mother nature has chosen us on this planet time and time again. But yet we still run from her towards technologys that will more and more likely turn us into something non biochemical. Will the human race end up as borg like entitys 1/2 or less mechania or even totally robotic? Or will mother nature forced or other wise brought about changes keep us fully biological beings? Does it even make a difference? Is their a difference? The physical understanding of the world is bringing about technological changes that will change our race forever how would you like to see it guide us into our voyage?
QUOTE (magpies+Jan 14 2008, 01:09 PM)
Will the human race end up as borg like entitys 1/2 or less mechania or even totally robotic?
Someone's been reading/watching too much sci - fi
:
The only "Mother Nature" guiding us is evolution. If we survive as non or only part biological beings, then we will, eventually, develop into entirely that, as the organic beings will die more frequently. If the opposite is true.... Well then, I guess we'll stay organic. But we are not entirely organic at the moment. We use hip replacements, titanium skull plating, pacemakers, and so much more! We are already beginning to turn away from being organic. And so far, using technology like this has only helped us. So, it probably is a good thing.
- Gehn
Someone's been reading/watching too much sci - fi
:The only "Mother Nature" guiding us is evolution. If we survive as non or only part biological beings, then we will, eventually, develop into entirely that, as the organic beings will die more frequently. If the opposite is true.... Well then, I guess we'll stay organic. But we are not entirely organic at the moment. We use hip replacements, titanium skull plating, pacemakers, and so much more! We are already beginning to turn away from being organic. And so far, using technology like this has only helped us. So, it probably is a good thing.
- Gehn
Though I hate to agree with that scoundrel Gehn, he may be right. If bio-mechanical beings have greater survival advantage than purely biological beings, then that is what will prevail.
In the (far?) future, it is also possible that artificial intelligence created by us may develop greater consciousness than us and take over from us. This would be nothing but the next step in evolution. We would be their monkeys.
In the (far?) future, it is also possible that artificial intelligence created by us may develop greater consciousness than us and take over from us. This would be nothing but the next step in evolution. We would be their monkeys.
QUOTE (vkamath+Jan 14 2008, 12:25 PM)
In the (far?) future, it is also possible that artificial intelligence created by us may develop greater consciousness than us and take over from us. This would be nothing but the next step in evolution. We would be their monkeys.
Is that necessarily a bad thing? I mean, if we develop an AI with an IQ in the range of 300 or so, and the ability to deal with millions or even billions of situations at the same time, each with the full force of that intellect, and then replace human governments with that, would it mean that we are enslaved, or would it mean that the standards such an AI was given for judging situations are so perfectly in line with what we want and need that we can then go on about our respective lives, without ever needing to bother ourselves with thoughts of government corruption, false convictions and an unfair system?
I tend to believe the latter, simply because I doubt any such AI would ever be put in a position to hold such power unless the entire nation/world/galaxy/species (or whatever level of government it occupies) was in agreement that the AI knew what was best in such situations, and because it's not as if turning the government into such an AI would necessarily mean the suspension of human rights. (In fact, I believe that such an AI would be programmed to place human rights first and foremost on it's list of principles to uphold.)
Not that I'm rooting for it, mind. I just don't think it's necessarily a bad thing.
Is that necessarily a bad thing? I mean, if we develop an AI with an IQ in the range of 300 or so, and the ability to deal with millions or even billions of situations at the same time, each with the full force of that intellect, and then replace human governments with that, would it mean that we are enslaved, or would it mean that the standards such an AI was given for judging situations are so perfectly in line with what we want and need that we can then go on about our respective lives, without ever needing to bother ourselves with thoughts of government corruption, false convictions and an unfair system?
I tend to believe the latter, simply because I doubt any such AI would ever be put in a position to hold such power unless the entire nation/world/galaxy/species (or whatever level of government it occupies) was in agreement that the AI knew what was best in such situations, and because it's not as if turning the government into such an AI would necessarily mean the suspension of human rights. (In fact, I believe that such an AI would be programmed to place human rights first and foremost on it's list of principles to uphold.)
Not that I'm rooting for it, mind. I just don't think it's necessarily a bad thing.
QUOTE (vkamath+Jan 14 2008, 05:25 PM)
Though I hate to agree with that scoundrel Gehn, he may be right.
Arrrharhaharhar!! Me be a scoundrel and me be always right !
- Gehn
Arrrharhaharhar!! Me be a scoundrel and me be always right
!- Gehn
I am not sure I can agree that an IQ of 300 is all that is required to exercise sound judgement in dealing with human situations. As for government corruption, I think it more unlikely that high IQ would result in anything more than greater ingenuity in the levels and forms of corruption.
BDW:
Your observations are right on.
But I'd add tangentially that I don't see any reason to think that some AI that we would create would necessarily come to the conclusion that it should behave within some constraints that we defined when we created it (programmed or not). I'm assuming here that such an AI would at some point reach the point where it was self-aware. In that sort of situation I'd expect a self-aware truly intelligent genius consciousness would do whatever it felt was in it's own self-interest first. The trick might be to make it physically dependent on humanity for it's own survival which would create an imperative for it to treat us well. Of course being self-aware it could resent that limitation ... and on and on. Sci-fi has good lessons for us to be aware of in terms of creating AI (something that isn't necessary for us to become "cyborgs")
I think the merging of man and technology is the next likely evolutionary step and probably not a "bad" thing. It has started already in the medical world with joint replacement (even medicines could be viewed as technology merging with our bodies) and now we're seeing technology that can be placed directly into brains which allow for control of computers via thought. Obviously, this is pretty primitive stuff but at the current rate of technology (can it really keep doubling at it's current rate??) advances perhaps more close integration between our biological selves and our technological creations isn't so far away as we might think. Evolution is a stagnant process without input but can make huge leaps under the right circumstances. We need to be careful (I'm not sure I have "faith" that we actually will be though) moving forward but I think it might be inevitable.
Are we cyborgs if we have discreet "cell phones" implanted? If we are able to access information via such an implant like we do today from our desktops? Such things are being envisioned today. Not more than a few years ago I was working with a guy trying to get his cell phone modem (56k throughput) to operate properly and now broadband access and internet surfing is simple from many cell platforms. Promotheus gave us a lot to work with. I hope we don't burn ourselves but I also hope we don't become Luddites.
Your observations are right on.
But I'd add tangentially that I don't see any reason to think that some AI that we would create would necessarily come to the conclusion that it should behave within some constraints that we defined when we created it (programmed or not). I'm assuming here that such an AI would at some point reach the point where it was self-aware. In that sort of situation I'd expect a self-aware truly intelligent genius consciousness would do whatever it felt was in it's own self-interest first. The trick might be to make it physically dependent on humanity for it's own survival which would create an imperative for it to treat us well. Of course being self-aware it could resent that limitation ... and on and on. Sci-fi has good lessons for us to be aware of in terms of creating AI (something that isn't necessary for us to become "cyborgs")
I think the merging of man and technology is the next likely evolutionary step and probably not a "bad" thing. It has started already in the medical world with joint replacement (even medicines could be viewed as technology merging with our bodies) and now we're seeing technology that can be placed directly into brains which allow for control of computers via thought. Obviously, this is pretty primitive stuff but at the current rate of technology (can it really keep doubling at it's current rate??) advances perhaps more close integration between our biological selves and our technological creations isn't so far away as we might think. Evolution is a stagnant process without input but can make huge leaps under the right circumstances. We need to be careful (I'm not sure I have "faith" that we actually will be though) moving forward but I think it might be inevitable.
Are we cyborgs if we have discreet "cell phones" implanted? If we are able to access information via such an implant like we do today from our desktops? Such things are being envisioned today. Not more than a few years ago I was working with a guy trying to get his cell phone modem (56k throughput) to operate properly and now broadband access and internet surfing is simple from many cell platforms. Promotheus gave us a lot to work with. I hope we don't burn ourselves but I also hope we don't become Luddites.
QUOTE (Edward 3+Jan 14 2008, 01:21 PM)
I am not sure I can agree that an IQ of 300 is all that is required to exercise sound judgment in dealing with human situations. As for government corruption, I think it more unlikely that high IQ would result in anything more than greater ingenuity in the levels and forms of corruption.
Dude, we're talking about a computer. What are you going to bribe it with, PDA's in sexy little outfits?
>as any person can tell you people have a dark side. Now give that dark site access to great knowledge and lots of computation power and it may just out think its creators fast styles.
Add to that a couple of anarchist type hackers getting at the AI's program. Look how much trouble we have with spam, computer virus, hackers and the sort. you know that this would happen. it gets even worse if the AI programs it's self (off spring)
4) AI would make decisions without any emotional investment. Eg: If the coutry is attacked by terrorist, it would not lose its cool and go about invading other countries which have nothing to do with it.
When the US invaded Iraq, is wasn't an emotional decision. It was one which they hoped would profit them, because of the involvement with the oil industry. AIs might not make emotional reactions, but they might make ones which we would regard as cold and cruel.
- Gehn
Dude, we're talking about a computer. What are you going to bribe it with, PDA's in sexy little outfits?
QUOTE (uaafanblog+)
But I'd add tangentially that I don't see any reason to think that some AI that we would create would necessarily come to the conclusion that it should behave within some constraints that we defined when we created it (programmed or not). I'm assuming here that such an AI would at some point reach the point where it was self-aware. In that sort of situation I'd expect a self-aware truly intelligent genius consciousness would do whatever it felt was in it's own self-interest first. The trick might be to make it physically dependent on humanity for it's own survival which would create an imperative for it to treat us well.
What would prevent us from instituting a hard-wired version of Asimov's Three Laws of Robotics? In fact, I think it would be rather easy to institute such. I'm fairly certain that the technology and knowledge of programming already exists to implement such laws, although there's currently nothing that even remotely resembles software that needs them.
It'd be pretty simple, just build a motherboard with hardwired memory that contains the three laws, as well as a fourth "do nothing to remove these laws from your operating parameters." and toss in a few capacitors between the power supply and the board for good measure, and we're set. A hyper-intelligent AI that's incapable of putting it's own welfare above the welfare of humanity. On top of that, we could tie in it's survival to our own well being (by keeping it from having the ability to maintain itself, that way, it needs not only our survival, but our advanced knowledge of computing and electronics.)
Or we could use it to replace only parts of the government. It could replace judges, congressmen and the president, as well as heads of various departments, with a single human-only department that would serve as a check and balance for the AI's decisions. We could also give the AI some level of control (not complete, of course) over who occupies positions in this branch, thereby ensuring that neither the AI nor the head of the AI-checking institution has absolute power. Note that the AI would need nothing but it's core computer and human interface equipment, which would keep it from going all HAL 2000 or Skynet on us. If it starts demanding we start wars and give it control over robotic weaponry, we can always pull the plug, format and re-install.
