On the topic of conscious AI
-
AI and theoretical discussions about consciousness seem to be popular these days. I have a very clear-cut view of what conscious AI is, but I've noticed the news articles covering the topic are somewhat.. on the surface in their approach. Which got me wondering: what does the community here think? My opinion can be summarized into 2 key point: - there's no measurable distinction between natural and artificial - as a result, human consciousness is the primary example of what you would call conscious AI The main practical difference, is that human consciousness is running electrical signals on top of a amalgamation of est. 37.2 trillion cells; AI is running electrical signals on an non-reactive silicon substrate. The main functional difference, is that humans are trained in observation, to copy the behavior of other humans. Given the size of our data set (7 billion?) and the time it takes us to get good at complex tasks (21 years for the best of us, 30+ for others?) I feel like we're pretty quick to dismiss our software and CPU based counterparts as less capable. When I look outside, I see people smarter than my Roomba, sure. But the gap doesn't seem too huge.
Having been out of my body several times, and having spent time studying accommodating philosophies and meditating, thinking, and feeling, I've come to the conclusion that my brain is a tool of my consciousness, but it is not me. Having said that, everything has at least a rudimentary consciousness, since everything is a product of consciousness that cannot be separated from its maker, so it might be possible for an AI to exhibit some consciousness, and our brains themselves to have some consciousness. But we are each more than that.
-
To Be Fair... Have you defined Intelligence? I studied AI at the University... If something is or appears Intelligent... Does it matter? And what is Intelligence without Empathy? (Dangerous, choose: Dalek, or psychopaths) What makes use HUMAN is the BioChemistry + Intelligence + Emotions/Empathy. The BioChemistry is why we get bored and STOP working on things. We have to eat, sleep. Our abilities change dynamically based on this. So, we have to have an operating system on top of an operating system. For the record, I was considering doing my Masters on defining Intelligence as "self-organizing hierarchical pattern recognition" And if you think of Mensa and IQ tests... The toughest questions are the most complex patterns to identify, and those are the answers that increase your scores. I fear machine intelligence because if it does not understand Empathy/pain/suffering... Many things are easy. (Kill the homeless, feed them to the poor as free food. Test every child early, those not capable of expanding the race are terminated). Those are OBVIOUS solutions... And HORRIBLY WRONG for a HUMAN. The seem almost like the Communist decisions!
I am not biochemical. I am a spirit inhabiting a biochemical machine.
-
AI and theoretical discussions about consciousness seem to be popular these days. I have a very clear-cut view of what conscious AI is, but I've noticed the news articles covering the topic are somewhat.. on the surface in their approach. Which got me wondering: what does the community here think? My opinion can be summarized into 2 key point: - there's no measurable distinction between natural and artificial - as a result, human consciousness is the primary example of what you would call conscious AI The main practical difference, is that human consciousness is running electrical signals on top of a amalgamation of est. 37.2 trillion cells; AI is running electrical signals on an non-reactive silicon substrate. The main functional difference, is that humans are trained in observation, to copy the behavior of other humans. Given the size of our data set (7 billion?) and the time it takes us to get good at complex tasks (21 years for the best of us, 30+ for others?) I feel like we're pretty quick to dismiss our software and CPU based counterparts as less capable. When I look outside, I see people smarter than my Roomba, sure. But the gap doesn't seem too huge.
KBZX5000 wrote:
there's no measurable distinction between natural and artificial
There are significant differences. First, here is the physical difference of the machines involved. Second, there is no proof that all of human consciousness is limited to its physical constraints, meaning no AI would be able to replicate any part of human consciousness that is not physical. Third, AI is incapable of maturing and growing on its own resulting in creativity. An AI, if asked which flower is prettier, can answer as its human minders have taught it in its training data. What AI cannot do is tell you why all on its own. And that, even a young human child can do.
-
I am not biochemical. I am a spirit inhabiting a biochemical machine.
Prove the spirit exists!
-
Prove the spirit exists!
Prove that consciousness exists!
-
Prove that consciousness exists!
