In my last post (quite a while ago now), I talked about IBM’s Watson playing Jeopardy. I introduced the Turing test and gave a short analysis of Watson’s performance. The discussion that ensued examined some interesting points and I’d like to elaborate further on some of these issues. In this post I’d like to talk about consciousness and in future posts make the jump to discussing conscious robots. For most of this post, I’ll be taking notes from a paper that appeared in the AIAA Journal of Aerospace Computing, Information, and Communication [1].
Definitions
To even get the discussion off the ground we’ve got to disambiguate a few important terms, namely intelligence, autonomy, and consciousness.
Intelligence
Kurzweil defined intelligence as “a very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience” [2].
Autonomy
Autonomy, OTOH, “refers to systems capable of operating in the real-world environment without any form of external control for extended periods of time” [3].
It’s important to note that intelligence and autonomy can be largely uncoupled. For example, an autonomous aircraft can operate (i.e. fly, communicate information, etc.) for long periods of time, but it may not reason, or comprehend ideas. OTOH, a supercomputer that has software simulating intelligence might not be able to do anything of interest, but it can reason, plan, and solve problems.
Consciousness
Clearly people disagree about what it means to be conscious. Descartes’ dualism is still alive and well in both philosophical and theological circles, including Mormonism. If we include the dualist notion in our definition of consciousness, we may as well end the discussion here and now. Technology cannot put a “spirit” or “soul” into a robot, therefore, there cannot be any conscious robots ever. If you take this view, I invite you to elaborate in the comments below with reasoned arguments as to your position.
I am not convinced of the dualist nature of human beings, and consider it a distinct possibility, indeed one that seems to be supported ever more by our advancing scientific knowledge, that humans are physical machines whose consciousness has sprung as a natural consequence of evolution and natural selection.
With this possibility in mind, a well accepted definition [4] of consciousness revolves around the concept of self-awareness – which seems to separate us from most everything else in the natural world. The “state of awareness” includes:
- Subjectivity: our own ideas, moods, and sensations are experienced directly
- Unity: all sensors are melded into one experience (i.e. we don’t consciously separately process vision data from our eyes and audio data from our ears, rather those data are processed and combined into an overall sensory “experience”).
- Intentionality: experiences have meaning beyond the current moment.
It is worth mentioning that consciousness is also associated with attention, which brings objects, concepts, etc. into our consciousness. This allows us to process huge amounts of information.
Implications
If we accept these definitions, we can make some interesting observations. First, there are
other animals that appear to be “conscious.” Dolphins, elephants, gorillas, magpies, chimps, orangutans, orcas, and some pigs exhibit self-awareness by passing the mirror test. It is interesting to note that dogs, cats, and young human babies all fail the mirror test (an important point that will be elaborated on in later posts). In fact, humans don’t generally pass the mirror test until they are 18 months of age.
Second, this POV certainly has far reaching implications for various theologies. If humans are only physical machines, is there a life after death, or do we cease to exist when we die? Does such a POV imply that spiritual experiences are nothing more than neurons firing together with an active sub-conscious and limbic system? If other animals exhibit “consciousness” and we accept the dualist position, does this mean other animals are capable of spiritual experiences?
Third, if we adopt this POV, we can make a very strong case that humans will indeed invent robots capable of obtaining consciousness. We’ll explore this, and more thoroughly discuss consciousness in the next post.
For now, what are your thoughts on consciousness, and the possibility of developing conscious robots, or more likely, robots capable of obtaining consciousness?
[poll id=”22″]
[poll id=”23″]
[1] Long, Lyle and Kelley, Troy. “Review of Consciousness and the Possibility of Conscious Robots,” Journal of Aerospace Computing, Information, and Communication. Vol. 7. February 2010.
[2] Kurzweil, R., “The Singularity is Near,” Penguin Group, London, 2005.
[3] Bekey, G.A., “Autonomous Robots: From Biological Inspiration to Implementation and Control,” MIT Press, Cambridge, MA, 2005.
[4] Dennett, D.C., “Consciousness Explained,” Back Bay Books, Boston, MA, 1992.
Wheat & Tares 
The earth has a soul. I see no reason why created things may not have them.
Though you haven’t defined the status of cyborgs, if I make a pig/computer does it still have a soul?
This is just too tantalizing to pass up. A great post.
First, I want to consider the possibility – a kind of Mormon-ey idea, if you will, that human souls occur at the highest level of organization (that is, it is an emergent property) of cells, systems, histories, memory, all working together and coming from them, a sort of “ensoulment.” This intends that the “real us” or the “spiritual us” comes at the highest, not lowest level of organization. Our spirit is not some kind of core, real thing. We are the real thing.
Second, all that this “real thing” (the soul) requires to transcend death is a medium to support it. We are hearing, for instance, about “downloading” our brains into computers. If God is imminent in creation, intimately knowing us (and, as well, knowing the array of our molecules) why could that memory of us, His consciousness, serve as the carrying medium for human consciousness (heck, chimpanzee consciousness) after death?
Why not before birth?
And, who says that humans did not first emerge (this is in the Mormon premortality) as physical beings? Why would physical Gods have “spirit” children anyway? That is not how things work.
So, what I am hastily suggesting is:
Yes, Virginia, it is all physical
Mortality is the humans “proper frame”
That when not embodied, human consciousness has to reside on some inferior medium (spirit, the light of Christ)
Consider the possibility of antemortal embodiment because without embodiment we cannot begin to exist
Joseph Smith’s term “intelligence” refers to the highest, not the lowest level of human organization.
Great post.
