Discover the best of the web!
Learn more about Digg by taking the tour.
Can Machines Become Conscious? Yes. A New Test Offers Proof.
spectrum.ieee.org — In some quarters it is taken for granted that within a generation, human beings —including you, if you can hang on for another 30 years or so—will have an alternative to death: being a ghost in a machine. We are confident that we are within a few decades of building a computer, a simulacrum, that can feel pain and pleasure, and fall in love.
- 146 diggs
- digg it
- bullcutter, on 06/05/2008, -2/+5No, the answer is no.
First of all, in regards to the Turing Test (can't distinguish the response of an AI device from a response by a human), the author of the article admits:
"No machine or program comes close to pulling off such a feat today."
Second of all, the author of the article strips down the definition of "consciousness" to something that can be conceivably reproduced by today's computing technology. It is a far cry from the definition of consciousness as applied to neuroscience and biology, which takes into account the collective action of specific neuroanatomic areas that gives rise to conscious behavior.
This is what the authors have to say about our attempts to successfully model the human brain thusfar:
"Simulating such a gigantic neural network model in the hope of seeing consciousness emerge, with millions of parameters whose values are only vaguely known, will not happen in the foreseeable future."
The human brain is the most complicated structure in the known universe, and if we can't model the nervous system of a 302-celled roundworm, what chances do we have of correctly modelling human consciousness??
We can make programs and machines like ALICE the chatbot that mimic what can be described as conscious-like behavior, but they are still rather incapable of generating original thought.
I majored and went to graduate school in neuroscience specifically so I could learn how to one day upload my brain to a computer before I die and live forever, but from the first few weeks into my introductory courses (spent a whole semester learning about a single neuron), I realized that this will not be quite possible, for a loooong time to come...- Culyt, on 06/05/2008, -1/+4The Blue Brain project already uploaded a rats neocortical column, we might not have a working model of a roundworm, but we do have a working simulation of something more complex on the molecular level (10,000 neurons, 1,000,000,000 synapses). They plan on a whole rat in 2 years and a human in 10.
http://en.wikipedia.org/wiki/Blue_Brain
http://bluebrain.epfl.ch/
We don't need a exact model or in depth understanding of how stuff works in order to actually implement it, we can either take what we have and make a copy, this is how they did the rat, they built a robot to get the electrical charge of each neuron and put it into a super computer.
In addition to that, shortly after the human level computing power is reached we will have such massive amounts of excess of computer power that even if we can't get it working, we can just leave it up to digital evolution, this already happens when we teach robots to recharge themselfs or catch a ball, these are not ALICE type programs, they are actually learning, they have been programmed to program themselfs. What would a very basic neural network that can do these things would do if we made it millions of times more complex. The main problems will be finding optimal evolutionary algorithms and defining what is considered a good evolutionary trait and providing an environment complex enough to stimulate intelligence.
However I believe we will be able to get very precise models of how the brain functions, brain scanning technologies are improving exponentially like computers are, nano robots could be used to give us information on the state of every single neuron in the brain, or just current brain scanning techniques vastly improved and combined together.
Worst come to the worse we will be able to simulate the atoms of the brain itself, although its unlikely that that level of detail will be needed, and its obviously very inefficient and the amount of mass required to simulate it would be more than brain itself (although hopefully less than a whole human body and the amount on energy required).
☢- bullcutter, on 06/06/2008, -0/+1"The main problems will be finding optimal evolutionary algorithms and defining what is considered a good evolutionary trait and providing an environment complex enough to stimulate intelligence."
Agreed. Our intelligence took billions of years of trial-and-error evolution to develop. There is no way to speed up this process without your results suffering a loss of applicability to the real world. You can speed up the rate at which mutation happens but then you are already losing some applicability to true human neurobiological development.
Again, the development of conscious awareness can be modeled but not recreated, nobody has the time or the resources. (billions of years and an entire, dynamic planet).