Well we have 2 choices here. Either it does, or it does not exist. If we see a person completely knocked out, we refer to them as unconscious. Therefore, we accept unconsciousness as a state of being. And the opposite of that, being consciousness. QED. Now, we can argue degrees of consciousness. But that REQUIRES the acceptance of consciousness. BTW, it exists because it is a "state" of being. It can be proven. Spirit, on the other hand, is something we "Possess". But it takes up no space, cannot be weighed or measured in the physical world. Therefore, it is simply a LABEL on part of us that we FEEL connects to other people, and not in ways we can completely explain, and many don't understand, and yet some say they don't have and it is a hallucination. Finally, as a word, it can have Positive or Negative meaning. Positive: He really brings a lot of Spirit to the game. Negative: I had to Fly Spirit! :-)
-
AI and theoretical discussions about consciousness seem to be popular these days. I have a very clear-cut view of what conscious AI is, but I've noticed the news articles covering the topic are somewhat.. on the surface in their approach. Which got me wondering: what does the community here think? My opinion can be summarized into 2 key point: - there's no measurable distinction between natural and artificial - as a result, human consciousness is the primary example of what you would call conscious AI The main practical difference, is that human consciousness is running electrical signals on top of a amalgamation of est. 37.2 trillion cells; AI is running electrical signals on an non-reactive silicon substrate. The main functional difference, is that humans are trained in observation, to copy the behavior of other humans. Given the size of our data set (7 billion?) and the time it takes us to get good at complex tasks (21 years for the best of us, 30+ for others?) I feel like we're pretty quick to dismiss our software and CPU based counterparts as less capable. When I look outside, I see people smarter than my Roomba, sure. But the gap doesn't seem too huge.
I am fascinated by AI, and perhaps more so by those who endeavor to create it. Most use the standard of "complete human intelligence" as the bar we need to cross. I would propose we set that bar much lower to actually achieve some success instead of just hype. What about a dog? Dogs are not tremendously intelligent; however, they can learn behaviors AND they can make choices, so they DO possess some intelligence. Granted, some choices such as chasing cars, attacking a bigger dog, etc., may not be "wise" choices, BUT I know of no robotic dog that can decide whether it prefers kibble or moist food. Some progress toward this level of decision making is being accomplished, but such progress is very slow and nowhere near the projections of anyone. Nonetheless, once we achieve "dog", maybe work our way up to "primate". "Human" is still a long way away. Nature took millennia to increase intelligence in humans to the current state. I think it tremendously arrogant to think we can achieve those same results in a few decades. "Complex tasks" can take many forms, but are constantly being learned throughout lives: crawling, walking, talking, using tools, MAKING tools. The biggie that humans seem to excel, and other animals fail miserably, is recognizing time, alternate possibilities for the future, and deciding accordingly.
-
I am fascinated by AI, and perhaps more so by those who endeavor to create it. Most use the standard of "complete human intelligence" as the bar we need to cross. I would propose we set that bar much lower to actually achieve some success instead of just hype. What about a dog? Dogs are not tremendously intelligent; however, they can learn behaviors AND they can make choices, so they DO possess some intelligence. Granted, some choices such as chasing cars, attacking a bigger dog, etc., may not be "wise" choices, BUT I know of no robotic dog that can decide whether it prefers kibble or moist food. Some progress toward this level of decision making is being accomplished, but such progress is very slow and nowhere near the projections of anyone. Nonetheless, once we achieve "dog", maybe work our way up to "primate". "Human" is still a long way away. Nature took millennia to increase intelligence in humans to the current state. I think it tremendously arrogant to think we can achieve those same results in a few decades. "Complex tasks" can take many forms, but are constantly being learned throughout lives: crawling, walking, talking, using tools, MAKING tools. The biggie that humans seem to excel, and other animals fail miserably, is recognizing time, alternate possibilities for the future, and deciding accordingly.
