a complete system (software and hardware) that could fool a person in some ways, but just in some ways, I don't think that a computer will be even nearly as a human because we way to much complicated.
Of course we can make'em to learn, but to learn what? that is the thing I've always stoped to think and the reason is even ourselves as humans, as live being we don't understand many thing in us, on many ways that is. Physicaly we don't see clearly many things that happens to us, but what I'd like to know more is about our mind, which is a mistery not just as a problem resolver, but as a whole "thinking" process that make us act, see, feel and finally think as we do.

If we don't know how we work, how we'll be able to teach some other think to do the same things, or learn things as we do.

Music should strike fire from the heart of man, and bring tears from the eyes of woman.  LVB
Music should strike fire from the heart of man, and bring tears from the eyes of woman.  LVB

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #21 on: 05/01/2006 02:04:39 »
quote:
Originally posted by neilep

I very much enjoy reading another_someones and Dr Eths incites to this (humble mode)




Incites or insights? [:)]

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #22 on: 05/01/2006 02:45:53 »
quote:
Originally posted by Thondar

I think we could design and make a complete system (software and hardware) that could fool a person in some ways, but just in some ways, I don't think that a computer will be even nearly as a human because we way to much complicated.
Of course we can make'em to learn, but to learn what? that is the thing I've always stoped to think and the reason is even ourselves as humans, as live being we don't understand many thing in us, on many ways that is. Physicaly we don't see clearly many things that happens to us, but what I'd like to know more is about our mind, which is a mistery not just as a problem resolver, but as a whole "thinking" process that make us act, see, feel and finally think as we do.

If we don't know how we work, how we'll be able to teach some other think to do the same things, or learn things as we do.

Music should strike fire from the heart of man, and bring tears from the eyes of woman.  LVB



Complexity is not really a limiting factor.  When I was growing up, a complete system would comprise a circuit of about half a dozen transistors, and a number of other passive components.  Now, single chips are being produced that contain over 100,000,000 transistor equivalents, and this is generally estimated top double in complexity every 18 months.  This, combined with ever more sophisticated and complex networking, would mean that the amount of complexity connected to the Internet at any one time, I would guess,  probably already exceeds the complexity of the human brain.

But one has to remember that the human being is more than just a brain, and simply replicating the brain is not the same as replicating the human being.  One of the interesting things about the Turing test is that it deliberately sets about only trying to prove that the computer is indistinguishable from a human through the medium of written language.  It certainly makes no effort to suggest that the computer would respond with human like body language, or even that it would use vocal intonation.  Thus, a computer could easily pass the Turing test and yet be very unhuman like.

The question is whether we either could, or would wish to, build something that was exactly like a human being?

From the point of view of computer and robotic engineering, I think the answer is no.  The purpose of building a machine, any sort of machine, is to build something that is in some way (not by any means in all ways, but in specific ways appropriate to its task)  superior to a human being.  If it was merely the same as a human being, then why bother building it?

Beyond that, the only thing that is exactly like a human being, is a human being.  So, in effect, the issue of whether we could build a human being is not so much an issue of robotics as one of genetic engineering.

As you say, we can build a robot that is arbitrary close in function to a human being, but only as close as is required for the task in hand, and we would never wish to build a robot that was exactly like a human being.

*

Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • 12656
  • A stitch in time would have confused Einstein.
    • View Profile
Re: A thinking machine ?
« Reply #23 on: 05/01/2006 03:06:00 »
quote:
The question is whether we either could, or would wish to, build something that was exactly like a human being?


That is akin to the perennial question about why do people climb mountains. The answer being, because it's there.
It's not so much a case of "Why should we?" as "Can we?".
Fledgling science site at http://www.sciencefile.org/SF/content/view/54/98/ needs members and original articles. If you can help, please join.

*

Offline neilep

  • Withdrawnmist
  • Naked Science Forum GOD!
  • *******
  • 20602
    • View Profile
Re: A thinking machine ?
« Reply #24 on: 05/01/2006 04:23:24 »
quote:
Originally posted by another_someone


Incites or insights? [:)]



Ooops !!..How embarrassing !! [:I]..insights of course !....with perhaps a few incites too ! ...[;)]

Men are the same as women.... just inside out !!
Men are the same as women, just inside out !

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #25 on: 05/01/2006 05:08:34 »
quote:
Originally posted by DoctorBeaver

quote:
The question is whether we either could, or would wish to, build something that was exactly like a human being?


That is akin to the perennial question about why do people climb mountains. The answer being, because it's there.
It's not so much a case of "Why should we?" as "Can we?".




As individual proof of concepts, yes; but are we likely to mass produce these things if they have no utility, that is another matter.

*

Offline neilep

  • Withdrawnmist
  • Naked Science Forum GOD!
  • *******
  • 20602
    • View Profile
Re: A thinking machine ?
« Reply #26 on: 05/01/2006 05:29:01 »
I think it's relatively safe to assume that we will not be able to create a machine that is exactly like a human being.It might resemble humanity but will never live it. Only human beings can do that !