Interesting post I have given this a lot of thought here is some of what I have considered. Intelligence is the easiest much has already been achieved but computers are a high contrast (binomial) medium and not easily adapted to the human grayscale contrast of vertical layers each increasing in contrast from perhaps 3 subconscious to maybe 5 conscious layers and the horizontal varying contrast of fading memory not to mention the software necessary to oversee and coordinate these. In addition a sense of self will probably require some scheme of selective “me” related short-term memory laid over similar long term-memory. All of this is complicated by interaction with a relational database that offers more hits to lower levels so that new ideas tend to precipitate upward to higher levels. It has the potential to take thinking to a new level by dropping out emotion related issues that interfere with clear thinking like jealousy, embarrassment, insult, fear of loss of creditability, competitive intellectualisation but to be ethical and moral should include some level of empathy. Autonomy is a balanced decision making process where components are weighted and voted but this can be very complex because it takes place on multiple levels.
Re Stephen-
I’m curious about what you mean when you say “the earth has a soul.” Could you elaborate a bit more?
Re Mark Clifford-
Very nice ideas. I really like what you’ve described.
Re Howard-
I’m so glad you’re already wading into the details. That’s exactly what I hope to talk about in the future. It’s clear that we currently do not have near the processing power in modern computers to rival what a human can do. In most of the stuff I’ve read, our modern computers have about as much computation power, memory, etc. as a cockroach. Yet, we do not have near the autonomy yet. But we’re getting there.
Re All-
I’ve noticed that in the polls (at least so far) people believe there is a soul/spirit, but also believe we’ll certainly achieve conscious robots. Does this imply that people believe that consciousness can be obtained without a soul, or that our robots will somehow obtain a soul? If the former (and you voted this way), are souls decoupled from consciousness, or merely uncorrelated, or what?
Re All
I’m not sure what you mean by soul but I believe our spirits are our minds at a subconscious level and they contribute greatly to our thinking and creativity on a conscious level in addition this communication must be the pathway of inspiration. I also believe our spirits are innocent and therefore we construct a psyche to interface with the mortal world. God must work with natural laws not magic so AI done completely would include a spirit but this poses a challenge much greater than the others.
I don’t buy the fact the earth has a soul. Just because something was created spiritually before created physically, doesn’t mean it has a soul.
My house had a blueprint that spiritually created the idea of what it should look like, prior to creating it physically. My house doesn’t have a soul…although my kids think its haunted…I don’t believe it.
I personally like the hand and glove analogy I taught on my mission. A person is both body and spirit. The resurrected savior has both body and spirit.
Great post.
I think I mentioned this in another post, but it is a thought experiment that applies perfectly here. What if we replaced a singlue neuron with a machine representation of that neuron – would we still be “human” or “conscious”? I would say yes, much like replacing a knee or a heart valve to replace a function of the body with something mechanical.
What if we continued replacing neurons in a similar fashion? Is there a point where we would NOT be conscious? What if we replaced them all?
Good question Mike S. I think we would remain conscious however any difference in frequency response between neuron and machine would affect our processing speed.
Yes, Virginia, it is all physical
Then why does it seem that mine was the first vote for the physicalist view?
Furthermore, I reject the characterization of LDS teachings as dualist. Joseph Smith, at least, was not a dualist. D&C 131:7-8 would seem to shoot a big hole in that notion.
Most people nowadays who criticize dualism have no idea what it actually is, and have never encountered its strongest arguments. Even Descartes himself never said that our true selves are on the other side of some kind of screen, and the matter which makes up our bodies is a glove that we put on our true selves. The dualism argued against by modern scientists is a complete straw-man that was never advocated by serious philosophers of mind.
Anyone who thinks that someday materialism will yield qualia is just guilty of “materialism-of-the-gaps” thinking. There is no way we can yield consciousness from matter even in principle… ever. We still have the binding problem and the unity of consciousness problem and the non-algorithmic nature of consciousness to deal with.
A good refutation of popular arguments against dualism can be found at Edward Feser’s blog here, here, here, and here. Churchland is supposedly one of the most-respected in the philosophy of mind, and even he doesn’t even understand what dualist philosophy actually says.
Re Syphax-
DISCLAIMER: I’m not a philosopher, and readily admit that. However, I have studied some of the arguments both ways so I know a couple things.
When I read this, what I read is “people who don’t see things the way I do, don’t understand what dualism actually says.” Perhaps you don’t intend it that way. I think there are VERY good arguments against dualism as proposed by Descartes. I’ve read some of Feser’s stuff before. I like him and his style, but he absolutely has that same arrogance about him and his writing IMHO. He’s always right, no matter what (and a bit condescending IMO).
Also, while I acknowledge your point about the “materialism-of-the-gaps” (a knife which cuts both directions), I don’t believe history and experience is in your favor. Science and technology have a way of filling in the gaps that appeared to be only fillable by God. This isn’t any proof, but personally, I think it is a reasonable inference that science WILL fill the gaps of consciousness and we’ll have conscious robots.
I’m not sure why you conclude that consciousness is non-algorithmic. Do you mean that consciousness isn’t deterministic? That’s surely a point to be debated. Nevertheless, it’s easy to write algorithms that are non-deterministic (or at least appear to be so to within the same precision that humans appear to be non-deterministic).
Finally, I think you fall into the trap that I see too many philosophers falling into. You’re emphatically making a final declaration that is limited to the rules of propositional logic. Frankly, I reject the notion that we are limited to those rules in our thinking, and I reject the notion that the world hinges on logic, mathematical formulae, etc. In other words, I think we build mathematical formulations, and logical worldviews that approximate the world around us. But I don’t think any of them perfectly describe the world around us.
Re Last Lemming-
Interesting. You’re right, I failed to acknowledge this distinction. It seems then, that taking this view would allow someone to believe in conscious robots, the physicalist view, and a life after death, and “spirit” view.
I feel like Joseph is splitting hairs here, perhaps trying to play both sides of the field. Any physical material will have physical properties, by definition. If the material cannot be discerned (except “by purer eyes,” clearly not realizable under physically creatable circumstances), then we might infer that it has no physical properties, and therefore isn’t physical is it?