A while back, there was an article about OpenWorm, which takes a similar point of view on the subject. I believe the problem with using "natural emerging intelligence" as a classifier, is that we are inherently biased. We assume there's something special about our brain, because the idea makes us feel good. Evolution, to me, is a layman's way of saying that our cellular composition is dynamic, rather than static. I'd argue that your brain is just one of the tools your cells rely on to avoid dying en masse on a cellular level. I'd also argue that we have not really evolved a lot in the past 5000 years; instead, the concepts we share have evolved tremendously. At the end of the day, our brain is only a set of neurons that can record and replay chemical patterns. The concepts we share are broken down into words, which break down into chemical impulses, which we imperfectly repeat at some point further in time. We rely on software when we need a sequence of functions that can be replayed perfectly. If you take a step back, it seems like we place a lot of value in the act of storing and replaying patterns. My hypothesis is that intelligence is an emergent property from the patterns we share, not the mechanism that stores it.
-
KBZX5000 wrote:
there's no measurable distinction between natural and artificial
There are significant differences. First, here is the physical difference of the machines involved. Second, there is no proof that all of human consciousness is limited to its physical constraints, meaning no AI would be able to replicate any part of human consciousness that is not physical. Third, AI is incapable of maturing and growing on its own resulting in creativity. An AI, if asked which flower is prettier, can answer as its human minders have taught it in its training data. What AI cannot do is tell you why all on its own. And that, even a young human child can do.
First point: I don't believe we are actively making conscious decisions. We create iPhones because it's in our nature, similar to how a tree will grow fruit. Regardless of how that makes either of us feel, I don't know of any known method to measure the difference between the two. Second point: I don't believe human consciousness is tied to the physical constraints of the human body. I believe consciousness emerges when ideas are successfully shared between 2 people. I do not believe there's anything inherently special about a person, but I do think we have strong feelings about the concept of being special. Do you trust your gut feeling, or do you search for cold hard facts? I choose the latter, because it makes me feel better, ironically. Third point: When I asked a newborn which flower is prettiest, it failed to form any kind of sensible response. When I asked google, it showed me pictures of pretty flowers.
-
I question your assertion,
Quote:
People can only create copies of the things they know
There has to be creativity somewhere, or we'd still be living in caves with a complete absence of tools. Often our creativity manifests in small improvements to existing structures, but recognizing the need for improvement and designing it, I would assert, are inherently creative tasks.
Outside of a dog, a book is a man's best friend; inside of a dog, it's too dark to read. -- Groucho Marx
-
To Be Fair... Have you defined Intelligence? I studied AI at the University... If something is or appears Intelligent... Does it matter? And what is Intelligence without Empathy? (Dangerous, choose: Dalek, or psychopaths) What makes use HUMAN is the BioChemistry + Intelligence + Emotions/Empathy. The BioChemistry is why we get bored and STOP working on things. We have to eat, sleep. Our abilities change dynamically based on this. So, we have to have an operating system on top of an operating system. For the record, I was considering doing my Masters on defining Intelligence as "self-organizing hierarchical pattern recognition" And if you think of Mensa and IQ tests... The toughest questions are the most complex patterns to identify, and those are the answers that increase your scores. I fear machine intelligence because if it does not understand Empathy/pain/suffering... Many things are easy. (Kill the homeless, feed them to the poor as free food. Test every child early, those not capable of expanding the race are terminated). Those are OBVIOUS solutions... And HORRIBLY WRONG for a HUMAN. The seem almost like the Communist decisions!
I define intelligence as: - an emergent behavior that occurs when a group of self-sustaining pattern engines successfully exchange ideas over an extended period of time. Feelings are tools; they short-circuit our thought process with previously established follow-up actions. It saves time and stops our neurons from getting overly exerted, mostly, but as a side effect it also makes our thought process more rigid. Fun side note: I really hate Mensa. They kept stalking me for years, trying to sucker me into joining their retarded little club. They tell me I'm smart and yet they treat me like an idiot. I'm not paying anyone who wastes my time.
-
Having been out of my body several times, and having spent time studying accommodating philosophies and meditating, thinking, and feeling, I've come to the conclusion that my brain is a tool of my consciousness, but it is not me. Having said that, everything has at least a rudimentary consciousness, since everything is a product of consciousness that cannot be separated from its maker, so it might be possible for an AI to exhibit some consciousness, and our brains themselves to have some consciousness. But we are each more than that.
