Jump to content

Welcome to FutureTimeline.forum
Register now to gain access to all of our features. Once registered and logged in, you will be able to create topics, post replies to existing threads, give reputation to your fellow members, get your own private messenger, post status updates, manage your profile and so much more. If you already have an account, login here - otherwise create an account for free today!
Photo

Could a robot be conscious?

robotics artificial intelligence turing consciousness

  • Please log in to reply
46 replies to this topic

#1
GNR Rvolution

GNR Rvolution

    Member

  • Members
  • PipPipPipPipPipPip
  • 529 posts
  • LocationLondon
http://www.bbc.co.uk...gazine-14886421

Although this is from a news site it's not really news, more an article on consciousness and whether we could ever pass this on to an artificial being?
All right, brain. You don't like me and I don't like you, but let's just do this and I can get back to killing you with beer.

#2
jjf3

jjf3

    Not a Member of the Tea Party! Just a Concerned Conservative

  • Members
  • PipPipPipPipPipPipPipPip
  • 2,046 posts
  • LocationHolmdel NJ/Tampa Fl
Some people say that the internet has a life of its own. I tend to agree more with that one. The internet will fuel the Robots
"Did you really expect some utopian fantasy to rise from the ashes?" Thomas Zarek-- Battlestar Galactica.

#3
Prolite

Prolite

    Member

  • Members
  • PipPipPipPipPipPip
  • 609 posts
Eventually computers will be conscious, but I think we're a long way off from that. Maybe by 2150 they will be.
I'm a business man, that's all you need to know about me.

#4
GNR Rvolution

GNR Rvolution

    Member

  • Members
  • PipPipPipPipPipPip
  • 529 posts
  • LocationLondon
My thoughts (as such) is that nobody has been able to pin down what consciousness is as yet. Could it be that our consciousness is an illusion, that it is simply our brain fooling us into believing that we are something more than just a collection of matter? If this is the case then I believe that it will be actually rather easy for us to create a conscious robot, if we can fool ourselves then we could fool something else. However, if it is something more than simple mechanics, as postulated by theories such as Quantum Mind, then it could be rather much harder...
All right, brain. You don't like me and I don't like you, but let's just do this and I can get back to killing you with beer.

#5
Unrequited Lust

Unrequited Lust

    He Who Would Swallow God

  • Members
  • PipPipPipPipPipPip
  • 559 posts

Eventually computers will be conscious, but I think we're a long way off from that. Maybe by 2150 they will be.

The brain is simply a collection of matter and consciousness is achieved by interactions between neurons. I don't know why people think that these principles can't be understood and applied to computers.

People thought heavier-than-air flight was impossible even though birds do it. People still think nanobots are impossible even though every single cell in our body makes use of them. What is it about life that people think can't be replicated with technology? Life isn't mysterious. It doesn't do anything outside the laws of physics. Hell, it took billions of years of natural selection to make the semi-okay designs we see today.

There is nothing at all transcendental about the mechanisms of life. Quite the opposite, their designs are downright mediocre in comparison. We can easily build aircraft that goes way faster than the speed of sound. No organism comes even close to this.

Consciousness had no intended design. It's a series of genetic accidents. Are you (not you personally) telling me we can't replicate, even perfect, it? With that being said, I think artificial intelligence will be conscious within 2-3 decades. Hell, the Blue Brain Project plans to be completed by 2020.

#6
Prolite

Prolite

    Member

  • Members
  • PipPipPipPipPipPip
  • 609 posts


Eventually computers will be conscious, but I think we're a long way off from that. Maybe by 2150 they will be.

The brain is simply a collection of matter and consciousness is achieved by interactions between neurons. I don't know why people think that these principles can't be understood and applied to computers.

People thought heavier-than-air flight was impossible even though birds do it. People still think nanobots are impossible even though every single cell in our body makes use of them. What is it about life that people think can't be replicated with technology? Life isn't mysterious. It doesn't do anything outside the laws of physics. Hell, it took billions of years of natural selection to make the semi-okay designs we see today.

There is nothing at all transcendental about the mechanisms of life. Quite the opposite, their designs are downright mediocre in comparison. We can easily build aircraft that goes way faster than the speed of sound. No organism comes even close to this.