It seems to me that latter-day saints are dualists, we just modify our vernacular a bit to give a nod to science and physicalism, but throw in the “purer eyes” bit to understand why we can’t actually have any evidence for the “more refined matter.”
But the articles that I linked show that most of Churchland’s stated objections to dualism are not actually dualist arguments at all, and don’t affect dualism at all. I’m not saying those who disagree with me don’t know what they’re talking about. I am talking about Feser’s articles that clearly state that Churchland is arguing against straw men.
I think materialism has accounted for material things, yes, and I expect to see plenty of scientific evidence for material things because the scientific method was developed specifically to understand the material world. But saying that adding up material can somehow get to consciousness is like saying that if we continue to calculate the digits of pi, it will one day get to 4. It’s a different category. There is no “particle” or “wave” that will ever account for consciousness, and the burden of proof I believe is on the materialist to justify such an assertion.
I mean consciousness is non-algorithmic in the way that Penrose stated.
Finally, I don’t understand why you would undermine the use of propositional logic in philosophy, and then state “I think we build mathematical formulations, and logical worldviews that approximate the world around us. But I don’t think any of them perfectly describe the world around us.” Interesting philosophy.
I would hope that any theory of consciousness that emerges is capable of unifying the concept across both physical and spiritual realms,
I’m not happy with dualistic views of reality, although O do think there can be dual descriptions of a single reality.
“other animals that appear to be “conscious.” Dolphins, elephants, gorillas, magpies, chimps, orangutans, orcas, and some pigs exhibit self-awareness”
Intelligent. When they can build a microchip or airplane, let’s talk. Heck, for that matter, let’s start with building a fire.
We are children of God. That is what seperates us from any other living creature. That is what gives us dominion over them — kill them, eat them or use them to plow our field. As I tell my animal loving sister ( who wont eat meat for this reason), one person’s meal is another person’s pet.
Will what animals have you gotten to know?
Re Syphax-
This is interesting. To me it makes perfect sense to say we could add up material and build something capable of obtaining consciousness. And the analogy to pi isn’t a good one in my book. When we build systems we interface many parts together. One of the biggest challenges facing engineering right now is that few people know how to be system engineers. The school of thought has been that if we create all the parts independently, then stick them together, things will work as designed. Turns out that’s not the case. Unhandled cases arise, and the system takes on different characteristics. There has been a push to develop engineers who can design for this.
The point? Adding up individual parts creates a synergistic effect with unknown quantities and qualities. The physicalist would claim precisely that humans are indeed the addition of many small parts to create something capable of consciousness.
This is the other big problem I see with naysayers. Most serious AI people, psychologists, etc. believe we will build a robot capable of obtaining consciousness. I don’t know of anyone who believes we’ll ever be capable of building a completely conscious robot in one shot (which is perhaps your point with which I agree). This makes sense since humans are not conscious (with reasonable arguments) when they’re born, but become conscious with time, growth, experience, etc. Humans have the capability of becoming conscious but it’s not a given. So I agree we can’t put a bunch of materials together and get consciousness, but we CAN, IMHO, put a bunch of materials together and it’s possible that consciousness will emerge.
Also, this is where I get lost on the dualist arguments. You claim it falls to materialists to explain consciousness, and I think that’s backward. To me, the default position is the physicalist view and it’s incumbent on dualists to explain why I should believe in a non-physical part of me when it appears everything else in the universe is indeed physical. It’s akin to saying that the default position on the cause of disease was demonic possession and it fell to materialists to convince us that germ theory was appropriate. Indeed that’s the way it was, but it was totally backward and impeded progress for way too long!
Thanks for your insights!! I’ve enjoyed the discussion.
Oh, and I don’t mean to undermine propositional logic. I don’t reject it or anything. I just think it falls short of perfectly describing the world, just like mathematics. As a result, I have a very hard time with definitive absolutes.
I think it’s interesting, and quite ironic, that one of the side effects of this emergent property of consciousness is that it allows those who possess it to believe that they are magically separate from all of its parts, when it is all of those parts that allow them to think about themselves in the first place.
This is an area of study that deeply intrigues me. It’s like a logic alarm that flashes red until I answer it. I find it interesting that we think that we could develop consciousness merely from a machine. I am not sure what part of me asks if people honestly believe this, but I do find it intriguing that we think technology could develop this, of the which, logically it never has any chance.
That said, let me explain. There is a difference between something that is living and something that is not such as a machine or computer/robot etc. That difference is that one experiences senses and the other doesn’t. Our decisions are based upon intelligence coupled with our senses, including spiritual intuition. Computers on the other hand only function according to input. It makes no real intelligent decision like we do and never will have thta capacity. Why? Because it can’t be made to “feel”. I don’t care how sophisticated you made that robot, as long as no part of it was incorporating living tissue, I could bash it with a sledgehammer and it wouldn’t feel pain, wouldn’t feel anger, etc.
This of course leads into the question of living things- are they mere robotic machines? No- absolutely not! We know from experience that living things are different than mere “machines”. That difference is obvious. So what is it that makes life so intelligent and “conscious”? I beg to say that it is our spirit. Like it says in the scriptures, God made things both that can be acted upon and things that can act for themselves. Rocks, minerals, etc, all are objects that are merely acted upon. They have no life and no conscious. Things made from these structures will only have the capacity to be acted upon be it a watch, computer or even a robot. Living substance on the other hand is blessed with special intelligent parts and given a spirit of intelligence and things made from this substance can thus act for themselves.
This barrier cannot be crossed. The one question I always throw out is this- If we think it is possible to create a robot with a conscious then it can be said that true intelligence can be written in the form of a mathematic operation. But, math, no matter how complex, cannot be written to express “feelings”, a trait that would be required for true consciousness.
So, no, we shall never be able to create a robot or machine with a conscious and thus no true intelligence.
jmb275 – such that pi never adds up to 4, but 3.9_ when it goes far enough, does. I had not thought of that approach.