Have you noticed that during meditation, you're mostly suppressing your though loop and enforcing minor sensory deprivation, until the residual brain activity becomes the primary thought loop? I honestly don't recommend it. I recommend segmentation: use a completely different set of skills, thoughts and feelings for a part of the day. That way, the other neurons can get some R&R, and you're emotions have more trigger moments, which helps to keep the chemical balance. Thought cycles have an intrinsic value. If I did nothing with them, it would feel like a waste.
-
First point: I don't believe we are actively making conscious decisions. We create iPhones because it's in our nature, similar to how a tree will grow fruit. Regardless of how that makes either of us feel, I don't know of any known method to measure the difference between the two. Second point: I don't believe human consciousness is tied to the physical constraints of the human body. I believe consciousness emerges when ideas are successfully shared between 2 people. I do not believe there's anything inherently special about a person, but I do think we have strong feelings about the concept of being special. Do you trust your gut feeling, or do you search for cold hard facts? I choose the latter, because it makes me feel better, ironically. Third point: When I asked a newborn which flower is prettiest, it failed to form any kind of sensible response. When I asked google, it showed me pictures of pretty flowers.
Fortunately, neither I nor any other adult is governed by what you believe. Yes, I make conscious decisions all the time, not counting the basic functions of the autonomic nervous system. I endeavor to use reason as much as humanly possible, so that reason governs and directs my emotion-based feelings. You asked Google to show pretty flowers. Google showed you what it’s human data sources determined were at least flowers, and at most pretty to them. Not a good test. Ask a 5 year old girl WHY she thinks the flower prettiest to her is pretty, and she can tell you. Ask Google (or Bing) to pick its one prettiest flower, and if it can do that, ask it why is it the prettiest. The response will make clear why AI is not, nor ever will be, able to attain its own independent consciousness. Science fiction is all fun and games until one starts thinking, without reason, it is real.
-
Have you noticed that during meditation, you're mostly suppressing your though loop and enforcing minor sensory deprivation, until the residual brain activity becomes the primary thought loop? I honestly don't recommend it. I recommend segmentation: use a completely different set of skills, thoughts and feelings for a part of the day. That way, the other neurons can get some R&R, and you're emotions have more trigger moments, which helps to keep the chemical balance. Thought cycles have an intrinsic value. If I did nothing with them, it would feel like a waste.
I get your point and engage in a variety of activities during the day. On the other hand, meditation has enabled me to do and sense things of a psychic nature that the monkey mind usually blocks.
-
I define intelligence as: - an emergent behavior that occurs when a group of self-sustaining pattern engines successfully exchange ideas over an extended period of time. Feelings are tools; they short-circuit our thought process with previously established follow-up actions. It saves time and stops our neurons from getting overly exerted, mostly, but as a side effect it also makes our thought process more rigid. Fun side note: I really hate Mensa. They kept stalking me for years, trying to sucker me into joining their retarded little club. They tell me I'm smart and yet they treat me like an idiot. I'm not paying anyone who wastes my time.
From your definition... There is no intelligence without behavior? The wise old monk isn't intelligent unless he behaves? And is it behavior that we MEASURE to score Intelligence? Your definition is close to what I imagine, but behavior throws me as well as no indication of the depth involved to indicate a level of intelligence. The more complicated a pattern you can easily recognize, the higher your intelligence, as is the higher your ability to predict the outcome of a set of events. (Just watch FAIL videos to prove the inverse!)
-
Fortunately, neither I nor any other adult is governed by what you believe. Yes, I make conscious decisions all the time, not counting the basic functions of the autonomic nervous system. I endeavor to use reason as much as humanly possible, so that reason governs and directs my emotion-based feelings. You asked Google to show pretty flowers. Google showed you what it’s human data sources determined were at least flowers, and at most pretty to them. Not a good test. Ask a 5 year old girl WHY she thinks the flower prettiest to her is pretty, and she can tell you. Ask Google (or Bing) to pick its one prettiest flower, and if it can do that, ask it why is it the prettiest. The response will make clear why AI is not, nor ever will be, able to attain its own independent consciousness. Science fiction is all fun and games until one starts thinking, without reason, it is real.