Consciousness had no intended design. It's a series of genetic accidents. Are you (not you personally) telling me we can't replicate, even perfect, it? With that being said, I think artificial intelligence will be conscious within 2-3 decades. Hell, the Blue Brain Project plans to be completed by 2020.


Why did you make your entire rant out to assume that I didn't believe that computers could become conscious? ..which is by the way, factually incorrect based on what I said. LoL you even quoted what I said and STILL made the supposition that I didn't support computer consciousness. You should seriously think about becoming a politician. ::cough:: smells like tea ::cough:: ;)
I'm a business man, that's all you need to know about me.

#7
GNR Rvolution

GNR Rvolution

    Member

  • Members
  • PipPipPipPipPipPip
  • 529 posts
  • LocationLondon

The brain is simply a collection of matter and consciousness is achieved by interactions between neurons. I don't know why people think that these principles can't be understood and applied to computers.


You may be right, but this isn't proven as yet, and it may be that the brain is more complicated than we think. Besides, what is defined as being conscious? The ability to simply perceive, to have self-awareness? Or is it the ability to create, to invent, to find the beautiful meaning in the little things? Many tests involving consciousness involve language, which should be relatively simple to learn, but parrots can mimic language?

As mentioned in other threads, I believe that computers can and will become self-aware, but will they truly be conscious, or are they simply following a complex series of commands? Are genetic algorithms the first step towards a computer that is creative?
All right, brain. You don't like me and I don't like you, but let's just do this and I can get back to killing you with beer.

#8
Unrequited Lust

Unrequited Lust

    He Who Would Swallow God

  • Members
  • PipPipPipPipPipPip
  • 559 posts



Eventually computers will be conscious, but I think we're a long way off from that. Maybe by 2150 they will be.

The brain is simply a collection of matter and consciousness is achieved by interactions between neurons. I don't know why people think that these principles can't be understood and applied to computers.

People thought heavier-than-air flight was impossible even though birds do it. People still think nanobots are impossible even though every single cell in our body makes use of them. What is it about life that people think can't be replicated with technology? Life isn't mysterious. It doesn't do anything outside the laws of physics. Hell, it took billions of years of natural selection to make the semi-okay designs we see today.

There is nothing at all transcendental about the mechanisms of life. Quite the opposite, their designs are downright mediocre in comparison. We can easily build aircraft that goes way faster than the speed of sound. No organism comes even close to this.

Consciousness had no intended design. It's a series of genetic accidents. Are you (not you personally) telling me we can't replicate, even perfect, it? With that being said, I think artificial intelligence will be conscious within 2-3 decades. Hell, the Blue Brain Project plans to be completed by 2020.


Why did you make your entire rant out to assume that I didn't believe that computers could become conscious? ..which is by the way, factually incorrect based on what I said. LoL you even quoted what I said and STILL made the supposition that I didn't support computer consciousness. You should seriously think about becoming a politician. ::cough:: smells like tea ::cough:: ;)

I know, I didn't quote correctly. Most of my rant was directed at the idea that computers cannot be conscious, not you. I quoted you because I think 140 years from now is just way too long. Can you imagine 140 years ago? that was 1870. 18 freaking 70. When they knew nothing compared to what we know today. Now take in the fact that information technology improves exponentially (yes, it's a fact). Even if we compare the progress from 1870-2011 linearly, you really think it'll take until 2050 to get conscious computers? If we're measuring it exponentially, then the idea that it'll take that long becomes ludicrous.