Rob Osborn — so you think that self-awareness is not possible with machines? That was the entire point of the OP’s reference to the mirror test.
Degree of self awareness seems to correlate with development and intelligence. So it might arise if that is all there is.
I’m not so sure, but not certain it is impossible either.
Stephen,
Self awareness is impossible from a mere machine, no matter how sophisticated and advanced it is. Sure, it could be programmed to know it’s own signature attributes by looking in a mirror at itself but it could never be made to “feel”. It could also never come up with anything intelligently novel on it’s own, a known trait of true intelligent entities.
AI may get more and more intense because of the extensive programming and algorythms used. It may even appear on the surface to be able to make decisions like humans but it will always come down to merely computing a programmed output from gathered input information. Humans do not make decisions like that. Our decisions are based on feelings, impressions, etc.
There is a science fiction series whose original sources had Mormon connections, Battlestar Galactica. (It’s being rebroadcast on BBC America beginning in June.) Its theme was exactly whether and how humans were different from machines (Cylons). During its original run I spent some time on chat boards for the series, and discovered how passionately people feel about this issue.
Until we know what consciousness, feelings, etc. actually are, our arguments will tend to be circular. I.e., because humans ARE machines, of course machines can become conscious. Or, alternatively, because humans ARE MORE THAN machines, machines can never become conscious.
So, what’s your theory of consciousness, Ron?
Then I could understand what you’re saying.
Firetag,
Describing consciousness is like explaining what color is to a blind man who has never experienced it. Where does consciousness come from? What is it’s workings? Who knows? It’s a true enigma that is at the moment outside of our knowledge. But, we can state through other precise testing and observation where consciousness is not and will never be. For instance- we know that a calculator can only do exactly what it is programmed to do. It can’t do anything more. We also know that machines such as calculators and computers and robots can’t be programmed to re-write programs already existing giving itself new and more novel information. We also know that senses such as hearing, sight, sound, feelings, etc, are only found to be part of living things. Non-living things, no matter how constructed, do not yield the capacity to have senses that are either testable or observable. Senses are in large part what makes us intelligent beings. It’s what makes us unique.
We have the ability, unlike machines, to take all the sum of our parts, our conditions, and couple that with our feelings and senses and make uniquely intelligent decisions based upon the sum of those things. A machine, no matter how sophisticated can only do tasks exactly how it is programmed to do. Machines do not have neither senses nor feelings to base those calculations off of. Neither does it have a truly smart and intelligent engine to create new novel information like we are capable of. Intelliegence- true intelligence is not the mere sum of mathmatical propositions. Why I decide I like sex or candy, or going to the movies is not some mathematical calculation- it cannot be written like that in which we program computers. It really is that simple!
We simply are not machines. We are living breathing and intelligent entities that are capable of intelligence that no machine can be capable of aquiring.
I really enjoyed this post—not so much the comment thread. One of the things my English Degree provided me is knowledge that the quest for an absolute definition of anything is always frustrating. No. Almost always frustrating. I’m curious to see where future posts will go with this topic. The social and ethical implications need deliberated. Here is an example:
I’m reminded that some electronic equipment are called “slave units.” If any of these devices develop awareness, let alone a sense of morality or yearning, I wonder how eager we enlightened humans will be to give them rights and equality. Perhaps we will give them a little freedom and a list of commandments.
It’s also tempting to only speak of computers rising to consciousness in scenarios where we graciously accept them or they terrorize us. There is also a context in which we might envy them, going so far as to attempt transferring our consciousness over to electronic formats and realms. This would be done as a means to achieve immortality. Such a feat is a long way off if it is possible at all. But it’s already a pipedream in some circles that have included persons of wealth and influence.
At any rate, my frequent dissatisfaction with my environment–and my attempts to modify said environment to my liking–are the best evidences I can point to that I am conscious, aware, and maybe autonomous. Put another way, when my heart breaks I write poetry.
Many basic senses have already been demonstrated in robots. What are feelings if they are not complex electrical and chemical signals? As such they can be reduced to algorithms.
But senses for humans are what makes us “feel”, meaning- what makes us have joy. Robots do not experience feelings of any kind. Demonstrate through observations that a robot can “feel” and I will believe you. But, I am not going to hold my breath cause it won’t happen.
Feelings are distinctly separate from senses but senses provide rich stimulus for feelings even joy. Sure artificial feeling has not been developed but electrical and chemical signals can be modeled using algorithms. When artificial feelings are developed they will be interconnected to a robot’s senses and thinking.
Artificial feelings? This is the exact thing that separates something from being conscious or not. Are you saying that if we can closely mimick true feelings with artificial feelings then we can achieve consciousness? What defines artificial feelings? Is it a robot that changes it’s level of activity according to some programmed input? How would a robot feel bad about accidently breaking a dish it was washing? Is it programmed to occasionally drop things breaking them? Is it also programmed to react in a thus exact manner according to how it dropped it? For that matter, is it programmed to commit suicide if it bothers itself so much over breaking the dish? I have no idea how one could mimick true feelings in a computer program. Computer programs do not take into account any true feelings. It’s always and only- if this then that, if that then this, etc etc. This brings up another interesting point about choice and agency. Can a computer be programmed to do things according to it’s own will? No, it can only act according to how it’s programmed. It doesn’t express any free will because not only is it against it’s programming, it doesn’t have any intelligence to act upon to excersize any free will.
I am surprized we even can imagine robots and computer programs with a conscious. Machines are not capable of consciousness. Will the AI community ever get over the fact that a machine will always be just that- a machine incapable of true emotions, feelings, etc?
Rob:
“Where does consciousness come from? What is it’s workings? Who knows? It’s a true enigma that is at the moment outside of our knowledge. But, we can state through other precise testing and observation where consciousness is not and will never be.”
That’s the essence of what I see as the problem with your argument — the leap from “Who knows?” to “will never be.?