Show me a computer or machine that can build a copy (or otherwise "spawn") of itself first and then we'll talk about the possibility consciousness in machines.- tehbored, on 06/06/2008, -1/+1First of all, so what if it's only a model? That doesn't mean it can't be just as good or even better than the real thing. Recreating the development of intelligence is nothing but a novelty if it can't be used for anything. Models can be used. Evolving an intelligence from scratch seems like a huge waste of time, but since nature has already done it for us, why not just improve on what we already have?
Second, it seems perfectly possible to recreate intelligence from scratch, but as you say it would take tremendous amounts of time. Certainly we could figure out a way to have computers help us along. It doesn't have to be done in the physical world, it can be done virtually. Though even so it could still take many decades since it is such a huge task.
Also, you can live forever without uploading your brain. Death is perfectly preventable with the right medicines and technologies. We're not that far off.
- tehbored, on 06/06/2008, -1/+1First of all, so what if it's only a model? That doesn't mean it can't be just as good or even better than the real thing. Recreating the development of intelligence is nothing but a novelty if it can't be used for anything. Models can be used. Evolving an intelligence from scratch seems like a huge waste of time, but since nature has already done it for us, why not just improve on what we already have?
- bullcutter, on 06/06/2008, -0/+1"The main problems will be finding optimal evolutionary algorithms and defining what is considered a good evolutionary trait and providing an environment complex enough to stimulate intelligence."
- rmtatum, on 06/06/2008, -2/+1Is it not only not possible from the biological perspective, it is not possible because man has a spirit. We are spiritual and physical beings. Computers will always be limited. Artificial intelligence should really be called computer augmentation. The machines are not really intelligent. They are only as smart as the programmer makes them to be. Mankind, however, was created in the image of God. Man will never be able to surpass God's intelligence by creating machines that are equal to or greater than himself. Man still does not completely understand his own body!
- Culyt, on 06/05/2008, -1/+4The Blue Brain project already uploaded a rats neocortical column, we might not have a working model of a roundworm, but we do have a working simulation of something more complex on the molecular level (10,000 neurons, 1,000,000,000 synapses). They plan on a whole rat in 2 years and a human in 10.
- scamerica, on 06/05/2008, -5/+2I predict that the first machine sparked to consciousness will be totally insane. Consciousness is a delicate equilibrium between madness and sanity and achieving a conscious machine that is mentally sane will be a great challenge.
- monoa, on 06/05/2008, -1/+3A prediction based on what? You're spouting uninformed nonsense, as usual.
- tehbored, on 06/06/2008, -0/+2Science fiction probably. This theme dates all the way back to Frankenstein.
- Waiting2awake, on 06/06/2008, -0/+1I am intrigued by this "sanity" you speak of...
- Culyt, on 06/05/2008, -1/+4I predict the first machine will have a fondness for pie and enjoy classical jazz. ☢
- ordig, on 06/06/2008, -0/+3Open the pod bay doors HAL...
- monoa, on 06/05/2008, -1/+3A prediction based on what? You're spouting uninformed nonsense, as usual.
- Rabbittt, on 06/05/2008, -1/+2Our Universe is a machine that is in the process of becoming self-aware..
- bullcutter, on 06/06/2008, -0/+2it already has, through the human mind. we are the universe, er.. we came from bits of it anyway.
- reisrocks, on 06/06/2008, -0/+2told you not to log onto digg when on those shrooms ;)
- 3tcp, on 06/05/2008, -2/+1"Can machines become conscious?"
This only way to answer this discussion is with a philosophical debate - one that has been had many times over and never ended with a consensus yes or no answer. The ability to create a computer program that may behave as if it were my dead grandpa has no implications regarding whether or not it is my dead grandpa.
Ones existence cannot be uploaded and ones personality changes, so a program that simulates my dead grandpa would only do so for his personality at a given point in time. If it were programed to evolve as if it were real then it still wouldn't be my grandpa. It would just be a simulation based on my grandpa's personality... if he had known that he was really just a computer program. His further experiences that determined the changes in his personality would be heavily influenced by the fact that it's not a real person. It's like the plot lines of data/the doctor on star treks merged with every 'clone' movie or tv episode out there. - DeskFlyer, on 06/06/2008, -1/+2Anyone who's seen Short Circuit knows this already.