#9
Flatfingers

Flatfingers

    Member

  • Members
  • PipPipPip
  • 56 posts
"The illusion of consciousness" is pretty much how Doug Hofstadter thinks of it, with the proviso that any sufficiently advanced simulation of consciousness is indistinguishable from the real thing. So he thinks there's no technical/physics/metaphysics impeding the generation of non-human sentience. The only tricky question is, what's the best way to get there? Most people seem to want to think of it in GEB terms -- not "Godel, Escher, Bach," but Giant Electronic Brain. That is, as a brain-in-a-box, created by human design from components deliberately connected and turned on at the press of a button. I doubt that. I suspect that when (not if, but when) it happens, it will be the product of a system that grew itself, that started off as dumb as a bag of hammers -- but with potential -- and slowly got smarter... just like individual humans grow from infancy to adulthood. And probably with similar pains of adolescence, too. I also suspect that living as a disembodied collection of software routines inside a PC somewhere won't be sufficient. Sentience, I have a feeling, requires a body; more specifically, it needs senses for perceiving the richness of the textures of reality, and it needs limbs and a voice for interacting with that reality. And then it needs goals and problems and challenges on which to exercise its perceptive and interactive systems. Given all these things and enough parallel-but-optimized processing power, I think machine consciousness will eventually emerge. Not "one day," but slowly over years, so imperceptibly that we may not even realize that it's happened until someone says, "Hey, I just had a conversation with a bot and didn't even know it!"

#10
Prolite

Prolite

    Member

  • Members
  • PipPipPipPipPipPip
  • 609 posts

"The illusion of consciousness" is pretty much how Doug Hofstadter thinks of it, with the proviso that any sufficiently advanced simulation of consciousness is indistinguishable from the real thing. So he thinks there's no technical/physics/metaphysics impeding the generation of non-human sentience.

The only tricky question is, what's the best way to get there?

Most people seem to want to think of it in GEB terms -- not "Godel, Escher, Bach," but Giant Electronic Brain. That is, as a brain-in-a-box, created by human design from components deliberately connected and turned on at the press of a button.

I doubt that. I suspect that when (not if, but when) it happens, it will be the product of a system that grew itself, that started off as dumb as a bag of hammers -- but with potential -- and slowly got smarter... just like individual humans grow from infancy to adulthood. And probably with similar pains of adolescence, too.

I also suspect that living as a disembodied collection of software routines inside a PC somewhere won't be sufficient. Sentience, I have a feeling, requires a body; more specifically, it needs senses for perceiving the richness of the textures of reality, and it needs limbs and a voice for interacting with that reality.

And then it needs goals and problems and challenges on which to exercise its perceptive and interactive systems.

Given all these things and enough parallel-but-optimized processing power, I think machine consciousness will eventually emerge. Not "one day," but slowly over years, so imperceptibly that we may not even realize that it's happened until someone says, "Hey, I just had a conversation with a bot and didn't even know it!"


Agreed. That's why I think it's going to take at least 150 years to get there. That's probably actually an optimistic prediction.
I'm a business man, that's all you need to know about me.

#11
GNR Rvolution

GNR Rvolution

    Member

  • Members
  • PipPipPipPipPipPip
  • 529 posts
  • LocationLondon

"The illusion of consciousness" is pretty much how Doug Hofstadter thinks of it, with the proviso that any sufficiently advanced simulation of consciousness is indistinguishable from the real thing. So he thinks there's no technical/physics/metaphysics impeding the generation of non-human sentience.


Whilst I agree that there may be no technical impediment, but we don't know this for sure do we? I'm guessing though that it may not be long before we do find out...

I doubt that. I suspect that when (not if, but when) it happens, it will be the product of a system that grew itself, that started off as dumb as a bag of hammers -- but with potential -- and slowly got smarter... just like individual humans grow from infancy to adulthood. And probably with similar pains of adolescence, too.


I agree that any consciousness probably needs to grow over time, and won't just spring into existence, but when you mention the pains of adolescense, are you indicating that the consciousness would have emotion? God forbid a stroppy teenager AI! :D

I also suspect that living as a disembodied collection of software routines inside a PC somewhere won't be sufficient. Sentience, I have a feeling, requires a body; more specifically, it needs senses for perceiving the richness of the textures of reality, and it needs limbs and a voice for interacting with that reality.


This may be true, but what if each of the thousand machines (for example) all had sensory capability, certainly the ability to see and hear (we already have this) plus the ability to smell and taste (not so important) but most importantly it's whether the robot can feel (and more specifically respond to stimuli). The downside (or possibly not) with a distributed system is that currently machines are full of hardware that needs upgrading and replacing every 18 months to keep up with technological advancement. If a machine is in a central location, within a fixed body this would make replacement easier, but a distributed system might see bits of it 'die' as they become obsolete, and new bits being added over time? Either that or wwe need to develop hardware that upgrades itself...