I think God used evolution a lot. I know personally compelling evidence that traces our linage back to things that were clearly not conscious. Made from the dust of the earth isn’t just metaphor. Something inanimate became animated and then conscious. Confident that that happened, but not understanding how, I have no confidence that God hasn’t rigged things so that continues to happen.
I think the interesting challenge is not “prove machines can be conscious”, it’s prove that humans are not organic machines.
Rob:
I should add that I’m interested in a specific theory of consciousness that I wrote about in an earlier post here on W&T. It has a lot of interesting features in deriving aspects of Mormon belief. For example, combine it with modern cosmologies of parallel universes, and you get the pre-existence of human spirits popping out as a freebie. As both a believer and a scientist, I like to play with theories with such features.
http://www.wheatandtares.org/2010/10/08/the-spirit-of-the-earth/
Rob 28 No it will require more than mimicking feelings to achieve consciousness it will also require a “me” working memory and “look down” programs that monitor and interact with multiple layers of programing. Yes computers typically use if then logic but they are not limited to that for instance there is fuzzy logic that allows multivariable fluid reasoning. In 3 I roughly outlined architecture I believe will be required.
It is not useful for you or a robot to feel badly about breaking a dish so why should robot feelings be attached to something like that?
Firetag,
I personally don’t believe in evolution. I see evidence pointing to an intelligent and purposeful creation.
Are we machines? Define machines? If we state a machine is a object that operates according to input that has moving parts and serves a purpose then in that sense we may be considered some kind of machine. But the machines we talk of with computers and robots which utulyze no living tissue are purely input output contraptions that only do exactly what they are programmed to do. Under that definition then by no means are humans “machines”.
I am not sure how, but our spirit of intelligence adds the ability of moving us beyond input output machines. In a sense we are an “operator” that is housed in the machine. Coupled together we have functions totally above and beyond that of man-made yet highly sophisticated engineered machines.
The ultimate test comes down to building a computer circuit that enables and encompasses emotions and true consciousness. This has wholly been shown to be impossible.
Howard,
Feelings and emotions are a part of what defines our conscious state. If a computer cannot be made to have these attributes, then it can have no conscious. How we “feel” has a lot to do with how and what decisions we make. One would first have to make a robot “feel” before it could make decisions like humans. The AI guys say it can be done but I know they must be smokin something funny.
Rob wrote we are an “operator” that is housed in the machine Excellent if by “we” you mean our spirit I totally agree well said. But what is our spirit if it is not knowledge housed in fine matter? Is some sort of magic also implied?
I don’t think the goal is to make decisions like humans if by that you mean to include the full range of human emotion. Feelings are not truth they are simply a response to stimulus some are useful to the goals of a robot like pain or fear which might prevent damage or destruction but a robot wouldn’t need to feel mad, glad or sad unless it was expected to be say a nanny or nurse.
Consciousness does not depend on feelings, feelings depend on consciousness. My consciousness level is unrelated to my emotions isn’t yours?
Rob:
I also see evidence of an intelligent and purposeful creation. Its purpose is to allow the Creator to experience all of his own potentialities, and His intelligence is shown by His use of evolution as a highly effective way of doing so.
God loves evolution, IMO.
Our Spirit’s may indeed “operate” our physical bodies, but I think the relationship is that spirit is to person as mind is to neuron. I think the spirit is a collective property of many copies and variants of my physical body that have, do, and will exist throughout spacetime (and probably a lot of other spacetimes as well). That’s how they can exist before the earth was formed and why they can not be destroyed upon death.
Just as an ant’s brain is not complex enough to house much of a mind, a single human mind is not complex enough to house much of a spirit. God “built” our spirits out of an infinite number of physical copies of our minds. (The existence of the physical copies is a pretty robust prediction of modern cosmologies, and the “worlds without number” in the Book of Moses is really quite comfortable with this.)
Firetag,
Sounds too much like “Back to the Future” science fiction stuff, eh eh.
I believe it is possible at some level (probably spiritual) to look at the mechanical operations as to why we think and make decisions and thus have consciousness. But the minute we try to place math equations on the why and how we find ourselves in unchartable territory. It would basically mean that intelligence can all be formulated from many complex layers into a string of perfect mathematical operations. It would basically mean that intelligence could be written as a program. But this is impossible. We are dealing specifically with free will and how it needs to be properly defined ont he “thought” level. For instance- are all of my actions and feelings based entirely upon the exact measurement of past interactions or is there always a choice that can be made at any given junction and thus render all past interactions as noncomputable into figuring what move I will make next? In other words-
Why do I make the choices I make? Is it based entirely off of a multi-layered mechanical type process that computes all past information in a mathematical way or is it something far beyond what can be represented as math. This gets right to the heart of the matter- If we can show that my “consciousness” is nothing but the direct functions of mathematics and can represent all the properties in the form of math, then by George…we shall be the richest men on the planet!
But I can asure you, this stuff makes for great science fiction entertainment but loses all practicality when trying to actually do it.
Howard,
The minute we start exploring robotic consciousness, is the very minute we chart the course into the fantasy land of science fiction.
What are AI progarammers really after- a way to mimick human consciousness in robotics, or how to actually plot the course into actual and real consciousness that is housed in a robot? I think it is real in some aspects for a robot to mimick (trick us into thinking without it actually happening) consciousness and appear more human or alive, but the reality of it is the program running the robot- the AI engine, is just a very very complex program of “if this then that”. The more of those layers you could add would help trick us into believing it was real but in reality it is all the same over and over again. At no point of giving it more and more complex features and programming is it going to actually become “smart” and gain some type of actual “conscious”.
To the extent that humans act rationally it is relatively easy to write a formula. When they act irrationally are they irrationally irrational or rationally irrational? This is a question for psychology and becomes more difficult to answer when we’re dealing with abnormal psychology but I think it’s fair to say that within normal psychology they act rationally irrational which means a formula can be written. So how do we correct our irrational but normal psychological behavior by exercising our autonomy through our consciousness or by accepting outside information in the form of new to us knowledge.