What would prevent us from instituting a hard-wired version of Asimov's Three Laws of Robotics? In fact, I think it would be rather easy to institute such. I'm fairly certain that the technology and knowledge of programming already exists to implement such laws, although there's currently nothing that even remotely resembles software that needs them.
It'd be pretty simple, just build a motherboard with hardwired memory that contains the three laws, as well as a fourth "do nothing to remove these laws from your operating parameters." and toss in a few capacitors between the power supply and the board for good measure, and we're set. A hyper-intelligent AI that's incapable of putting it's own welfare above the welfare of humanity. On top of that, we could tie in it's survival to our own well being (by keeping it from having the ability to maintain itself, that way, it needs not only our survival, but our advanced knowledge of computing and electronics.)
Or we could use it to replace only parts of the government. It could replace judges, congressmen and the president, as well as heads of various departments, with a single human-only department that would serve as a check and balance for the AI's decisions. We could also give the AI some level of control (not complete, of course) over who occupies positions in this branch, thereby ensuring that neither the AI nor the head of the AI-checking institution has absolute power. Note that the AI would need nothing but it's core computer and human interface equipment, which would keep it from going all HAL 2000 or Skynet on us. If it starts demanding we start wars and give it control over robotic weaponry, we can always pull the plug, format and re-install.
QUOTE (BigDumbWeirdo+)
Is that necessarily a bad thing? I mean, if we develop an AI with an IQ in the range of 300 or so, and the ability to deal with millions or even billions of situations at the same time, each with the full force of that intellect, and then replace human governments with that, would it mean that we are enslaved, or would it mean that the standards such an AI was given for judging situations are so perfectly in line with what we want and need that we can then go on about our respective lives, without ever needing to bother ourselves with thoughts of government corruption, false convictions and an unfair system?
I tend to believe the latter, simply because I doubt any such AI would ever be put in a position to hold such power unless the entire nation/world/galaxy/species (or whatever level of government it occupies) was in agreement that the AI knew what was best in such situations, and because it's not as if turning the government into such an AI would necessarily mean the suspension of human rights. (In fact, I believe that such an AI would be programmed to place human rights first and foremost on it's list of principles to uphold.)
Not that I'm rooting for it, mind. I just don't think it's necessarily a bad thing.
It may Not be possible for a AI with higher intelligence to reach a evolutionarily stable state without learning self preservation or "selfishness". Also it is hard to imagine what purpose would humans serve with our lowly intelligence to AI beings of IQ around 300 or higher.
Yes, I agree that AI would never be purposely put to hold such power. But it may happen that we build increasingly "intelligent" AI, that we may push beyond a "tipping point" unknowingly. This may be analogous to how sometimes we cannot fully predict all the scenarios of operation in a complex computer program.
I would think, even if Asimov's laws are hardwired into AI, the evolutionary pressures and human errors could "tip the scale" in favor of the AI. Another factor to consider is that, intelligence of an organism is nothing but its ability to adapt to a given environment. Once AI has taken over the management of our environment, our intelligence may start to decline.
I tend to believe the latter, simply because I doubt any such AI would ever be put in a position to hold such power unless the entire nation/world/galaxy/species (or whatever level of government it occupies) was in agreement that the AI knew what was best in such situations, and because it's not as if turning the government into such an AI would necessarily mean the suspension of human rights. (In fact, I believe that such an AI would be programmed to place human rights first and foremost on it's list of principles to uphold.)
Not that I'm rooting for it, mind. I just don't think it's necessarily a bad thing.
It may Not be possible for a AI with higher intelligence to reach a evolutionarily stable state without learning self preservation or "selfishness". Also it is hard to imagine what purpose would humans serve with our lowly intelligence to AI beings of IQ around 300 or higher.
Yes, I agree that AI would never be purposely put to hold such power. But it may happen that we build increasingly "intelligent" AI, that we may push beyond a "tipping point" unknowingly. This may be analogous to how sometimes we cannot fully predict all the scenarios of operation in a complex computer program.
I would think, even if Asimov's laws are hardwired into AI, the evolutionary pressures and human errors could "tip the scale" in favor of the AI. Another factor to consider is that, intelligence of an organism is nothing but its ability to adapt to a given environment. Once AI has taken over the management of our environment, our intelligence may start to decline.
If you really think computers can think more moraly or apperciate art/life better then you can welcome to crazy town population you imo...
Inorder for something like AI to become self aware it will have to understand concepts like love and hate. And as any person can tell you people have a dark side. Now give that dark site access to great knowledge and lots of computation power and it may just out think its creators fast styles. The problem with a computer that thinks is that it can execute its commands faster then you can say oups im sorry take that back. Something that keeps humans from wiping ourselfs out is the fact that we have to wait along time for words to travel from our brains to our mouths so we have time to realize we just made a mistake. Computer AI may not have this benefit. Of course it probably wont do anything to harm itself terrilby but it could easly do something that could cause many ppl if not all ppl to die on earth in realitive time with no takebacks oups im sorry.
And thats if the machine is only slightly over taken by the dark side of the force... If it goes full out nuts why did you create me... OH I KNOW so I could be a slave GG.
Inorder for something like AI to become self aware it will have to understand concepts like love and hate. And as any person can tell you people have a dark side. Now give that dark site access to great knowledge and lots of computation power and it may just out think its creators fast styles. The problem with a computer that thinks is that it can execute its commands faster then you can say oups im sorry take that back. Something that keeps humans from wiping ourselfs out is the fact that we have to wait along time for words to travel from our brains to our mouths so we have time to realize we just made a mistake. Computer AI may not have this benefit. Of course it probably wont do anything to harm itself terrilby but it could easly do something that could cause many ppl if not all ppl to die on earth in realitive time with no takebacks oups im sorry.
And thats if the machine is only slightly over taken by the dark side of the force... If it goes full out nuts why did you create me... OH I KNOW so I could be a slave GG.
>as any person can tell you people have a dark side. Now give that dark site access to great knowledge and lots of computation power and it may just out think its creators fast styles.
Add to that a couple of anarchist type hackers getting at the AI's program. Look how much trouble we have with spam, computer virus, hackers and the sort. you know that this would happen. it gets even worse if the AI programs it's self (off spring)
AI spam So now your vacuum cleaner is going to escape to the mall to buy a Nokia phone, your fridge just ordered a bulk suppliy if viagra and your washing machine stops because it's too busy downloading porn 
So now your vacuum cleaner is going to escape to the mall to buy a Nokia phone, your fridge just ordered a bulk suppliy if viagra and your washing machine stops because it's too busy downloading porn
People want to build a nice benevolent little AI to run everything for us. Great, I can imagine what it would be like in a world run by the salvation army.
What we need in an overlord class AI is an understanding of what makes humans tick. Build its core operating system around the seven sins.
Now that will be an AI humans could relate to.
What we need in an overlord class AI is an understanding of what makes humans tick. Build its core operating system around the seven sins.
- Lust
- Gluttony
- Greed
- Sloth
- Wrath
- Envy
- Pride
Now that will be an AI humans could relate to.
This speculation is pure science fiction. We are light years from even beginning to understand consciousness, much less from reconstructing what took nature millions of years to perfect. It certainly won't be a "program" that "runs" on a processor to magically calculate it into existence. The understanding that is required first, makes the technological advancements needed look akin to choosing a colour. Today's AI community are 'rubbing two sticks together' awaiting other fields to emerge and make some advancements in understanding consciousness. Oh and I worked as a programmer for six years so don't give me dribble about parallel processing and such; still not one mm closer to what would be required.
I would never waste my time hearing speculation from a modern day AI or computer guy on this subject. I would highly recommend a book by Roger Penrose ; 'The Emperor's New Mind', and subsequent works. They are not intended to be solutions, but are more realistic than the sci-fi-ai that dominates the subject.
I would never waste my time hearing speculation from a modern day AI or computer guy on this subject. I would highly recommend a book by Roger Penrose ; 'The Emperor's New Mind', and subsequent works. They are not intended to be solutions, but are more realistic than the sci-fi-ai that dominates the subject.
Hi BigDumbWeirdo et al,
So you would be happy to be ruled by a computer - which operates to a set of rules - and you seem to think it possible to encapsulate all facets of the human condition in such a rule-set - with no scope to deviate from these rules to cope with complex, unforeseen developments? No thanks - not a world I would want to share with you!!!
So you would be happy to be ruled by a computer - which operates to a set of rules - and you seem to think it possible to encapsulate all facets of the human condition in such a rule-set - with no scope to deviate from these rules to cope with complex, unforeseen developments? No thanks - not a world I would want to share with you!!!
Thats the misunderstanding here,... the presumption that thought can be recreated via a algorithm; If such a AI thingy existed it wouldn't be governed by set calculated 'rules', ... it would have to be conscious and so in all respects except form a 'person',... any less of a definition while it may meet Alan Turing's standard, would not be considered 'intelligent'.
... most human problems cannot be solved by calculation as they have few black/white answers, because most of our problems lack enough facts to make it better than judgements. AI cannot be algorithmic and meet its ultimate goal of consciousness at the same time.
I recall hearing decades ago the idea of developing software to make cars drive themselves. Every once and a while I'll see a show or article again mentioning the idea, and i just roll my eyes. To use traditional programming or any algorithmic methodology to do the simplest human acts is immensly complicated. (fuzzy logic is bs as well).
I recall hearing decades ago the idea of developing software to make cars drive themselves. Every once and a while I'll see a show or article again mentioning the idea, and i just roll my eyes. To use traditional programming or any algorithmic methodology to do the simplest human acts is immensly complicated. (fuzzy logic is bs as well).
QUOTE (Edward 3+Jan 15 2008, 07:34 AM)
Hi BigDumbWeirdo et al,
So you would be happy to be ruled by a computer - which operates to a set of rules - and you seem to think it possible to encapsulate all facets of the human condition in such a rule-set - with no scope to deviate from these rules to cope with complex, unforeseen developments? No thanks - not a world I would want to share with you!!!
Several people in this thread have misinterpreted this point, yours is just the most recent, so I will elaborate using your post to provide context.
There would be no "Ruling." No world ruled over by a near-omniscient AI.
There would be "Governing." In the same way that other humans govern us now.