Quote:
Ask a 5 year old girl WHY she thinks the flower prettiest to her is pretty, and she can tell you.
Have you ever done this? Actual ask a child "why"? When I ask a 5 year old girl why she likes something, the answer is a stunted and confused mess, which means she's making up the answer on the spot. Fibbing is a nifty function of our brain, which generates most of our "on the fly" thinking. 5 year olds still suck at it, so it's easy to spot while it's happening. But, that's not consciousness, that's just real-time retrieval of deep-stored memory, mashed through a lexicon and syntax parser.
-
From your definition... There is no intelligence without behavior? The wise old monk isn't intelligent unless he behaves? And is it behavior that we MEASURE to score Intelligence? Your definition is close to what I imagine, but behavior throws me as well as no indication of the depth involved to indicate a level of intelligence. The more complicated a pattern you can easily recognize, the higher your intelligence, as is the higher your ability to predict the outcome of a set of events. (Just watch FAIL videos to prove the inverse!)
You can only recognize patterns if you have been exposed to them. This means you always need a precursor before intelligence can be established. I believe the exchange of ideas is the necessary initial precursor If the wise old monk has never moved, he can't be wise. He needs to exchange ideas before he can become wise. Once he has intelligence, inaction will diminish it. We currently score intelligence by testing the adoption and retention rate of measurable patterns. Someone who enjoys looking for patterns is considered smart. Someone who does not enjoy looking for patterns is considered dumb. So wait, does that mean I can become super smart, just reading / researching patterns? Yes. Exactly that. By any conventional measurement standard we possess, that's the thing that makes you smart. The first problem with that is that most of our patterns are assumed to be self-evident so they're never really written down in one place. Second problem is that people jealously guard the patterns they know because it gives them a measurement of power. You could call them trade secrets, but to me that seems to imply the information is somehow complex, which it often isn't.
-
You can only recognize patterns if you have been exposed to them. This means you always need a precursor before intelligence can be established. I believe the exchange of ideas is the necessary initial precursor If the wise old monk has never moved, he can't be wise. He needs to exchange ideas before he can become wise. Once he has intelligence, inaction will diminish it. We currently score intelligence by testing the adoption and retention rate of measurable patterns. Someone who enjoys looking for patterns is considered smart. Someone who does not enjoy looking for patterns is considered dumb. So wait, does that mean I can become super smart, just reading / researching patterns? Yes. Exactly that. By any conventional measurement standard we possess, that's the thing that makes you smart. The first problem with that is that most of our patterns are assumed to be self-evident so they're never really written down in one place. Second problem is that people jealously guard the patterns they know because it gives them a measurement of power. You could call them trade secrets, but to me that seems to imply the information is somehow complex, which it often isn't.