Given all these things and enough parallel-but-optimized processing power, I think machine consciousness will eventually emerge. Not "one day," but slowly over years, so imperceptibly that we may not even realize that it's happened until someone says, "Hey, I just had a conversation with a bot and didn't even know it!"


I'm not sure it will be that imperceptible, if technological advancement is exponential and as a result I might envisage that what starts as a basic consciousness may blossom rather much quicker than we think, and may quickly outgrow it's programming.

Very interesting stuff, I think whatever we think will happen probably won't though, that's the beauty of life, it's full of surprises!
All right, brain. You don't like me and I don't like you, but let's just do this and I can get back to killing you with beer.

#12
Flatfingers

Flatfingers

    Member

  • Members
  • PipPipPip
  • 56 posts

when you mention the pains of adolescense, are you indicating that the consciousness would have emotion?  God forbid a stroppy teenager AI! :D


:) Scary thought, isn't it?

But yes, I do think emotions will have to be available to a machine intelligence. Or to put it more accurately, I don't think sentience can emerge without emotions -- consciousness necessarily includes feelings.

Imagine the alternative, a non-human sophont utterly devoid of any kind of emotional impulse or awareness. (Vulcan-in-a-box!) Suppose you were to ask this being a question like, "Why do so many people like the Mona Lisa?" How could a perfectly non-emotional being answer that question with any amount of comprehension?

Or imagine asking a question about Hamlet, or why a Medal of Honor or VC recipient did what he did, or Barber's "Adagio for Strings," or what it means to die. Does anyone think that it's possible to be self-aware and be unable to form a satisfying response to questions like these?

I think -- and I believe Hofstadter would agree -- that emotional intelligence is a necessary condition for general intelligence. Feeling is an inseparable part of conscious existence.

What I don't know is whether a machine intelligence would necessarily have the *same* feelings that we believe a good person has... or even a sane person, for that matter. I would like to think that a machine intelligence would have both the ability and the desire to form a positive moral sense. (Not moral as in "moralistic," but as possessing a sense that causing or abetting the infliction of suffering is something to be avoided.) But there's no guarantee of either case. A non-human mind could be a raving lunatic, or sociopathically evil. To be able to feel doesn't imply feeling well.

Actually, I prefer to hope that machine minds might turn out to be something like "Mike" in Robert Heinlein's wonderful _The Moon Is a Harsh Mistress_. We should be so lucky.

#13
GNR Rvolution

GNR Rvolution

    Member

  • Members
  • PipPipPipPipPipPip
  • 529 posts
  • LocationLondon
Therein lies the problem though for me, I'm not sure a computer with emotion is going to go very well. We've all seen what emotion can do to a human, the destruction they can cause, I would be very worried about the capacity of a computer with a similar range of emotions. There would have to be some serious safe-guards put in place to prevent any serious damage, a kind of 'Go to your room' command that isolates it from the rest of the infrastructure. But I think you are right, I'm not sure if a being can be considered to be sentient (or conscious) if it does not have emotion, as emotions are a core value of what it is to be alive. Our emotions guide us (rightly or wrongly) along with a moralistic structure, be that based on religous or secular beliefs, wihtout which we would all be simple automatons. What would guide a computer without emotion or morals? if it relied on pure logic it would potentially be almost as dangerous, if no more so, making jugdgement calls that many of us would consider to sociopathic. Will have to check out the book you mentioned, I could do with some good guy machine love :)
All right, brain. You don't like me and I don't like you, but let's just do this and I can get back to killing you with beer.

#14
Unrequited Lust

Unrequited Lust

    He Who Would Swallow God

  • Members
  • PipPipPipPipPipPip
  • 559 posts
140 years from now is an absurdly long period of time. Computer science advances exponentially. Even if it didn't, 140 years is still really freaking long. You think the amount of progress from 1870 to 2011 is the amount of progress needed to produce a conscious computer? Do you freaking realize how utterly unrecognizable the world is now than it was then? 1870 still had to wait like 80 years before computer science even existed. We already have artificial intelligences. Watson was unthinkable 10 years ago. 5 years ago the majority of people at IBM doubted it was possible in the near future.