If you're going to confuse the two, then you could honestly claim that we are currently being ruled over by a selfish, oil hungry bully who was spoiled by his father's wealth and power his whole life, and wants nothing more than any dictator: to rule unopposed and be remembered as the savior of his people, despite having restricted the rights of his people. (There are a few good things to say about him, but not many, and they aren't relevant here.)
That doesn't sound very desirable, but if you're confusing "ruling" with "governing" it's accurate in your views.
Now if you describe "The President of the United States" with those terms, you'd be accurate, (if not COMPLETELY accurate, considering the lack of description of his few good qualities.)
It's the same principle with the AI I spoke of. It would have decision-making power, but not absolute power. Hence, it would not be ruling over us. It would be making economic decisions, distributing funding, dictating foreign policy, etc...
So you would be happy to be ruled by a computer - which operates to a set of rules - and you seem to think it possible to encapsulate all facets of the human condition in such a rule-set - with no scope to deviate from these rules to cope with complex, unforeseen developments? No thanks - not a world I would want to share with you!!!
Several people in this thread have misinterpreted this point, yours is just the most recent, so I will elaborate using your post to provide context.
There would be no "Ruling." No world ruled over by a near-omniscient AI.
There would be "Governing." In the same way that other humans govern us now.
If you're going to confuse the two, then you could honestly claim that we are currently being ruled over by a selfish, oil hungry bully who was spoiled by his father's wealth and power his whole life, and wants nothing more than any dictator: to rule unopposed and be remembered as the savior of his people, despite having restricted the rights of his people. (There are a few good things to say about him, but not many, and they aren't relevant here.)
That doesn't sound very desirable, but if you're confusing "ruling" with "governing" it's accurate in your views.
Now if you describe "The President of the United States" with those terms, you'd be accurate, (if not COMPLETELY accurate, considering the lack of description of his few good qualities.)
It's the same principle with the AI I spoke of. It would have decision-making power, but not absolute power. Hence, it would not be ruling over us. It would be making economic decisions, distributing funding, dictating foreign policy, etc...
Government by an AI would do no better than a president. The factor that would equalize the two is the problem that everyone being governed has a different idea of how to resolve issues, just as the electorate does now. This is human nature and there is no way around it. After a bad decision is made whose criteria would be used to adjust the AI algorithm for next time?
QUOTE (Edward3+)
Hi BigDumbWeirdo et al,
So you would be happy to be ruled by a computer - which operates to a set of rules - and you seem to think it possible to encapsulate all facets of the human condition in such a rule-set - with no scope to deviate from these rules to cope with complex, unforeseen developments? No thanks - not a world I would want to share with you!!!
Keeping aside for a moment the question on the possibility of encapsulating all facets of the human condition in a rule set, there are advantages as well as disadvantages to being ruled by AI.
Advantages
1) AI is would be free from selfishness (at-least initially).
2) AI would calculate the risk Vs reward much more accurately.
3) AI would never require rest and be available 24/7. It would never be on vacation.
4) AI would make decisions without any emotional investment. Eg: If the coutry is attacked by terrorist, it would not lose its cool and go about invading other countries which have nothing to do with it.
Disadvantages
1) AI would make decisions without any emotional investment. Eg: If killing billions of people in another country is strategically advantageous for its country, it may do it. (Unless it has been specifically programmed not to kill people, which is a disadvantage in itself during war).
2) It would require some people to maintain it in the long run, this may leave a gap for manipulation by vested interests.
coming to the question of the possibility of encapsulating all facets of the human condition in a rule set -
Since there is no one correct way to rule the country, there would be challenges when programming it. Eg: Would it be conservative or liberal?
So you would be happy to be ruled by a computer - which operates to a set of rules - and you seem to think it possible to encapsulate all facets of the human condition in such a rule-set - with no scope to deviate from these rules to cope with complex, unforeseen developments? No thanks - not a world I would want to share with you!!!
Keeping aside for a moment the question on the possibility of encapsulating all facets of the human condition in a rule set, there are advantages as well as disadvantages to being ruled by AI.
Advantages
1) AI is would be free from selfishness (at-least initially).
2) AI would calculate the risk Vs reward much more accurately.
3) AI would never require rest and be available 24/7. It would never be on vacation.
4) AI would make decisions without any emotional investment. Eg: If the coutry is attacked by terrorist, it would not lose its cool and go about invading other countries which have nothing to do with it.
Disadvantages
1) AI would make decisions without any emotional investment. Eg: If killing billions of people in another country is strategically advantageous for its country, it may do it. (Unless it has been specifically programmed not to kill people, which is a disadvantage in itself during war).
2) It would require some people to maintain it in the long run, this may leave a gap for manipulation by vested interests.
coming to the question of the possibility of encapsulating all facets of the human condition in a rule set -
Since there is no one correct way to rule the country, there would be challenges when programming it. Eg: Would it be conservative or liberal?
QUOTE (vkamath+Jan 15 2008, 06:34 PM)
4) AI would make decisions without any emotional investment. Eg: If the coutry is attacked by terrorist, it would not lose its cool and go about invading other countries which have nothing to do with it.
When the US invaded Iraq, is wasn't an emotional decision. It was one which they hoped would profit them, because of the involvement with the oil industry. AIs might not make emotional reactions, but they might make ones which we would regard as cold and cruel.
- Gehn
QUOTE (Gehn+)
When the US invaded Iraq, is wasn't an emotional decision. It was one which they hoped would profit them, because of the involvement with the oil industry. AIs might not make emotional reactions, but they might make ones which we would regard as cold and cruel.
Yes, it was not a emotional decision for the decision makers, but it was sold to Americans as a emotional decision by invoking their anger. Post 9-11 there was a lot of sabre-rattling that went on and still going on. Comments were made by the US president such us "If you are not with us, you are with them", which is ridiculous.
Overall I think USA overreacted to 9-11 (which is a topic for a different thread). Hence the security scenario is only worse compared to before 9-11. AI may do a better job.
Yes, it was not a emotional decision for the decision makers, but it was sold to Americans as a emotional decision by invoking their anger. Post 9-11 there was a lot of sabre-rattling that went on and still going on. Comments were made by the US president such us "If you are not with us, you are with them", which is ridiculous.
Overall I think USA overreacted to 9-11 (which is a topic for a different thread). Hence the security scenario is only worse compared to before 9-11. AI may do a better job.
QUOTE (gski+Jan 15 2008, 12:26 PM)
Government by an AI would do no better than a president. The factor that would equalize the two is the problem that everyone being governed has a different idea of how to resolve issues, just as the electorate does now. This is human nature and there is no way around it. After a bad decision is made whose criteria would be used to adjust the AI algorithm for next time?
What makes you think that it would make a mistake in the first place? And even if it did, why wouldn't it learn from it's mistake the way we do? After all, it's intelligent.
To vkamath:
I don't see why such an AI needs emotions or emotional considerations. A judge's job is to decide, based on circumstances, what evidence and behavior to allow or ban from a trial. An AI could do that job without the slightest hint of emotions, and likely better. (As a personal example, I was driving my camaro a few years ago when a pregnant, single mother changed lanes without checking her blind spot and rammed me, running me off the road. She had expired insurance, and when I finally got my day in court to collect the money she owed me for damages to my vehicle, the judge sided with her, simply because she was a pregnant single mother, and he felt sympathy for her. I eventually got the money she owed me {thanks to a no-nonsense female judge} but justice was initially denied me because of the judge's emotions.)
One could argue that such a system would result in rulings that always abide by the letter of the law, but that would be ignoring the legal concept of "intent." If it can be shown that one intended to break the law, then no amount of legal maneuvering can save them.
Similarly, economic decisions can be made with no consideration for human emotions. We may one day elect a president who chooses to dump large amounts of money into the CIA, to fund foreign assets who would support political candidates in their home countries who would dictate foreign policies favorable to the US at the expense of their own countries well being, in return for personal considerations from the US. This is a distinct possibility to happen on a large scale, and on small scales, happens on a daily basis. An AI with no emotions of fear or greed would not institute such a policy, recognizing that widening the economic gap between the US and the rest of the world would be harmful in the long run. (Consider the following: We currently import diamonds from Africa at cost X. If we followed the example I gave, we could see political leaders in Africa who arbitrarily lowered the price of diamonds to 0.5X. This then would reduce profits for the companies and individuals mining those diamonds, which in turn, would cut production, bringing costs up to >X, and engendering distrust and resentment of local governments as well as the US. If instead, we offered to pay 1.5X for the next 5 years for diamonds, profits would rise for the companies and individuals mining those diamonds, allowing for the purchase of better equipment and the hiring of more workers, in addition to the increase in wages existing workers might see, which itself would engender appreciation for the US, and could, over the course of a decade or two, drop the price of diamonds to a stable 0.5X, which is stable because of the advanced equipment and trained workers already present in the minds. The larger companies might not feel so good about the overall lowering of their profit margins, but this would be salved by the increase in gross profits that came from lowered overhead and increased market.{which itself would result from the decrease in value, as more uses for diamonds are opened up.} )
I would imagine it would be moderate. Such would please some of the people all of the time, and all of the people some of the time.
I would imagine it would be moderate. Such would please some of the people all of the time, and all of the people some of the time.
When the US invaded Iraq, is wasn't an emotional decision. It was one which they hoped would profit them, because of the involvement with the oil industry. AIs might not make emotional reactions, but they might make ones which we would regard as cold and cruel.
I disagree with that. Greed is an emotion.
I disagree with that. Greed is an emotion.
Hmmm....
If we programmed an AI, we would make it always take the most profitable decisions. It would merely be carrying out the action with the most gain for the government it belongs to, and it would not be feeling any emotion. We will usually carry out the action which is most profitable for us in a resonable way, but we will do it automatically. It wouldn't be greed in the case of the AI, but it would be in the case of Bush.
- Gehn
What makes you think that it would make a mistake in the first place? And even if it did, why wouldn't it learn from it's mistake the way we do? After all, it's intelligent.
To vkamath:
I don't see why such an AI needs emotions or emotional considerations. A judge's job is to decide, based on circumstances, what evidence and behavior to allow or ban from a trial. An AI could do that job without the slightest hint of emotions, and likely better. (As a personal example, I was driving my camaro a few years ago when a pregnant, single mother changed lanes without checking her blind spot and rammed me, running me off the road. She had expired insurance, and when I finally got my day in court to collect the money she owed me for damages to my vehicle, the judge sided with her, simply because she was a pregnant single mother, and he felt sympathy for her. I eventually got the money she owed me {thanks to a no-nonsense female judge} but justice was initially denied me because of the judge's emotions.)