- EndersGame21, on 06/06/2008, -0/+2Can't remember the science fiction author who said this, but "being human can be summed up in two words: 'like me.'" I wonder how long it will be until a computer analyzes a human's emotions and says "hey, that's just like me."
- coboman, on 06/06/2008, -0/+3There is a huge difference between being able to act as a conscious being and actually being conscious.
Deep Blue acted as an intelligent chess player. So much so that it defeated Kasparov, the best chess player there has been.
That doesn't mean it was intelligent, or that it knew how to play chess. It just looked that way.
Consciousness is a subjective phenomenon. It's existence is in the mind of the beholder. The ability of an individual to recreate and live the world inside of his own mind. This can only be achieved by living beings. Machines will be able to imitate consciousness, they already do. But it is only and act and it will never be the real thing. - Gabberwok, on 06/06/2008, -1/+2There will certainly be conscious machines in the future, our brains are complicated but there's nothing magical about them. We'll never be able to "download" our minds however - it's hard enough to visualize synapses in vitro. Mapping every synapse of the brain could theoretically be possible (although it would almost certainly require destroying the brain), but there would still be critical information missing about the state of each cell (protein expression levels, resting potential, etc. etc.) . It ain't going to happen.
- reisrocks, on 06/06/2008, -0/+2Dugg down for being rambling bullcrap that won't happen within our lifetimes.
- blankman, on 06/06/2008, -2/+1an alternative to death is uploading your memories into a computer? Don't these people understand that "You" will still be dead, just because there is a copy of your brain on a computer. Copying yourself doesn't 'transfer" you. Whatever makes us aware of ourselves is either a biological product specific to ourselves, or a soul. Neither of which can be transfered that we know of. A "transfer" into a computer may make that computer conscious somehow, but it wouldn't be "our" consciousness.
- bullcutter, on 06/06/2008, -1/+1yes thats true, good point. that's another good reason I stopped my pursuit.
every time somebody got "beamed" in Star Trek, they died and began life anew as an exact copy of their old selves, who thought they were still their old selves as far as they were concerned.
the same principle applies to the whole "neural uploading" thing.
i suppose it'd just be easier to keep your brain alive in a jar. maybe put it on a nice, fully-armored mech body or something.- blankman, on 06/07/2008, -1/+1well in fairness to Star Trek, the transporter didn't really make a copy of the person in another place. It simply took them apart molecule by molecule and reassembled those same molecules in another location. In that situation it's most likely we'd still be ourselves.
But just think of the computer brain download for example. Let's say we did it while still alive, you'd still be you right? Just because a computer now has your memories and may act like you doesn't make it you.
People really shouldn't be so afraid of death. You will either live on for eternity if there is really a soul, or you will cease to exist. Either way you'll be fine. It's your family and those that care about you left behind that suffer.- bullcutter, on 06/07/2008, -2/+1How does death not occur when somebody is taken apart molecule by molecule?
And really the pursuit of life extension isn't really a fear of death so much as a desire to use breakthroughs in technology to extend life, which has been going on in some form or another for thousands of years. Only with recent developments in our understanding of human genetics and cell biology did it really become an organized movement, in the last 20 years or so, but when we develop a technology it is only natural to pursue the limits of its beneficence to mankind.
- bullcutter, on 06/07/2008, -2/+1How does death not occur when somebody is taken apart molecule by molecule?
- blankman, on 06/07/2008, -1/+1well in fairness to Star Trek, the transporter didn't really make a copy of the person in another place. It simply took them apart molecule by molecule and reassembled those same molecules in another location. In that situation it's most likely we'd still be ourselves.
- bullcutter, on 06/06/2008, -1/+1yes thats true, good point. that's another good reason I stopped my pursuit.
Digg is coming to a city (and computer) near you! Check out all the details on our