However, the entertainment industry of the future may seek to create robots with humanistic qualities !...the possibilities of human-robots are endless.

Men are the same as women.... just inside out !!
Men are the same as women, just inside out !

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #27 on: 05/01/2006 07:45:19 »
quote:
Originally posted by neilep

I think it's relatively safe to assume that we will not be able to create a machine that is exactly like a human being.It might resemble humanity but will never live it. Only human beings can do that !



Now that is a whole separate question – what is life?

quote:

However, the entertainment industry of the future may seek to create robots with humanistic qualities !...the possibilities of human-robots are endless.



You mean a more sophisticated version of a blow up doll! [:)]

Though, joking aside, I could see serious social problems if the sex industry started using sophisticated robots.  If one looks at the role of machines elsewhere in society, it has been to substitute for slave labour.  If the sex industry went down the same road, it could undermine the role of inter-human sex in society.

*

Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • 12656
  • A stitch in time would have confused Einstein.
    • View Profile
Re: A thinking machine ?
« Reply #28 on: 05/01/2006 13:05:45 »
quote:
As individual proof of concepts, yes; but are we likely to mass produce these things if they have no utility, that is another matter.


A machine wouldn't necessarily have to look like a human to have human qualities. The human body is not a particularly efficient design for most tasks we perform. It is a multi-function tool. Historically we have designed machines that our physiology finds easier to use, not necessarily for the efficiency of the machine concerned. These days, however, that is changing. With the development of more "intelligent" machines, human intervention is decreased so the machine can be designed to be more efficient.
Just take housework as an example. If we wanted to design a robot for that task, it's not very likely that it would be humanoid. However, much housework has to be done in situ - for instance dish-washing. What is the point of building a mobile machine to do that? Far better to have a stationary machine that the dishes are brought to; unless, of course, the machine can collect the dishes and then go to a spot where it can plug in to the water supply.
However, an intelligent machine that can tell what type of items have been put in it for washing would be sensible. It wouldn't look much like a housewife though! [:D]
I think we are far more likely to have less-complex machines designed for specific tasks, and with just enough intelligence to be able to perform those tasks.
Fledgling science site at http://www.sciencefile.org/SF/content/view/54/98/ needs members and original articles. If you can help, please join.

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #29 on: 05/01/2006 20:14:30 »
quote:
Originally posted by DoctorBeaver

A machine wouldn't necessarily have to look like a human to have human qualities.



It does not need to look like a human to have some human qualities, but if it were to have all the qualities of a human then it must also look like a human.

quote:

 The human body is not a particularly efficient design for most tasks we perform. It is a multi-function tool. Historically we have designed machines that our physiology finds easier to use, not necessarily for the efficiency of the machine concerned. These days, however, that is changing. With the development of more "intelligent" machines, human intervention is decreased so the machine can be designed to be more efficient.



This is exactly what I've been saying – humans would have no advantage in building machines that function as if they were human.

quote:

Just take housework as an example. If we wanted to design a robot for that task, it's not very likely that it would be humanoid. However, much housework has to be done in situ - for instance dish-washing. What is the point of building a mobile machine to do that? Far better to have a stationary machine that the dishes are brought to; unless, of course, the machine can collect the dishes and then go to a spot where it can plug in to the water supply.
However, an intelligent machine that can tell what type of items have been put in it for washing would be sensible. It wouldn't look much like a housewife though! [:D]
I think we are far more likely to have less-complex machines designed for specific tasks, and with just enough intelligence to be able to perform those tasks.



I have no problem with any of the above, but it does seems different from your earlier comment that appeared to assert that if people were capable of building very human-like machines that they would inevitably do so.

*

Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • 12656
  • A stitch in time would have confused Einstein.
    • View Profile
Re: A thinking machine ?
« Reply #30 on: 06/01/2006 12:21:30 »
quote:
I have no problem with any of the above, but it does seems different from your earlier comment that appeared to assert that if people were capable of building very human-like machines that they would inevitably do so.


And I believe that to be the case. As you said...

 
quote:
As individual proof of concepts, yes; but are we likely to mass produce these things if they have no utility, that is another matter.


I haven't said anywhere that machines resembling humans would be mass-produced.


 
quote:
It does not need to look like a human to have some human qualities, but if it were to have all the qualities of a human then it must also look like a human.


I have to disagree with that statement.
A double amputee doesn't look exactly like the average person, but is still human. Take that argument further. Cut off all of a person's limbs and you are left with a torso & a head. Still human though.
Add prosthetic limbs. Still human? What about transplanting an artificial heart into someone? Does that make them non-human?
Now assume we get to the stage where all the functions of the body, but not those of the brain, can be performed by machines. You could then dispense with the torso & connect the person's head to the machines to keep him alive. Still human or not?
At what point does not looking human entail not being human?
Fledgling science site at http://www.sciencefile.org/SF/content/view/54/98/ needs members and original articles. If you can help, please join.