One could argue that such a system would result in rulings that always abide by the letter of the law, but that would be ignoring the legal concept of "intent." If it can be shown that one intended to break the law, then no amount of legal maneuvering can save them.
Similarly, economic decisions can be made with no consideration for human emotions. We may one day elect a president who chooses to dump large amounts of money into the CIA, to fund foreign assets who would support political candidates in their home countries who would dictate foreign policies favorable to the US at the expense of their own countries well being, in return for personal considerations from the US. This is a distinct possibility to happen on a large scale, and on small scales, happens on a daily basis. An AI with no emotions of fear or greed would not institute such a policy, recognizing that widening the economic gap between the US and the rest of the world would be harmful in the long run. (Consider the following: We currently import diamonds from Africa at cost X. If we followed the example I gave, we could see political leaders in Africa who arbitrarily lowered the price of diamonds to 0.5X. This then would reduce profits for the companies and individuals mining those diamonds, which in turn, would cut production, bringing costs up to >X, and engendering distrust and resentment of local governments as well as the US. If instead, we offered to pay 1.5X for the next 5 years for diamonds, profits would rise for the companies and individuals mining those diamonds, allowing for the purchase of better equipment and the hiring of more workers, in addition to the increase in wages existing workers might see, which itself would engender appreciation for the US, and could, over the course of a decade or two, drop the price of diamonds to a stable 0.5X, which is stable because of the advanced equipment and trained workers already present in the minds. The larger companies might not feel so good about the overall lowering of their profit margins, but this would be salved by the increase in gross profits that came from lowered overhead and increased market.{which itself would result from the decrease in value, as more uses for diamonds are opened up.} )
QUOTE
Since there is no one correct way to rule the country, there would be challenges when programming it. Eg: Would it be conservative or liberal?
I would imagine it would be moderate. Such would please some of the people all of the time, and all of the people some of the time.
QUOTE (->
| QUOTE |
| Since there is no one correct way to rule the country, there would be challenges when programming it. Eg: Would it be conservative or liberal? |
I would imagine it would be moderate. Such would please some of the people all of the time, and all of the people some of the time.
When the US invaded Iraq, is wasn't an emotional decision. It was one which they hoped would profit them, because of the involvement with the oil industry. AIs might not make emotional reactions, but they might make ones which we would regard as cold and cruel.
I disagree with that. Greed is an emotion.
QUOTE (BigDumbWeirdo+Jan 15 2008, 07:07 PM)
I disagree with that. Greed is an emotion.
Hmmm....
If we programmed an AI, we would make it always take the most profitable decisions. It would merely be carrying out the action with the most gain for the government it belongs to, and it would not be feeling any emotion. We will usually carry out the action which is most profitable for us in a resonable way, but we will do it automatically. It wouldn't be greed in the case of the AI, but it would be in the case of Bush.
- Gehn
QUOTE (Gehn+Jan 15 2008, 02:14 PM)
Hmmm....
If we programmed an AI, we would make it always take the most profitable decisions. It would merely be carrying out the action with the most gain for the government it belongs to, and it would not be feeling any emotion. We will usually carry out the action which is most profitable for us in a resonable way, but we will do it automatically. It wouldn't be greed in the case of the AI, but it would be in the case of Bush.
- Gehn
If we programmed it in such a way, then it might. Or it might calculate probabilities of causes and effects years into the future, and instead of taking the first course of action I offered in my last post (on the subject of African Diamonds) it would take the second, seeing the greater profitable nature of actions which humanity would see as moral. This is something deadbeat fails to grasp in other threads: Logical behavior and moral behavior are often one and the same.
If we programmed an AI, we would make it always take the most profitable decisions. It would merely be carrying out the action with the most gain for the government it belongs to, and it would not be feeling any emotion. We will usually carry out the action which is most profitable for us in a resonable way, but we will do it automatically. It wouldn't be greed in the case of the AI, but it would be in the case of Bush.
- Gehn
If we programmed it in such a way, then it might. Or it might calculate probabilities of causes and effects years into the future, and instead of taking the first course of action I offered in my last post (on the subject of African Diamonds) it would take the second, seeing the greater profitable nature of actions which humanity would see as moral. This is something deadbeat fails to grasp in other threads: Logical behavior and moral behavior are often one and the same.
For BigDumbWeirdo: My point is that if it made a decision there would always be a group that thinks it is a mistake, just as now there is always some group to say a political decision is wrong.
QUOTE (gski+Jan 15 2008, 02:46 PM)
For BigDumbWeirdo: My point is that if it made a decision there would always be a group that thinks it is a mistake, just as now there is always some group to say a political decision is wrong.
It would be hard for such a group to justify it's claims, considering the vastly superior intellect of the AI, coupled with it's greater understanding of the results of it's decisions and the relevant legality.
It would be hard for such a group to justify it's claims, considering the vastly superior intellect of the AI, coupled with it's greater understanding of the results of it's decisions and the relevant legality.
The minute judges are replaced by AI is the minute they will be outlawed by his compatriots!
I shouldn't want a world where peoples jobs are taken by a machine...even though ideally that would potentially be the ultimate in objective reasoning if processed correctly in a court-room scenario.
I would rather not work with bots either. I prefer the real human face to look upon, and to ask the real human to lunch on our break. What would Ai talk about over lunch where they could not share the bond of food? Thier love life? Thier AI cats & dogs...the sickening reality of having too many dishes to do every night after work? I don't think so.
But if stepford wives could only see us now, we are well on our way to making them replace women. I wonder if they are making life-like love orifices?
http://www.youtube.com/watch?v=074fQp15aZc
http://www.youtube.com/watch?v=d_Zy2sqxe5w
http://www.youtube.com/watch?v=MY8-sJS0W1I
http://www.youtube.com/watch?v=074fQp15aZc
I shouldn't want a world where peoples jobs are taken by a machine...even though ideally that would potentially be the ultimate in objective reasoning if processed correctly in a court-room scenario.
I would rather not work with bots either. I prefer the real human face to look upon, and to ask the real human to lunch on our break. What would Ai talk about over lunch where they could not share the bond of food? Thier love life? Thier AI cats & dogs...the sickening reality of having too many dishes to do every night after work? I don't think so.
But if stepford wives could only see us now, we are well on our way to making them replace women. I wonder if they are making life-like love orifices?
http://www.youtube.com/watch?v=074fQp15aZc
http://www.youtube.com/watch?v=d_Zy2sqxe5w
http://www.youtube.com/watch?v=MY8-sJS0W1I
http://www.youtube.com/watch?v=074fQp15aZc
QUOTE (gski+Jan 15 2008, 10:26 AM)
Government by an AI would do no better than a president. The factor that would equalize the two is the problem that everyone being governed has a different idea of how to resolve issues, just as the electorate does now. This is human nature and there is no way around it. After a bad decision is made whose criteria would be used to adjust the AI algorithm for next time?
I think everyone does have different ideas on how to govern a country, a town or city or state for that matter. So who is programming these AI with the capability to satisfy hordes of people at once...good question. I am with you..it would be an impossible task to satisfy both wealthy, and then poverty stricken people at once....just as it would be hard to find a balance between consevatives and liberals.
Then it becomes WHO is the deciding factor in programming these machines? And how to make them think like people, and yet, with more objectivity and less emotional distraction than humans?
After a bad decision is made how easy to reprogram them to just walk the dog?
I think everyone does have different ideas on how to govern a country, a town or city or state for that matter. So who is programming these AI with the capability to satisfy hordes of people at once...good question. I am with you..it would be an impossible task to satisfy both wealthy, and then poverty stricken people at once....just as it would be hard to find a balance between consevatives and liberals.
Then it becomes WHO is the deciding factor in programming these machines? And how to make them think like people, and yet, with more objectivity and less emotional distraction than humans?
After a bad decision is made how easy to reprogram them to just walk the dog?
The fundamental flaw in all of this is the acceptance of IQ as a valid measure of capacity to make sound decisions - in fact IQ is not even an accurate measure of intelligence - it is probably no more than a measure of ability to perform in IQ tests.
QUOTE (BigDumbWeirdo+)
To vkamath:
I don't see why such an AI needs emotions or emotional considerations. A judge's job is to decide, based on circumstances, what evidence and behavior to allow or ban from a trial. An AI could do that job without the slightest hint of emotions, and likely better. (As a personal example, I was driving my camaro a few years ago when a pregnant, single mother changed lanes without checking her blind spot and rammed me, running me off the road. She had expired insurance, and when I finally got my day in court to collect the money she owed me for damages to my vehicle, the judge sided with her, simply because she was a pregnant single mother, and he felt sympathy for her. I eventually got the money she owed me {thanks to a no-nonsense female judge} but justice was initially denied me because of the judge's emotions.)
One could argue that such a system would result in rulings that always abide by the letter of the law, but that would be ignoring the legal concept of "intent." If it can be shown that one intended to break the law, then no amount of legal maneuvering can save them.
I believe emotional considerations have to be somehow programmed into the AI. For Eg: Nuking China may free up a lot of resources for use by us. But that doesn't mean we should do that.
I don't see why such an AI needs emotions or emotional considerations. A judge's job is to decide, based on circumstances, what evidence and behavior to allow or ban from a trial. An AI could do that job without the slightest hint of emotions, and likely better. (As a personal example, I was driving my camaro a few years ago when a pregnant, single mother changed lanes without checking her blind spot and rammed me, running me off the road. She had expired insurance, and when I finally got my day in court to collect the money she owed me for damages to my vehicle, the judge sided with her, simply because she was a pregnant single mother, and he felt sympathy for her. I eventually got the money she owed me {thanks to a no-nonsense female judge} but justice was initially denied me because of the judge's emotions.)
One could argue that such a system would result in rulings that always abide by the letter of the law, but that would be ignoring the legal concept of "intent." If it can be shown that one intended to break the law, then no amount of legal maneuvering can save them.
I believe emotional considerations have to be somehow programmed into the AI. For Eg: Nuking China may free up a lot of resources for use by us. But that doesn't mean we should do that.
QUOTE (BigDumbWeirdo+)
This is something deadbeat fails to grasp in other threads: Logical behavior and moral behavior are often one and the same.
Morality is a most complex subject. It is very hard to determine by logic. Eg: Is it possible to determine by logic if eating non-vegetarian food is good or bad?
If the answer is - Good. Then where do you draw the line? Is eating humans also good? What about monkeys?
If the answer is - Bad. Then should all of us become vegetarians? Can we eat fish or eggs?