I completely disagree with you that you can only recognize patterns you have been exposed to. The magic of pattern recognition is that Numerology is a perfect example. We can actually make up INSANE patterns and fit things to them... Also, you can study patterns in general, but below a certain IQ point, I don't think you can see them. In fact, in children, one of the key markers for IQ is not pattern recognition, but the desire to keep trying. My daughter was given an incredibly long word to try to pronounce (some chemical name, she was 7, she started reading 4), and the doctor said she outlasted everyone her age that he had tested. He had to STOP her from trying to get it right. That "Search" for the correct answer is a search for a discernible pattern you have not seen before. While I agree if you are the puzzle maker, solving the puzzle is easy... But there is a certain amount of magic to creating problems/puzzles. Take a LOW IQ individual and have them create a pattern problem for someone else to solve. You will quickly find a "Child-like" knock-knock joke formulation. (Those jokes, BTW, are a standard setup and pattern). We are drawn to such patterns. Exchanging them does NOT create intelligence, and NOT exchanging them does not DIMINISH intelligence. While I would argue that communicating complex intellectual ideas to others can CLEARLY helped develop the reasoning behind the ideas... Writing them down in a book is a future potential exchange. If the book is burned first, the OPPORTUNITY is lost, but the IQ of the person who wrote the book is NOT diminished. For the same reason. The Old Wise guy who sees the truth and beauty of the universe, and all of its inherent patterns is just as intelligent with or without ever talking to another human being. The point about this wise old person is "Who Cares?" Because their lack of interaction impacts nobody but "others", not themselves. they were fine with their life. The protection of knowledge... unfortunately as we advance as a civilization, knowledge becomes the truest medium of power and value. I, for example, just learned a TON about eating right and fasting that flies in the face of EVERYTHING we were told since the 1970s. yet in 8 weeks, I have lost 35lbs, and EVERY Blood Marker has improved. My risk of cancer, heart attack, and atherosclerosis are all markedly improved. Yep, we guard knowledge. And yep, we are often misled to the profit of others! But even those become patterns, and when you see them enough, you lea
-
I completely disagree with you that you can only recognize patterns you have been exposed to. The magic of pattern recognition is that Numerology is a perfect example. We can actually make up INSANE patterns and fit things to them... Also, you can study patterns in general, but below a certain IQ point, I don't think you can see them. In fact, in children, one of the key markers for IQ is not pattern recognition, but the desire to keep trying. My daughter was given an incredibly long word to try to pronounce (some chemical name, she was 7, she started reading 4), and the doctor said she outlasted everyone her age that he had tested. He had to STOP her from trying to get it right. That "Search" for the correct answer is a search for a discernible pattern you have not seen before. While I agree if you are the puzzle maker, solving the puzzle is easy... But there is a certain amount of magic to creating problems/puzzles. Take a LOW IQ individual and have them create a pattern problem for someone else to solve. You will quickly find a "Child-like" knock-knock joke formulation. (Those jokes, BTW, are a standard setup and pattern). We are drawn to such patterns. Exchanging them does NOT create intelligence, and NOT exchanging them does not DIMINISH intelligence. While I would argue that communicating complex intellectual ideas to others can CLEARLY helped develop the reasoning behind the ideas... Writing them down in a book is a future potential exchange. If the book is burned first, the OPPORTUNITY is lost, but the IQ of the person who wrote the book is NOT diminished. For the same reason. The Old Wise guy who sees the truth and beauty of the universe, and all of its inherent patterns is just as intelligent with or without ever talking to another human being. The point about this wise old person is "Who Cares?" Because their lack of interaction impacts nobody but "others", not themselves. they were fine with their life. The protection of knowledge... unfortunately as we advance as a civilization, knowledge becomes the truest medium of power and value. I, for example, just learned a TON about eating right and fasting that flies in the face of EVERYTHING we were told since the 1970s. yet in 8 weeks, I have lost 35lbs, and EVERY Blood Marker has improved. My risk of cancer, heart attack, and atherosclerosis are all markedly improved. Yep, we guard knowledge. And yep, we are often misled to the profit of others! But even those become patterns, and when you see them enough, you lea
It's kinda cool that your daughter got scored on her desire to "keep at it". That's an interesting metric, tbh. As an anecdote, I can share you my experience with low-IQ people coming up with puzzles: They are derivative in form and they're often incomplete, with multiple fitting solutions, and no indication of which solution will be considered "correct". A good riddle is therefor a complex pattern: it follows a structured set-up, has a clue to identify the correct answer, and has an elimination factor to exclude wrong results. By simply reading those properties, you have inevitably gained intelligence, because you automatically mesh your notion of a riddle with the properties I present. You might accept them or reject them, but you're bound by the conclusions you draw, recalling them partially the next time you have to come up with a riddle. This is the acquisition of intelligence. You digest, analyze, consolidate, repeat, forget the details. The higher your intelligence, the more patterns you can combine and reproduce with a measure of success, potentially opening up the way to more interactions and more patterns. In contrast, the more you revert to immediate self-gratification, the less complex everything becomes, which generally reduces the variance of the interactions you'll have, and the more dumb you become. Smart people who suddenly decide to watch TV all day and never go out again, don't stay smart. We're not machines that suddenly stop working, but we do deteriorate gradually over time.