FireTag 35 Great comment.
It doesn’t require that I be able to write the algorithm; it just requires that God be able to do it. Bruce has several posts in the archive that talk about this as I recall.
Firetag,
I seriously doubt that consciousness can be traced down to some algorithm, even if God himself were writing it. It all keeps coming back to if numbers, in the right sequence, can be formulated into true intelligence. But the funny part is that without a consciousness to begin with and to understand it, it means nothing by itself.
Howard,
I was in the store today and couldn’t decide which bacon I should buy- should I buy the more expensive brand, the thick cut, maple falvored, etc. I ended up picking a cheaper brand.
There were factors involved such as I was in a hurry, didn’t have a lot of extra cash, etc, but still wanted to enjoy my Sunday dinner tomorrow. Now of course these factors can be rationally and logically played out to some sort of general consensus and probably could have been narrowed down to a few choices based upon those rational factors.
But, and here is the big but, If I couldn’t choose between the few brands based upon the rational factors, why did I choose the one I did? I could answer that it finally came down to it just “feeling” right on the one I picked up rather randomly as I was handling the different packages. My “consciousness” had a large part into why I chose the one I did. I could even state to some degree that my feeling good about the one I chose caused me to have passion on that rather irrelevent choice, I mean really- it’s just bacon!
Now if we were to break this down on the spiritual particle level, what motions or processes were taking place for me to “feel” good about my random choice? And, is this a different process than we find in computer programs that work strictly with math operations?
Let me pose this question-
Do we suppose that we are only acting upon the outside inputs out of our control that effect us or do we actually have choice of our actions and the ability to control input and output actions?
Can a robot be programmed to act and control input and output information, even generating it’s own intelligent input and recognizing it was good based upon it’s output?
Rob I don’t think there would be any advantage to figuring out your formula for buying bacon that particular day why should the robot come to the same conclusion you did? One of the advantages of AI is better decisions can be made because they will be mostly emotion free. A better example would be crossing a busy street safely the test would be do you both make it to the other side okay not did you take the same path or same number of steps.
We do have choice of our actions but that varies with individual enlightenment. The answer to your second question would be yes that would be required to actually achieve AI.
In order to achieve true consciousness, the robot would have to make choices in the precise manner that I do. The robot would have to choose the bacon in the precise manner that I did. Is this entire post not about how a robot can achieve consciousness?
Also-
Suppose that one of the bacon packages I picked up had something wrong with it and my spirit gave me a bad feeling about that particular package. How do you program a robot with that state of conscious ability? Can a robot be programmed to have spiritual experiences? Is not spiritual experiences a part of our conscious abilities?
I disagree with your premise in 44 we are not cloning you mind we’re only attempting to create consciousness why should your emotional inefficiency be included in a robot beacon buying decision? The question you pose in 45 is an interesting one how do we receive inspiration? I believe it is received through our spirit at a low level of consciousness and precipitates upward this is why I believe the robot will in it’s final stages need some kind of a spirit equivalent.
But consciousness can be strongly argued to be just as much a spiritual experience as it is a physical experience.
Exactly what kind of spirit equivalent could be made? I am not sure there is any equivalence to our spirit.
Well we know that material things are modeled first in the spirit and it is made of fine matter it must house knowledge so I see the spirit as a wispier version of our physical bodies they must think in a similar manner and would therefore be programmable too.
Yeah but we are speaking of living tissue and substance rather than what is found in robots.
Yes we are. How does that make the programming fundamentally different?
We aren’t programmed- that’t the fundamental difference.
Robots only do what they are programmed to do. Humans do what they want to do- no programming involved.
Can you explain how we do what we want to do without logic and knowledge? Isn’t logic + knowledge programming?
Intelligence, autonomy and consciousness . . . I can think of some people who don’t meet this test.
I’m not sure what to make of the duality thing. I tend to think that even if consciousness “springs” up when our physical bodies are advanced enough, there is something indelible about it, an “extra” quality as you say. OTOH, that’s what Cornelius thought in Planet of the Apes.
When I think of “programming” I think of strict math operations that control exactly how a machine or computer functions. In humans I do not see where we compute and operate in this manner. We use logic and reasoning in ways far beyond the reach of math. For a robot to make a decision it can only do what it is exactly programmed to do so it really doesn’t make any decision, it only appears to. For a human to make a decision it takes into account many factors, judges them and then at that point it may go completely against all logic in it’s decision anyway. That is the human “conscious” factor. How would one write a program that could duplicate this in a robot? It is impossible.
We use logic and reasoning in ways far beyond the reach of math. No that’s an oxymoron you are thinking of simple common programing I’m talking about something more complex that’s probably never been done before but should possible through the layered architecture I spoke of. Your autonomous choices go through some kind of a logical decision tree think of it as a maze you reason your way through and performance improves with increased experience. For a human to make a decision it takes into account many factors, judges them and then at that point it may go completely against all logic in it’s decision anyway. So are you asserting that that humans are irrationally irrational or just emotional? Are you asserting that robots must clone this useless irrationally irrationality or emotionality to achieve consciousness?
Rob, you’ve gotten hung up on the wrong level of abstraction. You think that because programs are written with if-statements and while-loops they can never be more than an advanced calculator. That’s like saying that since there are only a handful of basic elements in the universe, things can only exist as dumb collections of molecules. It doesn’t make any sense.
You cannot demonstrate that your brain is not following a deterministic set of biological reactions by appealing to “feeling” or by asserting your freedom of choice of what toothpaste to buy at the store. Your feelings and your choices operate at the highest levels of abstraction, along with your consciousness. The actual fundamentals, the if-statements and while-loops, happen at the very lowest levels of abstraction. They are building blocks. Everything you see around you is made up of smaller, simpler parts which are themselves made up of even smaller, even simpler parts. It is the synergy of all of these parts that you perceive with your mind. It’s a wonderful thing, but it is not magic.