Morality is a most complex subject. It is very hard to determine by logic. Eg: Is it possible to determine by logic if eating non-vegetarian food is good or bad?
If the answer is - Good. Then where do you draw the line? Is eating humans also good? What about monkeys?
If the answer is - Bad. Then should all of us become vegetarians? Can we eat fish or eggs?
QUOTE (vkamath+Jan 15 2008, 03:42 PM)
I believe emotional considerations have to be somehow programmed into the AI. For Eg: Nuking China may free up a lot of resources for use by us. But that doesn't mean we should do that.
Of course not, but there are logical reasons why we should not do that. (The possibility of reprisal by Chinese-friendly nuclear powers, the possibility of reprisal by whatever Chinese military forces remain, the very real effect of condemnation by the rest of the world, the nuclear fallout that would result from such a large-scale nuclear assault, even given today's low-fallout nukes.)
For the most part, I agree, but difficulty and impossibility are not the same.
In the example you give, determining the morality by emotional standards is fairly simple: Simply decide whether or not animals deserve the right to not be eaten.
For that, we look to nature. Animals eat other animals, it is the natural way of things. Then we look to humanity. Strict vegetarian diets are not healthy (despite what vegetarians and vegans claim,) thus promoting such diets would increase the amount of suffering in the world.
Therefore, by two standards eating meat is good, as determined by purely emotional considerations.
Then we look at it from a logical perspective. The initial question is the same: Do animals deserve the right to not be eaten?
We look at the animal kingdom: animals are eaten by other animals, which contributes to evolution by way of natural selection. Then we look at humanity: Strict vegetarian diets are not healthy, and so we come once again to the unavoidable fact that vegetarianism diets contribute to medical problems, problems which cost resources to solve, thus taking precious resources away from curing less preventable illnesses.
Your example was not a particularly good one for illustrating the point. However, I still agree that it is often difficult to determine morality through the use of logic alone, and far easier to determine morality through emotional considerations.
A better example would be this: Your brother robbed a bank to feed his starving children after being laid off from the only job he's ever held. Do you turn him in?
Initially, logic and morality are at odds. Logically, you turn him in so as to avoid the possibility of jail time yourself. Emotionally, you do not, because he is your brother and because he did so to feed his children. However, when you start taking into account all the variables involved in the situation, things become far more complex for the logical side. Did your brother have opportunities other than committing a bank heist? Did your brother harm anyone during the heist? Will your brother be denied a fair trial if caught? etc, etc, etc...
Note that some of these questions apply to the emotional side of the matter, as well. When you really go in-depth, you can get the same answer from your emotions as you do from logic, or vice verse, but you must be willing to consider all the aspects, which is quite difficult to do when faced with a moral dilemma.
Now, to address the specific questions you asked:
For the most part, I agree, but difficulty and impossibility are not the same.
In the example you give, determining the morality by emotional standards is fairly simple: Simply decide whether or not animals deserve the right to not be eaten.
For that, we look to nature. Animals eat other animals, it is the natural way of things. Then we look to humanity. Strict vegetarian diets are not healthy (despite what vegetarians and vegans claim,) thus promoting such diets would increase the amount of suffering in the world.
Therefore, by two standards eating meat is good, as determined by purely emotional considerations.
Then we look at it from a logical perspective. The initial question is the same: Do animals deserve the right to not be eaten?
We look at the animal kingdom: animals are eaten by other animals, which contributes to evolution by way of natural selection. Then we look at humanity: Strict vegetarian diets are not healthy, and so we come once again to the unavoidable fact that vegetarianism diets contribute to medical problems, problems which cost resources to solve, thus taking precious resources away from curing less preventable illnesses.
Your example was not a particularly good one for illustrating the point. However, I still agree that it is often difficult to determine morality through the use of logic alone, and far easier to determine morality through emotional considerations.
A better example would be this: Your brother robbed a bank to feed his starving children after being laid off from the only job he's ever held. Do you turn him in?
Initially, logic and morality are at odds. Logically, you turn him in so as to avoid the possibility of jail time yourself. Emotionally, you do not, because he is your brother and because he did so to feed his children. However, when you start taking into account all the variables involved in the situation, things become far more complex for the logical side. Did your brother have opportunities other than committing a bank heist? Did your brother harm anyone during the heist? Will your brother be denied a fair trial if caught? etc, etc, etc...
Note that some of these questions apply to the emotional side of the matter, as well. When you really go in-depth, you can get the same answer from your emotions as you do from logic, or vice verse, but you must be willing to consider all the aspects, which is quite difficult to do when faced with a moral dilemma.
Now, to address the specific questions you asked:
Then where do you draw the line? Is eating humans also good? What about monkeys?
That is where the logical considerations get trickier in your question, but not too tricky. Do we eat humans? No, because each human has an unknown amount of potential to do things which could improve the lives of humanity as a whole. Do we eat monkeys? That requires a re-visit of the questions we used to determine if eating meat is good or bad. Are monkeys eaten in the wild? Is eating monkeys good nutrition? Then you will have determined a logical answer to that question. (I don't know the answer to either so I'm not going to answer them.)
Such an answer may be different from the emotional answer, but that doesn't make it less valid.
In this case the questions are purely human in nature. 'Vegetarians' who eat fish and eggs and cheese and drink milk are healthier than vegans who eat no animal products at all. Therefore, logically and emotionally, yes.
Of course not, but there are logical reasons why we should not do that. (The possibility of reprisal by Chinese-friendly nuclear powers, the possibility of reprisal by whatever Chinese military forces remain, the very real effect of condemnation by the rest of the world, the nuclear fallout that would result from such a large-scale nuclear assault, even given today's low-fallout nukes.)
QUOTE
Morality is a most complex subject. It is very hard to determine by logic. Eg: Is it possible to determine by logic if eating non-vegetarian food is good or bad?
If the answer is - Good. Then where do you draw the line? Is eating humans also good? What about monkeys?
If the answer is - Bad. Then should all of us become vegetarians? Can we eat fish or eggs?
If the answer is - Good. Then where do you draw the line? Is eating humans also good? What about monkeys?
If the answer is - Bad. Then should all of us become vegetarians? Can we eat fish or eggs?
For the most part, I agree, but difficulty and impossibility are not the same.
In the example you give, determining the morality by emotional standards is fairly simple: Simply decide whether or not animals deserve the right to not be eaten.
For that, we look to nature. Animals eat other animals, it is the natural way of things. Then we look to humanity. Strict vegetarian diets are not healthy (despite what vegetarians and vegans claim,) thus promoting such diets would increase the amount of suffering in the world.
Therefore, by two standards eating meat is good, as determined by purely emotional considerations.
Then we look at it from a logical perspective. The initial question is the same: Do animals deserve the right to not be eaten?
We look at the animal kingdom: animals are eaten by other animals, which contributes to evolution by way of natural selection. Then we look at humanity: Strict vegetarian diets are not healthy, and so we come once again to the unavoidable fact that vegetarianism diets contribute to medical problems, problems which cost resources to solve, thus taking precious resources away from curing less preventable illnesses.
Your example was not a particularly good one for illustrating the point. However, I still agree that it is often difficult to determine morality through the use of logic alone, and far easier to determine morality through emotional considerations.
A better example would be this: Your brother robbed a bank to feed his starving children after being laid off from the only job he's ever held. Do you turn him in?
Initially, logic and morality are at odds. Logically, you turn him in so as to avoid the possibility of jail time yourself. Emotionally, you do not, because he is your brother and because he did so to feed his children. However, when you start taking into account all the variables involved in the situation, things become far more complex for the logical side. Did your brother have opportunities other than committing a bank heist? Did your brother harm anyone during the heist? Will your brother be denied a fair trial if caught? etc, etc, etc...
Note that some of these questions apply to the emotional side of the matter, as well. When you really go in-depth, you can get the same answer from your emotions as you do from logic, or vice verse, but you must be willing to consider all the aspects, which is quite difficult to do when faced with a moral dilemma.
Now, to address the specific questions you asked:
QUOTE (->
| QUOTE |
| Morality is a most complex subject. It is very hard to determine by logic. Eg: Is it possible to determine by logic if eating non-vegetarian food is good or bad? If the answer is - Good. Then where do you draw the line? Is eating humans also good? What about monkeys? If the answer is - Bad. Then should all of us become vegetarians? Can we eat fish or eggs? |
For the most part, I agree, but difficulty and impossibility are not the same.
In the example you give, determining the morality by emotional standards is fairly simple: Simply decide whether or not animals deserve the right to not be eaten.
For that, we look to nature. Animals eat other animals, it is the natural way of things. Then we look to humanity. Strict vegetarian diets are not healthy (despite what vegetarians and vegans claim,) thus promoting such diets would increase the amount of suffering in the world.
Therefore, by two standards eating meat is good, as determined by purely emotional considerations.
Then we look at it from a logical perspective. The initial question is the same: Do animals deserve the right to not be eaten?
We look at the animal kingdom: animals are eaten by other animals, which contributes to evolution by way of natural selection. Then we look at humanity: Strict vegetarian diets are not healthy, and so we come once again to the unavoidable fact that vegetarianism diets contribute to medical problems, problems which cost resources to solve, thus taking precious resources away from curing less preventable illnesses.
Your example was not a particularly good one for illustrating the point. However, I still agree that it is often difficult to determine morality through the use of logic alone, and far easier to determine morality through emotional considerations.
A better example would be this: Your brother robbed a bank to feed his starving children after being laid off from the only job he's ever held. Do you turn him in?
Initially, logic and morality are at odds. Logically, you turn him in so as to avoid the possibility of jail time yourself. Emotionally, you do not, because he is your brother and because he did so to feed his children. However, when you start taking into account all the variables involved in the situation, things become far more complex for the logical side. Did your brother have opportunities other than committing a bank heist? Did your brother harm anyone during the heist? Will your brother be denied a fair trial if caught? etc, etc, etc...
Note that some of these questions apply to the emotional side of the matter, as well. When you really go in-depth, you can get the same answer from your emotions as you do from logic, or vice verse, but you must be willing to consider all the aspects, which is quite difficult to do when faced with a moral dilemma.
Now, to address the specific questions you asked:
Then where do you draw the line? Is eating humans also good? What about monkeys?
That is where the logical considerations get trickier in your question, but not too tricky. Do we eat humans? No, because each human has an unknown amount of potential to do things which could improve the lives of humanity as a whole. Do we eat monkeys? That requires a re-visit of the questions we used to determine if eating meat is good or bad. Are monkeys eaten in the wild? Is eating monkeys good nutrition? Then you will have determined a logical answer to that question. (I don't know the answer to either so I'm not going to answer them.)