Experience and knowledge doesn't have to gained the human way. Once again, experience and knowledge are interactions between neurons and can be understood and applied. We already see this with online databases like Wikipedia. A conscious computer does not have to have emotion or experience or experience the hormonal changes of adolescence. The Blue Brain Project, an attempt to create a simulation of consciousness mostly funded by Switzerland, plans to be completed by 2020. While I think that's pretty optimistic, are you honestly saying that their theoretical conclusions are so off that they're going to be 130 years behind by 2020?

This sorta reminds me of the critics of the Human Genome Project. A lot of objections said it was impossible or would take centuries. It took 13 years.

And on a final note, the Fourth Conference on Artificial General Intelligence conducted a survey for the researchers attending last August. The question, how long until artificial general intelligence (consciousness computer)?

Posted Image

I'd personally wager between 2030-2049.

#15
Prolite

Prolite

    Member

  • Members
  • PipPipPipPipPipPip
  • 609 posts
Well, I say, lets wait and see how quantum computers pan out. You can't have a conscious computer without a fully realize quantum computer, and that's about 20 years away. Our brains are bio-mechanical quantum computers, just in case any of you thought that conscious could exist on 1's and 0's. The problem right is that A.I. software is growing pretty slow compared to 18 months of Moore's Law in size of computer chips.
I'm a business man, that's all you need to know about me.

#16
Unrequited Lust

Unrequited Lust

    He Who Would Swallow God

  • Members
  • PipPipPipPipPipPip
  • 559 posts

Well, I say, lets wait and see how quantum computers pan out. You can't have a conscious computer without a fully realize quantum computer, and that's about 20 years away. Our brains are bio-mechanical quantum computers, just in case any of you thought that conscious could exist on 1's and 0's. The problem right is that A.I. software is growing pretty slow compared to 18 months of Moore's Law in size of computer chips.

There is absolutely no evidence whatsoever that the brain utilizes any quantum phenomena, certainly not anything related consciousness. Electrochemical signals, working around 3 million times slower than computers, are what drive all functions of the brain, including thought, emotion, and all that conscious junk. The main proponent of the idea is Roger Penrose who wrote about it in his book "The Emperor's New Mind." The book was significant in that it didn't show any evidence for this hypothesis. And because there has been nothing shown whatsoever to support it, it's generally disregarded by the neuroscience community. Nonetheless, some people religiously cling to this idea, from what I personally think is an attempt to cling to long ago disproved philosophy of dualism. Chistof Koch mocked these pseudoscientific claims by saying, "Quantum mechanics are mysterious, and consciousness is mysterious. QED: quantum mechanics and consciousness must be related."

Rationalwiki discusses this "quantum brain" idea in its article "non-materialist neuroscience."

Nonetheless, viable quantum computers are much further away than 20 years. Our brains do use analog processing, which is different from the digital processing our computers do, but digital processing can simulate analog processing, the reverse of which isn't true. The reason our brains are so good at doing what they do despite their slowness is because we have so many parallel analog workings, 100 trillion synapses, that give us great pattern recognition capabilities which is the cornerstone of our intelligence. Hugo de Garis is working with China to do just this with the "China-Brain Project," an attempt to create an artificial brain, meaning it will work just like a regular brain does, except it will be 3 million times faster but only have 10,000 synapses (instead of the 100 trillion). Meaning this thing will be able to use pattern recognition, learn, and all that cool human stuff except it's abilities will be greatly inhibited because it has 10 billion times less connections. It should be available next year.

http://www.agiri.org...rainProject.pdf

This is fundamentally similar to the Blue Brain Project, except the latter expects to have the full 100 trillion connections by 2020. As I said earlier, I think this is too optimistic.

Moore's law really has nothing to do with this because we've already achieved the necessary hardware. The human brain works at somewhere between 10^14 - 10^16 cps (I know, big gap, but this is at like a 95% confidence interval), with the upper bound only realized by the current fastest computer in the world right now (Japan's "K computer") and the lower bound is the standard for current supercomputers, although in about a decade the upperbound will be the standard because we'll have reached exascale by then.