The point I am trying to make here is that our decisions- our consciousness, is not the sum parts of what could be written on even an infinite amount of levels mathematically. We are not programmed machines. We do not think in binary terms like computers work. Making raional or irational decisions is not something one could program into a computer. Even random generator programs in computers aren’t really random at all. Beforehand, the program knows exactly what program it is running all the time even if that program runs according to random input from another source. It is always- “if this then that” no more and no less.. There is no deviation in a computer program to run in that manner unless somehow short-circuits or programming errors arise which is not part of it’s normal functioning or purpose anyway.
Sure, it creates quite the enigma to how the mind processes information, but we can’t think of it the same way we run computer programming. Consciousness is not the result of strict “if this then that” operations like a computer. Asuredly there are parts of consciousness that work in the manner of “if this then that”, but the entire process cannot work in this fashion. If it did then we truly wouldn’t be free agents, we would only be mere puppets for how events unfolded to us out of our control. That goes against every principle of the gospel.
Our unique ability to make decisions, have self awareness, consciousness, etc, goes without definition. All we know is that computer programs, no matter how sophisticated, cannot duplicate what we can do and process in the unique way we do as part of our spiritual and physical makeup. Trying to write down why we make a decision mathematically is useless in large part because we do not process inforamtion in mathmatical terms in entirety.
If one could program a computer to process information outside of the binary code of math then one could be at least on a better road that may lead to understanding human thought. We are just not going to get there though in these strict computer languages of strict math, our minds and spirits just don’t operate in that manner.
Without life, all that would exist in the entire universe would be just dumb collections of molecules. Somehow, there is an intelligent source involved and some thing have it and most don’t. God has it and created other forms of life to have this intelligence. One must possess this intelligence to break the mold of being dumb. Robots, calculators, etc will always be just dumb collections of molecules that can’t think or act for themselves.
You cannot tell from the simple fact that you perceive yourself as a free agent that your brain isn’t functioning deterministically at the very lowest levels. To assert this is begging the question.
Oh I’m not arguing with you there. There are a lot of things out there that go against the principles of the gospel. Fascinating, isn’t it? Even evolution goes against the principles of the gospel, as you have so aptly observed.
“if this then that” is not the basis of computer logic it actually a higher level logic than the binary system “1”, “not 1” of computer hardware. “If/then” logic allows the construction of logic gates such as not, and, nor, or etc. covering all the possible choices involved in decision making. You probably don’t realize all the logic gates your personal decision making goes through because much of it runs in the background of your subconscious.
I believe in God and that I have a spirit I recognize the difference between rock, computer, plant and animal. But isn’t there also a difference between a mouse and a man? What is that difference?
The rock contains no logic, the plant contains more logic than the rock, the mouse contains more logic than the plant and the man contains more logic than the mouse.
How much logic shall we give to the robot computer and where will it fall in this continuum?
Carson,
To state that my decisions are based deterministically is like stating that I have no control of my actions. So, according to that logic I might as well just do exactly what my physical body wants and committ all manners of sin because after all- I am just programmed that way and have no real control, only perceived control.
That is bad logic.
I personally believe that every choice we consciously make is not pre-determined to that point. This is different to how a computer program works because it already knows exactly what path it will make at the moment before that individual path is taken. If one were to print a copy of the program at any given moment and trace all of the math of that operation, it would all add up perfectly as to why it chose the path it did. Our conscious state cannot operate in that manner. Now I am not suggesting any alternate theories for why or how this works, but I am basing this off the logic of how gancy works and we have it, and because we have it, our minds can’t operate in deterministic manners.
Howard,
I will agree that the mind on many levels functions as a logic gate. That is not my argument. My argument is at the actual moment of decision, the moment of feeling, the moment of awareness, etc. I do not believe our mind uses any logic gate at that juncture. It uses something else. What? I am not sure.
…moment of decision, the moment of feeling, the moment of awareness… Rob it sounds like you are trying to describe the experience of consciousness here do you see these as existing simultaneously, sequentially or separately?
I am merely suggesting that these moments are arrived through non-mathmatical processes like that of a computer program..
Re ChildeJake
I’m so glad you brought this up! I plan to write on this in the last post of this series – I think it’s a very important question. As I think about the implications for conscious robots, I cannot help but think that perhaps we shouldn’t develop them at all. Not sure on that though yet.
Robots with rights? That ought to be novel, I await the post! This thought experiment into science fiction is intriguing me.
Rob:
See the movie “Bicentennial Man”.
Re Rob
I’m sorry I’m late to the discussion at hand. I didn’t think comments were still being made here.
Might I humbly suggest that you learn a little more about computation, physical processes, uncertainty, neuroscience, etc. before making such sharp conclusions?
Interacting with the real world is precisely why a robot might develop consciousness. The world is unpredictable, noisy, and prone to error. The logic mechanisms used to compute any action (output) come from operating on noisy, error prone, uncertain sensors (inputs).
What your argument boils down to is that you believe humans have free will, and are not deterministic and therefore we cannot make a conscious robot because robots are deterministic. But the two don’t have to be mutually exclusive as the compatibilists will tell you. Humans absolutely make some decisions deterministically (by all appearances). The free-will vs. determinism debate has raged for millenia.
I also am not convinced that humans are totally deterministic either. But I don’t think this excludes the possibility of conscious robots. Keep in mind that no one believes that we will create a robot, push the “start” button and that it will have consciousness. Much like humans, most people (scientists) believe we will create robots capable of developing consciousness.
The fact that you cannot articulate exactly how the interactions happen that cause you to make a certain decision is precisely why your arguments aren’t convincing. You speak in generalities, and make statements like “it’s obvious that…” without stating precisely why it’s so obvious.
Re Rob again-
To belabor the point just a little, how exactly do you know that you have free-will at all? And further, how do you know that a machine is deterministic?