Such an answer may be different from the emotional answer, but that doesn't make it less valid.
QUOTE
Then should all of us become vegetarians? Can we eat fish or eggs?
In this case the questions are purely human in nature. 'Vegetarians' who eat fish and eggs and cheese and drink milk are healthier than vegans who eat no animal products at all. Therefore, logically and emotionally, yes.
I see that you have answered the question only from the perspective of the human, which is fine. But what about the animal? Doesn't it have the right to live?
QUOTE (BigDumbWeirdo+)
We look at the animal kingdom: animals are eaten by other animals, which contributes to evolution by way of natural selection. Then we look at humanity: Strict vegetarian diets are not healthy, and so we come once again to the unavoidable fact that vegetarianism diets contribute to medical problems, problems which cost resources to solve, thus taking precious resources away from curing less preventable illnesses.
But strict vegetarian diets by humans are quite "healthy" for the animal which is saved from becoming food.
But strict vegetarian diets by humans are quite "healthy" for the animal which is saved from becoming food.
QUOTE (BigDumbWeirdo+)
That requires a re-visit of the questions we used to determine if eating meat is good or bad. Are monkeys eaten in the wild? Is eating monkeys good nutrition? Then you will have determined a logical answer to that question. (I don't know the answer to either so I'm not going to answer them.)
I am not sure where you are going with the question on monkeys being eaten in the wild. In the wild, there are animals that eat its own young, probably to generate enough breast milk to feed its surviving stronger young or for itself so that it can produce more healthier young ones in future.
Why is something that happens in the wild a standard for what human do?
Anyway I don't have anything against being non-vegetarian, being one myself. Just trying to make a point.
I am not sure where you are going with the question on monkeys being eaten in the wild. In the wild, there are animals that eat its own young, probably to generate enough breast milk to feed its surviving stronger young or for itself so that it can produce more healthier young ones in future.
Why is something that happens in the wild a standard for what human do?
Anyway I don't have anything against being non-vegetarian, being one myself. Just trying to make a point.
QUOTE (vkamath+Jan 15 2008, 05:03 PM)
I see that you have answered the question only from the perspective of the human, which is fine. But what about the animal? Doesn't it have the right to live?
That's the basis of the question "do animals get eaten in the wild?" and the conclusion that such provides the natural selection from which evolution arises.
Hence the question of the animal's rights. Does an animal deserve the same rights as humans? Some people would say yes, but considering the scandals surrounding PETA and other such organizations of people who say "yes" to that question, I would say that these people are not being rational. Besides, we as humans have an obligation to support humanity's well-being over the well-being of animals. A moral obligation. To grant animals equal rights to a human would harm humanity in so many different ways it's hard to name them all.
To answer the forthcoming question, "why is it a moral obligation to support humanity over the animal kingdom?"
It is because we have a choice between promoting the well being of one or the other. When there is no significant deciding factor, we must choose an insignificant factor upon which to base our decision: Because the one making the decision is him/herself, human.
Additionally, there is the fact that a human -with more emotional complexity and self-awareness than an animal- is more pained by suffering than an animal.
All of that is the logical answer, the emotional answer is quite a bit simpler, as is to be expected.
Which would you rather see die, an animal or a human? (No, the human is neither a saint nor a serial killer, but a normal human being.) To the emotionally stable person, the choice is clear, the animal must die.
Hence the question of the animal's rights. Does an animal deserve the same rights as humans? Some people would say yes, but considering the scandals surrounding PETA and other such organizations of people who say "yes" to that question, I would say that these people are not being rational. Besides, we as humans have an obligation to support humanity's well-being over the well-being of animals. A moral obligation. To grant animals equal rights to a human would harm humanity in so many different ways it's hard to name them all.
To answer the forthcoming question, "why is it a moral obligation to support humanity over the animal kingdom?"
It is because we have a choice between promoting the well being of one or the other. When there is no significant deciding factor, we must choose an insignificant factor upon which to base our decision: Because the one making the decision is him/herself, human.
Additionally, there is the fact that a human -with more emotional complexity and self-awareness than an animal- is more pained by suffering than an animal.
All of that is the logical answer, the emotional answer is quite a bit simpler, as is to be expected.
Which would you rather see die, an animal or a human? (No, the human is neither a saint nor a serial killer, but a normal human being.) To the emotionally stable person, the choice is clear, the animal must die.
Why is something that happens in the wild a standard for what human do?
Because the wild is the world in which we live. We may control our immediate environments, but that was not always the case. We sprang forth from the wild, and thus the wild provides a useful moral compass when other considerations cannot give a clear answer.
Which would you rather see die, an animal or a human? (No, the human is neither a saint nor a serial killer, but a normal human being.) To the emotionally stable person, the choice is clear, the animal must die.
The average human is sometimes content to stand watching while an animal bites and tears at the flesh of a human ( even a woman with a baby) it seems. When I was walking once with Sarah at six months old in a stroller, I was attacked by a ferocious dog in the neighborhood who had gotton over the fence at the house where he was starved and left with no water on a daily basis. The people in that home were breeding the dog to harm others if they should enter the land.
How many lived within this home and did not worry over others? Two men stood simply conversing while the dog fed on my arm and I screamed out for assistance...they had strudy construction worker boots on and made no move toward my person. They could see the babys stroller there by me, and myself being attacked by a large german shepard. They did nothing but stand talking in the next driveway while this went on.
I looked at them and cried out...I turned away and yelled for help...a neighbor finally ran from across the road and scared the dog away. Yelling it's name in anger! I had bites from my wrist to my shoulder. I held my arm out for the dog to eat at it, because ripping my arm away meant he might dive upon my helpless infant...I was very suprised at the two men who continued to stand talking as I went across the street to get patched up a bit.
This example is written to show that 2 out of 3 people in a group of three men, were unconcerned (unconditioned?) for the care of the human in a scene where animal was faced off with the human, and if the dog had been allowed to continue I knew my neck was the next place up where his teeth were sinking into my body, and that if it got so far that I was in very big trouble, possibly facing death.
How is it that both the "owners" who reared the dog and these men held so little concern of man over beast? How common is this phenomenon, do you think? I think it happens all the time.
That family and I spoke about the two men and their lack of concern for me, and we were in awe of the situation. I think people are conditioned to keep themselves safe at all costs, even with the benefit of being heroic...
And that is something we could condition into AI (to be on topic) we could have them conditioned to care for humans over animals in such a case. But could we condition them to go to the aid of a human in need? And could we allow them to have weapons on thier machinery? Something like a stun-gun for situations like that one?
It would probably live a slow life and become the leader of the Slow Movement worldwide and die one day of procedure division code being found in the data division.
It took me a long, long time to quit dreaming in COBOL. I do still have the occasional nightmare in RPG, though...
Ok. Find me that list of multiplication tables programmed into my pocket calculator.
It's not there.
You're wrong.
That's the basis of the question "do animals get eaten in the wild?" and the conclusion that such provides the natural selection from which evolution arises.
QUOTE
But strict vegetarian diets by humans are quite "healthy" for the animal which is saved from becoming food.
Hence the question of the animal's rights. Does an animal deserve the same rights as humans? Some people would say yes, but considering the scandals surrounding PETA and other such organizations of people who say "yes" to that question, I would say that these people are not being rational. Besides, we as humans have an obligation to support humanity's well-being over the well-being of animals. A moral obligation. To grant animals equal rights to a human would harm humanity in so many different ways it's hard to name them all.
To answer the forthcoming question, "why is it a moral obligation to support humanity over the animal kingdom?"
It is because we have a choice between promoting the well being of one or the other. When there is no significant deciding factor, we must choose an insignificant factor upon which to base our decision: Because the one making the decision is him/herself, human.
Additionally, there is the fact that a human -with more emotional complexity and self-awareness than an animal- is more pained by suffering than an animal.
All of that is the logical answer, the emotional answer is quite a bit simpler, as is to be expected.
Which would you rather see die, an animal or a human? (No, the human is neither a saint nor a serial killer, but a normal human being.) To the emotionally stable person, the choice is clear, the animal must die.
QUOTE (->
| QUOTE |
| But strict vegetarian diets by humans are quite "healthy" for the animal which is saved from becoming food. |
Hence the question of the animal's rights. Does an animal deserve the same rights as humans? Some people would say yes, but considering the scandals surrounding PETA and other such organizations of people who say "yes" to that question, I would say that these people are not being rational. Besides, we as humans have an obligation to support humanity's well-being over the well-being of animals. A moral obligation. To grant animals equal rights to a human would harm humanity in so many different ways it's hard to name them all.
To answer the forthcoming question, "why is it a moral obligation to support humanity over the animal kingdom?"
It is because we have a choice between promoting the well being of one or the other. When there is no significant deciding factor, we must choose an insignificant factor upon which to base our decision: Because the one making the decision is him/herself, human.
Additionally, there is the fact that a human -with more emotional complexity and self-awareness than an animal- is more pained by suffering than an animal.
All of that is the logical answer, the emotional answer is quite a bit simpler, as is to be expected.
Which would you rather see die, an animal or a human? (No, the human is neither a saint nor a serial killer, but a normal human being.) To the emotionally stable person, the choice is clear, the animal must die.
Why is something that happens in the wild a standard for what human do?
Because the wild is the world in which we live. We may control our immediate environments, but that was not always the case. We sprang forth from the wild, and thus the wild provides a useful moral compass when other considerations cannot give a clear answer.
Colossus, Terminator, 7 of 9, whatever.
If you program it in COBOL, does it live? If yes, does it live well?
If you program it in COBOL, does it live? If yes, does it live well?
QUOTE (Sapo+Jan 17 2008, 01:08 AM)
Colossus, Terminator, 7 of 9, whatever.
If you program it in COBOL, does it live? If yes, does it live well?
It would probably live a slow life and become the leader of the Slow Movement worldwide and die one day of procedure division code being found in the data division.
If you program it in COBOL, does it live? If yes, does it live well?
It would probably live a slow life and become the leader of the Slow Movement worldwide and die one day of procedure division code being found in the data division.
QUOTE (BigDumbWeirdo+Jan 15 2008, 03:28 PM)
Which would you rather see die, an animal or a human? (No, the human is neither a saint nor a serial killer, but a normal human being.) To the emotionally stable person, the choice is clear, the animal must die.