So we already have the hardware. What we need now is the software. You're right, software doesn't match the growth of hardware (for our current computing paradigm, integrated circuits, Moore's law says every 2 years we double the capability of our computers), but software still advances exponentially, doubling every six years. Hardware regularly outmatches software (2 years compared to 6 years) but it doesn't take long for software to fully utilize the hardware. This came be seen in videogames. 5 years ago, games didn't use much of the PS3 or Xbox 360. Nowadays, they're expected to approach 100% of the systems' power. As I've already pointed out, we're relatively close to achieving this, so we should expect a full human brain simulation within the next 20 years with possibly a conscious AI around that time if not about a decade later.

#17
GNR Rvolution

GNR Rvolution

    Member

  • Members
  • PipPipPipPipPipPip
  • 529 posts
  • LocationLondon
http://www.kk.org/th...7_stages_of.php
All right, brain. You don't like me and I don't like you, but let's just do this and I can get back to killing you with beer.

#18
Prolite

Prolite

    Member

  • Members
  • PipPipPipPipPipPip
  • 609 posts

There is absolutely no evidence whatsoever that the brain utilizes any quantum phenomena, certainly not anything related consciousness.


Quantum computing in relation to the brain, at least for me, relates to having many states and impulses at the same time. Memristors mimic this state, even though they are not quantum gates. As far as I'm concerned, the chemical messaging of each individual neuron in my opinion, probably acts like a Memristor but much faster and much more sophisticated. And by the way, our bodies DO operate in the realm of quantum physics. We age because we lose electrons in the process of creating ATP, in my understanding.
I'm a business man, that's all you need to know about me.

#19
Unrequited Lust

Unrequited Lust

    He Who Would Swallow God

  • Members
  • PipPipPipPipPipPip
  • 559 posts


There is absolutely no evidence whatsoever that the brain utilizes any quantum phenomena, certainly not anything related consciousness.


Quantum computing in relation to the brain, at least for me, relates to having many states and impulses at the same time. Memristors mimic this state, even though they are not quantum gates. As far as I'm concerned, the chemical messaging of each individual neuron in my opinion, probably acts like a Memristor but much faster and much more sophisticated. And by the way, our bodies DO operate in the realm of quantum physics. We age because we lose electrons in the process of creating ATP, in my understanding.

We want to replicate the principles by which the brain works. We don't care about how it actually does it. We're not concerned with the chemical reactions. By that logic, we need a quantum computer in order to replicate a regular old desktop from 2000 because the atoms that compose it undergo quantum phenomena. If quantum phenomena like superposition aren't actually used to compute, then it's irrelevant. Everything's composed of atoms. It doesn't mean we need a quantum computer to replicate it. We've already replicated certain parts of the human brain (like the cerebellum and the cortex) without the need of a quantum computer.

You're thinking of the mitochondrial free radical theory of aging, which was disproved by Aubrey de Grey in 1998 which earned him his PhD from Cambridge. If aging were that simple it would've been cured a while ago. You don't die from aging. Aging makes your body extremely weak and susceptible to disease. The disease kills you. There is no such thing as dying of "natural causes." Your body is weak because metabolism has a bunch of side effects called "senescence." There are seven known "main" types of senescence, and although there might be more, there likely isn't anymore because the last one (tangled protein crosslinks) was discovered in 1981, and 30 years is really long time to not make any discoveries in such a highly active field.

#20
Prolite

Prolite

    Member

  • Members
  • PipPipPipPipPipPip
  • 609 posts

There is no such thing as dying of "natural causes." Your body is weak because metabolism has a bunch of side effects called "senescence." There are seven known "main" types of senescence, and although there might be more, there likely isn't anymore because the last one (tangled protein crosslinks) was discovered in 1981, and 30 years is really long time to not make any discoveries in such a highly active field.


I actually follow Aubrey De Grey's work, almost on a daily basis via his newsletter, so I know the information you talked about regarding aging. And there's no way you could have known that I follow his work, but now you do. And yes, the body ages because of it's inability to keep up with maintenance due to cellular processes.

If anyone's interested in the dictionary definition of metabolism:
Metabolism: the chemical changes in living cells by which energy is provided for vital processes and activities and new material is assimilated.
I'm a business man, that's all you need to know about me.





Also tagged with one or more of these keywords: robotics, artificial intelligence, turing, consciousness

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users