Suppose I tell you that just because you program a machine to operate on input according to mathematical equations doesn’t mean it will do that (which is very very true). There are complex reasons why machines do not always behave in the way they are programmed. One example is “upsets” in the transistors that govern the binary operation of a computer. Sometimes things happen (noise, error, radiation, power leakage, etc.) in a transistor that cause it’s voltage levels to flip, thereby flipping a single bit in the binary computation. This results in the computer NOT acting according to the rules and mathematics by which it was programmed.
Another example is floating point error. Floating point operations within a computer are an approximation of the actual decimal numbers they represent. They are only accurate to within a certain precision. For example, 0.1 + 0.1 = 0.2 in decimal mathematics. But in floating point operations 0.1 + 0.2 != (does not equal) 0.2, EVER!! Why? Because 0.1 is impossible to represent as an IEEE floating point number.
The point is, although computers appear to be deterministic, they aren’t always. Like humans, they operate in an uncertain world, with uncertain consequences. Is this how they will develop consciousness? I don’t know, but I am fairly confident that at some point they will.
In a later post, I’ll discuss some of the technical requirements needed to have the computational power to even get on the road to consciousness.
jmb275,
One doesn’t have to be a computer nerd/engineer to understand and realize how the basic workings of a computer, machine, etc is completely different than how humans work.
The fact that you can’t articulate exactly how the interactions might happen to create consciousness in robots make your arguments wholly unconvincing. Somehow you think that just making computers more complex will make them develop conscious abilities is ridiculous.
I would beg to differ that most people think we will develop robots with consciouses. I think there is a small minority that thinks it is possible but for the most part, the majority of people know and realize that artificial intelligence will always be just that- artificial.
I don’t know waht religion you are but in LDS doctrine one has to believe in free will otherwise there is no existance. Making decisions for us means being able to freely choose between several options and having that decision not based in it’s entirety off of past events.
I find it interesting that theories delving into robotic consciousness remain as abstracts just on paper and nothing more. There is no science whatsoever proving any level of consciousness in a machine. Every knowledgeable person the planet knows that only living things possess true intelligence and conscious states. No machine, computer or non-living device ever invented or manufactured has ever developed any level of true intelligence or level of consciousness.
All your theories are stuck in science-fiction land and will continue to remain there until you can show how humans intelligence really works on some mathmatical or machine level. So it appears you are as guilty as I in explaining how human consciousness works.
Re Rob-
Fair enough. Thanks for setting me straight.
jmb275,
You asked-
“To belabor the point just a little, how exactly do you know that you have free-will at all? And further, how do you know that a machine is deterministic?”
I think it needs to be understood a little defferently because fromt he angle you propose we could debate all the day long if I really even exist at all- hum, am I even typing on a keyboard right now, etc.. We need to ask it better like this-
We can define free will as the testable function that an agent can choose and act according to ones desires when confronted with multiple options/decisions that can be made which lead to different conclusions.
Humans excersize free will, we constantly make choices in a real world environment.
Computers will do only that which they are exactly programmed to do. If a machine or computer makes an error it is only because of programming error or malfunction with an exact cause. This is not how humans work.
I work with mechanical machines all day long at my job. Sometimes they don’t function right. At those times I am called upon to fix them. Without any doubts whatsoever I can determine there is an exact cause to the malfunction and it will always be precisely logical and evident of it’s malfunction according to every law of science and physics.
Re Rob-
I see. Sounds like you have it all figured out.
jmb275,
Huh? I am as clueless as you!
What I do know from a lot of study is that humans are not machines. We don’t act like machines, we don’t break like machines, we don’t look like machines, etc. So, to make a machine human-like is rather impossible. It absolutely cannot be done. Men have been making machines since the beginning of time and have never made one capable of consciousness. It just can’t be done. The laws of science, physics, etc just do not allow this to happen. Why? I am not sure other than the fact it is a principle and law that exists. there is no bending or warping that law. Creating intelligence is impossible. It either exists already or it never can exist. The scriptures even state that intelliegence cannot be created- it has always existed. non-living things are thus not capable of becoming or gaining intelligence. The only possibility would be if somehow God placed an intelligent spirit in the machine, of the which I doubt he woudl or even could do.
Re Rob-
Typically people who are clueless don’t take such hardline positions on issues on which they are not an expert.
You’ve declared emphatically your conclusions using very absolute terms like “impossible,” “laws of physics,” “cannot,” “every knowledgable person knows,” “the only possibility,” etc. etc. To me this implies that you are an expert. You likely have spent years studying consciousness, artificial intelligence, computation, systems, robotics, etc. and have an advanced degree to be able to speak so certainly.
I commend you for your diligence on such an intriguing topic.
I’ve no doubt Rob is also an expert in evolution, too.
jmb275,
I am only clueless as to the specifics of how consciousness works- how we feel and reason, etc. But then again- so is all the world. No one knows that.
Please show me a principle or law used in computer engineering that allows for formulating the human experience and I think we can all agree. But you must understand as I do that there is no magic math formula that encodes for intelliegence let alone consciousness.
Re Rob
No thanks. I can see you have it figured out and have your mind made up. I think it’s great you’re able to be so certain about so many things.
Where the rubber meets the road is where it matters. Let’s be honest about what we are truly able to achieve, or try to achieve.
I think it wholly possible to give robots the perception of consciounsness, perhaps even mimick it in some great degree. But when it comes right down to it, have we really created true consciousness? The smartest computers we have in the world still only do specific tasks.
If we can’t figure out human consciousness, how can we possibly pattern a robot to have such? That’s like me getting in my car blindfolded and trying to drive from my home in Idaho to New York and seeing a Broadway play all the while obeying all the laws and getting all the directions right somehow.
I am curious however to see how one plans on creating this “consciousness” and placing it in a robot. I think it would be quite novel if it really happened or is possible. I just don’t see how one can take basic computer language and use it to formulate it into something that we have no idea how it works in the first place.