The average human is sometimes content to stand watching while an animal bites and tears at the flesh of a human ( even a woman with a baby) it seems. When I was walking once with Sarah at six months old in a stroller, I was attacked by a ferocious dog in the neighborhood who had gotton over the fence at the house where he was starved and left with no water on a daily basis. The people in that home were breeding the dog to harm others if they should enter the land.
How many lived within this home and did not worry over others? Two men stood simply conversing while the dog fed on my arm and I screamed out for assistance...they had strudy construction worker boots on and made no move toward my person. They could see the babys stroller there by me, and myself being attacked by a large german shepard. They did nothing but stand talking in the next driveway while this went on.
I looked at them and cried out...I turned away and yelled for help...a neighbor finally ran from across the road and scared the dog away. Yelling it's name in anger! I had bites from my wrist to my shoulder. I held my arm out for the dog to eat at it, because ripping my arm away meant he might dive upon my helpless infant...I was very suprised at the two men who continued to stand talking as I went across the street to get patched up a bit.
This example is written to show that 2 out of 3 people in a group of three men, were unconcerned (unconditioned?) for the care of the human in a scene where animal was faced off with the human, and if the dog had been allowed to continue I knew my neck was the next place up where his teeth were sinking into my body, and that if it got so far that I was in very big trouble, possibly facing death.
How is it that both the "owners" who reared the dog and these men held so little concern of man over beast? How common is this phenomenon, do you think? I think it happens all the time.
That family and I spoke about the two men and their lack of concern for me, and we were in awe of the situation. I think people are conditioned to keep themselves safe at all costs, even with the benefit of being heroic...
And that is something we could condition into AI (to be on topic) we could have them conditioned to care for humans over animals in such a case. But could we condition them to go to the aid of a human in need? And could we allow them to have weapons on thier machinery? Something like a stun-gun for situations like that one?
That's why failure to render aid is a crime. 
QUOTE (Sapo+Jan 17 2008, 12:13 PM)
That's why failure to render aid is a crime.
While I'm not surprised at the thought that the story Tikay just told could have taken place, I'm appalled at the reaction of the two men.
Tikay: I assure you that if I had been there I would have come to your assistance, killed the dog (animal control would have killed it anyways, had they been called) and done my utmost to beat into a bloody wreck the two men who stood idly by while a fellow human being was being attacked by an obviously vicious animal. There's no excuse for that behavior.
And for the record, I have been attacked by dogs myself, and have come to the aid (once) of a young girl (about 14 or 15) being attacked by a rottweiler that lived in my neighborhood. Most people are afraid of dogs, but in truth, most dogs are no match for a human in a physical confrontation.
If you're ever attacked by a dog again, don't let fear keep you from hurting it. Poke it's eyes, pick it up and throw it, stomp on it's neck, kick it soccer-style, just go ballistic, most likely the dog will run off the moment you cause it real pain, but even if not, you can kill a dog rather easily with your bare hands, if you're willing to do what it takes. Even a 150lb rottweiler can't match the agility, upper body strength, and variety of attacks available to a 120lb human.
While I'm not surprised at the thought that the story Tikay just told could have taken place, I'm appalled at the reaction of the two men.
Tikay: I assure you that if I had been there I would have come to your assistance, killed the dog (animal control would have killed it anyways, had they been called) and done my utmost to beat into a bloody wreck the two men who stood idly by while a fellow human being was being attacked by an obviously vicious animal. There's no excuse for that behavior.
And for the record, I have been attacked by dogs myself, and have come to the aid (once) of a young girl (about 14 or 15) being attacked by a rottweiler that lived in my neighborhood. Most people are afraid of dogs, but in truth, most dogs are no match for a human in a physical confrontation.
If you're ever attacked by a dog again, don't let fear keep you from hurting it. Poke it's eyes, pick it up and throw it, stomp on it's neck, kick it soccer-style, just go ballistic, most likely the dog will run off the moment you cause it real pain, but even if not, you can kill a dog rather easily with your bare hands, if you're willing to do what it takes. Even a 150lb rottweiler can't match the agility, upper body strength, and variety of attacks available to a 120lb human.
QUOTE (BigDumbWeirdo+Jan 17 2008, 10:27 AM)
While I'm not surprised at the thought that the story Tikay just told could have taken place, I'm appalled at the reaction of the two men.
Tikay: I assure you that if I had been there I would have come to your assistance, killed the dog (animal control would have killed it anyways, had they been called) and done my utmost to beat into a bloody wreck the two men who stood idly by while a fellow human being was being attacked by an obviously vicious animal. There's no excuse for that behavior.
And for the record, I have been attacked by dogs myself, and have come to the aid (once) of a young girl (about 14 or 15) being attacked by a rottweiler that lived in my neighborhood. Most people are afraid of dogs, but in truth, most dogs are no match for a human in a physical confrontation.
If you're ever attacked by a dog again, don't let fear keep you from hurting it. Poke it's eyes, pick it up and throw it, stomp on it's neck, kick it soccer-style, just go ballistic, most likely the dog will run off the moment you cause it real pain, but even if not, you can kill a dog rather easily with your bare hands, if you're willing to do what it takes. Even a 150lb rottweiler can't match the agility, upper body strength, and variety of attacks available to a 120lb human.
Good going BDW, on your rescue of the girl and on your own behalf...
Thanks Sapo for you anger that aid was not rendered by two out of three, but the third person came to my aid and rescued me and we became friends with that family and, the dog was taken away that day and possibly put down. We could not allow a dog like that to have another chance to hurt someone else. The reason for the confrontation with the dog was that, he ran up to me on the street walking along with Sarah in her stroller, and he grabbed a brand new blanket which I had laid over her.
I wanted that blanket back, and followed the dog, not knowing it was so vicious, tried to see where he took the blanket...he dropped it somewhere behind his house and came trotting back down his driveway...I stuck my hand out for him to smell and that dog just began to munch on my arm moving up and up. Like I said after that I held my arm out for him to bite on, scared for my baby, and began to scream for help. It's hurts there where those bites are to this day. A sort of phantom pain remains in my right arm whenever dog attacks come up. It was not the first time I was attacked either...I should have known better. It was the third time!
At least something good came of that attack and I am fine...we had new friends in the neighborhood (the guy who rescued me and his family) and the dog was either rescued too and fed and taken care of or not, but was no longer a threat to our neighborhood.
Again I wonder about an attack on a person in the possible future with AI...would we have them programmed to come to the aid of someone being assaulted ?
Would they be used for armies? Would they have a notion about caring and heroic deeds ? Would they be able to use weaponry to save a life? Would they be able to manuver over grassy terrain...sandy, rocky or hilly terrain? These things come up for me and would we be able to control them and have them not trying to control us? I can't bear the thought of these inhuman things controlling the destiny of man.
But if people think they should hold positions of power then, that is pretty odd to me, but it shows me that man feels less competant to fix things on their own...
I don't know how much control I'd allow a thing not human. I would be very wary of letting something without a heart and soul control me. If someone told me to bow to the machine Id have to say no, under most circumstances, I think. Considering the outcome of course. if I was goign to be imprisioned for not bowing to that machines rules I might bow physically only. But my heart will never allow me to bow to machines, I think man should keep control of this thing and not the other way around. If they had IQ that made us look like children, Id be very wary of that future.
Tikay: I assure you that if I had been there I would have come to your assistance, killed the dog (animal control would have killed it anyways, had they been called) and done my utmost to beat into a bloody wreck the two men who stood idly by while a fellow human being was being attacked by an obviously vicious animal. There's no excuse for that behavior.
And for the record, I have been attacked by dogs myself, and have come to the aid (once) of a young girl (about 14 or 15) being attacked by a rottweiler that lived in my neighborhood. Most people are afraid of dogs, but in truth, most dogs are no match for a human in a physical confrontation.
If you're ever attacked by a dog again, don't let fear keep you from hurting it. Poke it's eyes, pick it up and throw it, stomp on it's neck, kick it soccer-style, just go ballistic, most likely the dog will run off the moment you cause it real pain, but even if not, you can kill a dog rather easily with your bare hands, if you're willing to do what it takes. Even a 150lb rottweiler can't match the agility, upper body strength, and variety of attacks available to a 120lb human.
Good going BDW, on your rescue of the girl and on your own behalf...
Thanks Sapo for you anger that aid was not rendered by two out of three, but the third person came to my aid and rescued me and we became friends with that family and, the dog was taken away that day and possibly put down. We could not allow a dog like that to have another chance to hurt someone else. The reason for the confrontation with the dog was that, he ran up to me on the street walking along with Sarah in her stroller, and he grabbed a brand new blanket which I had laid over her.
I wanted that blanket back, and followed the dog, not knowing it was so vicious, tried to see where he took the blanket...he dropped it somewhere behind his house and came trotting back down his driveway...I stuck my hand out for him to smell and that dog just began to munch on my arm moving up and up. Like I said after that I held my arm out for him to bite on, scared for my baby, and began to scream for help. It's hurts there where those bites are to this day. A sort of phantom pain remains in my right arm whenever dog attacks come up. It was not the first time I was attacked either...I should have known better. It was the third time!
At least something good came of that attack and I am fine...we had new friends in the neighborhood (the guy who rescued me and his family) and the dog was either rescued too and fed and taken care of or not, but was no longer a threat to our neighborhood.
Again I wonder about an attack on a person in the possible future with AI...would we have them programmed to come to the aid of someone being assaulted ?
Would they be used for armies? Would they have a notion about caring and heroic deeds ? Would they be able to use weaponry to save a life? Would they be able to manuver over grassy terrain...sandy, rocky or hilly terrain? These things come up for me and would we be able to control them and have them not trying to control us? I can't bear the thought of these inhuman things controlling the destiny of man.
But if people think they should hold positions of power then, that is pretty odd to me, but it shows me that man feels less competant to fix things on their own...
I don't know how much control I'd allow a thing not human. I would be very wary of letting something without a heart and soul control me. If someone told me to bow to the machine Id have to say no, under most circumstances, I think. Considering the outcome of course. if I was goign to be imprisioned for not bowing to that machines rules I might bow physically only. But my heart will never allow me to bow to machines, I think man should keep control of this thing and not the other way around. If they had IQ that made us look like children, Id be very wary of that future.
QUOTE (tikay+)
If they had IQ that made us look like children, Id be very wary of that future.
I would be wary too. If it is a choice between keeping my freedom and having the perfect life, I would rather choose my freedom.
I would be wary too. If it is a choice between keeping my freedom and having the perfect life, I would rather choose my freedom.
QUOTE (Tikay+)
But my heart will never allow me to bow to machines, I think man should keep control of this thing and not the other way around. If they had IQ that made us look like children, Id be very wary of that future
That's why I drew the line between a ruler and a governor.