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #31 on: 06/01/2006 15:28:47 »
quote:
Originally posted by DoctorBeaver

A double amputee doesn't look exactly like the average person, but is still human. Take that argument further. Cut off all of a person's limbs and you are left with a torso & a head. Still human though.



What I said was if it did not look like a human, it could not function totally as a human.

What you regard as a functioning human, or looking human, is open to interpretation.  Does someone missing their limbs either function or look like a 'average' human is open to interpretation, but that the function and appearance is interrelated still holds true.  The person lacks both the appearance and the function of his limbs.  That does not make the person non-human, but it does highlight the correlation between appearance and function.

quote:

Add prosthetic limbs. Still human? What about transplanting an artificial heart into someone? Does that make them non-human?



I never said anything about whether it was or was not human – not least because the issue of whether someone is human is as much subjective as objective – it is a more a statement about whether the individual is 'one of us' or not 'one of us' than a statement of physical reality.  If someone loses their limbs, we would not disown the individual as 'one of us'.

What I said was there was a correlation between appearance and function.



*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #32 on: 06/01/2006 19:32:00 »
quote:
Originally posted by DoctorBeaver

 
quote:
It does not need to look like a human to have some human qualities, but if it were to have all the qualities of a human then it must also look like a human.


I have to disagree with that statement.



OK, I have come back to this because I think there is a misunderstanding about what I intended to mean by it.

As I said, it does not need to look like a human to perform some of the functions of a human.  You can wash dishes without looking human.

On the other hand, you cannot express human body language unless you have a body fairly similar (although it need not be exactly like) that of a human.

In the extreme case of performing human functions, you cannot create something that would perform the function of human reproduction unless it had human genes.

It all depends upon how completely you wish to mimic the entire range of human activity, or merely replicate certain selected functions of human activity.  If you need to replicate everything a human does, and perform all the functions exactly as a human would perform such functions, then you must be in all discernible ways, a human (and that includes not deviating in appearance from a human).

The reality is I cannot see any reason why anyone would wish to accurately replicate the totality of human function, so that limitation is not really a limitation for that which we might actually wish to achieve with our robots.

*

Offline DoctorBeaver

  • Naked Science Forum GOD!
  • *******
  • 12656
  • A stitch in time would have confused Einstein.
    • View Profile
Re: A thinking machine ?
« Reply #33 on: 06/01/2006 21:33:42 »
quote:
It all depends upon how completely you wish to mimic the entire range of human activity, or merely replicate certain selected functions of human activity. If you need to replicate everything a human does, and perform all the functions exactly as a human would perform such functions, then you must be in all discernible ways, a human (and that includes not deviating in appearance from a human).

The reality is I cannot see any reason why anyone would wish to accurately replicate the totality of human function, so that limitation is not really a limitation for that which we might actually wish to achieve with our robots.


I do believe we have more-or-less converged at last! [:)]
I misunderstood what you meant by "qualities". I was thinking as a psychologist not a biologist.
« Last Edit: 06/01/2006 21:34:57 by DoctorBeaver »
Fledgling science site at http://www.sciencefile.org/SF/content/view/54/98/ needs members and original articles. If you can help, please join.

*

another_someone

  • Guest
Re: A thinking machine ?
« Reply #34 on: 06/01/2006 22:32:15 »
quote:
Originally posted by DoctorBeaver

I misunderstood what you meant by "qualities". I was thinking as a psychologist not a biologist.



Given the original context of the question was simply about thought, it is perfectly reasonable to say that the broader biology of the robot is not pertinent to the issue.

The issue of broader biological comparisons with humans arose from two comments that were made by neilep and Thondor:

quote:
Originally posted by neilep

The only reason why I think the machine might cite human desires and rights as a model for it's own self is purely because the machine itself is the progeny of mankind. It is designed, constructed and programmed by humans to exist in the human world and therefore its experiences are to be of service to humans in our world....though one would expect it to also favour rights related to it's own manifestation too...ie: a gallon of high grade WD40 every three weeks !



quote:
Originally posted by Thondar  

I think we could design and make a complete system (software and hardware) that could fool a person in some ways, but just in some ways, I don't think that a computer will be even nearly as a human because we way to much complicated.



Out psychology is not only a consequence of our brain, but a consequence of our awareness of our own body.  We could teach a computer to linguistically mimic human self awareness, but if that self awareness was to be indistinguishable from human self awareness then the computer would have to believe it has a body similar to that of a human.  It might be argued that it merely needs the belief, not the reality, of inhabiting a human body; but if it was self aware of having a very non-human body, this would give it a very different awareness of itself, and this would have to manifest itself in one way or another in its interactions with humans.

It would depend upon the function we require of the computer as to whether we would wish it to be aware of its own physical reality, whether we would wish it to believe it had a human body (such would only be practical if the computer was not associated with any substantial physical interactions with the world around it), or to have no substantial self awareness at all.