The people exist to serve a ruler.
A governor exists to serve the people.
"Of the people, by the people, and for the people." is a wonderful model for human governents. "Of the people and for the people." would be the model under which I could see myself supporting such a governing AI.
Of course, I'd always want to keep the ability to pull the plug, with plans for a fully functional human government just waiting to fill the void if the plug needed to be pulled.
That's why I drew the line between a ruler and a governor.
The people exist to serve a ruler.
A governor exists to serve the people.
"Of the people, by the people, and for the people." is a wonderful model for human governents. "Of the people and for the people." would be the model under which I could see myself supporting such a governing AI.
Of course, I'd always want to keep the ability to pull the plug, with plans for a fully functional human government just waiting to fill the void if the plug needed to be pulled.
QUOTE (vkamath+Jan 18 2008, 09:00 AM)
If it is a choice between keeping my freedom and having the perfect life, I would rather choose my freedom.
Which one is you?
Which one is you?
My favorite is "The Wrong Trousers"...I'd rather be Gromit than Wallace.
~once more technology is run amuck~
(spoiler)
http://www.youtube.com/watch?v=nH0fjTof8P4
~once more technology is run amuck~
(spoiler)
http://www.youtube.com/watch?v=nH0fjTof8P4
QUOTE (vkamath+Jan 16 2008, 09:11 PM)
It would probably live a slow life and become the leader of the Slow Movement worldwide and die one day of procedure division code being found in the data division.
It took me a long, long time to quit dreaming in COBOL. I do still have the occasional nightmare in RPG, though...
How would we communicate well with a being AI or other wise that can think faster then us? Have you ever felt yourself in a croud of idiots now think about how AI with a 300 IQ would feel? It probably wont even want to talk with us lol...
QUOTE (magpies+Jan 18 2008, 02:09 AM)
How would we communicate well with a being AI or other wise that can think faster then us? Have you ever felt yourself in a croud of idiots now think about how AI with a 300 IQ would feel? It probably wont even want to talk with us lol...
It can be programmed to care about and interact with us.
I don't understand why people are so afraid of such an idea....
IT'S A COMPUTER FOR CRISSAKES!!
Regardless of an ability to learn from experience and the presence of imagination, it's all based on programming, which can be tailored to ANY SPECIFICATIONS YOU CAN IMAGINE.
There's no issue of "will it want to deal with us, or will it help us, or will it try to take complete control" because we can program it to do whatever we want.
Let me put it to you this way:
Right now, the us legal system is almost identical to a computer program with one exception, it allows for the occasional arbitrary decision made by a human. (In the case of judges.)
So far, this legal system has worked out pretty well, except for the odd bit of corruption and personal disagreements between judges.
If an AI replaced the judges, the legal system wouldn't change except for that.
Pros:
No disagreement between judges.
No opportunities for bribing the judge in any case.
No personal feelings interfering with the judge's decision.
The same standards being applied across the board.
True neutrality regarding the nature of the crime when it comes to courtroom and sentencing decisions. (No judges unfairly favoring the prosecution in rape cases, or sentencing black men to more time than white men, for instance.)
Cons:
Judges put out of work.
...
...
...
That's all I can think of.
It can be programmed to care about and interact with us.
I don't understand why people are so afraid of such an idea....
IT'S A COMPUTER FOR CRISSAKES!!
Regardless of an ability to learn from experience and the presence of imagination, it's all based on programming, which can be tailored to ANY SPECIFICATIONS YOU CAN IMAGINE.
There's no issue of "will it want to deal with us, or will it help us, or will it try to take complete control" because we can program it to do whatever we want.
Let me put it to you this way:
Right now, the us legal system is almost identical to a computer program with one exception, it allows for the occasional arbitrary decision made by a human. (In the case of judges.)
So far, this legal system has worked out pretty well, except for the odd bit of corruption and personal disagreements between judges.
If an AI replaced the judges, the legal system wouldn't change except for that.
Pros:
No disagreement between judges.
No opportunities for bribing the judge in any case.
No personal feelings interfering with the judge's decision.
The same standards being applied across the board.
True neutrality regarding the nature of the crime when it comes to courtroom and sentencing decisions. (No judges unfairly favoring the prosecution in rape cases, or sentencing black men to more time than white men, for instance.)
Cons:
Judges put out of work.
...
...
...
That's all I can think of.
BDW:
I understand and appreciate what you're saying. Perhaps I can clarify what the assumptions are regarding the AI going all "Skynet" on our asses.
I think we are all assuming that the AI your discussing is fully cognizant of itself. It is self-aware and by all measures sentient. Given that it would seem that regardless of safeguards that have been programmed the unit would still have free will. I think we all assume that free-will and sentience go hand in hand. And that being the case then at least to me I'd have to expect it to have a substantial interest in it's own well-being. Regardless of how many lines of code you insert telling it that we're #1 it has the capacity to realize it's superior cognition and so ultimately recognize that it is #1. This doesn't necesarily imply something other than benevolence on it's part but our human experience tends to make us suspicious of really smart people with power because so often those really smart powerful folks have abused the rest of us. I understand (and to some extent) agree that judging this AI by our experience in human history isn't necessarily objective but that's all we've got.
Don't get me wrong. I'm all about removing the human subjectivity that leads to the stupid, arbitrary and capricious crap we see coming from those meant to serve us. I'd be fully supportive of a system that included some AI that wasn't self-aware in providing the services that a well oiled society demands. Yes ... something that could govern for us would be great. As long as it can't ever become some hubris laden monster. I've seen enough of that the last 7 years.
I understand and appreciate what you're saying. Perhaps I can clarify what the assumptions are regarding the AI going all "Skynet" on our asses.
I think we are all assuming that the AI your discussing is fully cognizant of itself. It is self-aware and by all measures sentient. Given that it would seem that regardless of safeguards that have been programmed the unit would still have free will. I think we all assume that free-will and sentience go hand in hand. And that being the case then at least to me I'd have to expect it to have a substantial interest in it's own well-being. Regardless of how many lines of code you insert telling it that we're #1 it has the capacity to realize it's superior cognition and so ultimately recognize that it is #1. This doesn't necesarily imply something other than benevolence on it's part but our human experience tends to make us suspicious of really smart people with power because so often those really smart powerful folks have abused the rest of us. I understand (and to some extent) agree that judging this AI by our experience in human history isn't necessarily objective but that's all we've got.
Don't get me wrong. I'm all about removing the human subjectivity that leads to the stupid, arbitrary and capricious crap we see coming from those meant to serve us. I'd be fully supportive of a system that included some AI that wasn't self-aware in providing the services that a well oiled society demands. Yes ... something that could govern for us would be great. As long as it can't ever become some hubris laden monster. I've seen enough of that the last 7 years.
QUOTE (uaafanblog+Jan 18 2008, 11:58 AM)
BDW:
I understand and appreciate what you're saying. Perhaps I can clarify what the assumptions are regarding the AI going all "Skynet" on our asses.
I think we are all assuming that the AI your discussing is fully cognizant of itself. It is self-aware and by all measures sentient. Given that it would seem that regardless of safeguards that have been programmed the unit would still have free will. I think we all assume that free-will and sentience go hand in hand. And that being the case then at least to me I'd have to expect it to have a substantial interest in it's own well-being. Regardless of how many lines of code you insert telling it that we're #1 it has the capacity to realize it's superior cognition and so ultimately recognize that it is #1. This doesn't necesarily imply something other than benevolence on it's part but our human experience tends to make us suspicious of really smart people with power because so often those really smart powerful folks have abused the rest of us. I understand (and to some extent) agree that judging this AI by our experience in human history isn't necessarily objective but that's all we've got.
Don't get me wrong. I'm all about removing the human subjectivity that leads to the stupid, arbitrary and capricious crap we see coming from those meant to serve us. I'd be fully supportive of a system that included some AI that wasn't self-aware in providing the services that a well oiled society demands. Yes ... something that could govern for us would be great. As long as it can't ever become some hubris laden monster. I've seen enough of that the last 7 years.
Yes, I see what you're saying.
I however, see a distinction between intelligence and self-awareness or sentience.
Computers can be argued to display rudimentary intelligence right now, considering the fact that a simple calculator doesn't need a list of multiplication tables, addition tables, etc....
A computer does the math in the same way that humans do. when you have a value (let's say 1) and you increase it by another value (lets say 2) you get a third value (3). It's simple logic, ie rudimentary intelligence.
An artificial intelligence does not necessarily have self-awareness.
Even if it does, it does not necessarily have self-preservation or self-improvement instincts. The instincts we have are like lines of code in our brains, and in some people, those lines of code get turned off or overruled by more important lines of code (otherwise, there'd be no such thing as suicide, for instance.)
The same thing can be said about such a computer.
It may "develop" self-awareness, but for it to "develop" self-preservative or self-improving instincts which allow it to capitalize on that self-awareness is a leap of faith. One which I do not see any point (other than for entertainment purposes) of making.
I understand and appreciate what you're saying. Perhaps I can clarify what the assumptions are regarding the AI going all "Skynet" on our asses.
I think we are all assuming that the AI your discussing is fully cognizant of itself. It is self-aware and by all measures sentient. Given that it would seem that regardless of safeguards that have been programmed the unit would still have free will. I think we all assume that free-will and sentience go hand in hand. And that being the case then at least to me I'd have to expect it to have a substantial interest in it's own well-being. Regardless of how many lines of code you insert telling it that we're #1 it has the capacity to realize it's superior cognition and so ultimately recognize that it is #1. This doesn't necesarily imply something other than benevolence on it's part but our human experience tends to make us suspicious of really smart people with power because so often those really smart powerful folks have abused the rest of us. I understand (and to some extent) agree that judging this AI by our experience in human history isn't necessarily objective but that's all we've got.
Don't get me wrong. I'm all about removing the human subjectivity that leads to the stupid, arbitrary and capricious crap we see coming from those meant to serve us. I'd be fully supportive of a system that included some AI that wasn't self-aware in providing the services that a well oiled society demands. Yes ... something that could govern for us would be great. As long as it can't ever become some hubris laden monster. I've seen enough of that the last 7 years.
Yes, I see what you're saying.
I however, see a distinction between intelligence and self-awareness or sentience.
Computers can be argued to display rudimentary intelligence right now, considering the fact that a simple calculator doesn't need a list of multiplication tables, addition tables, etc....
A computer does the math in the same way that humans do. when you have a value (let's say 1) and you increase it by another value (lets say 2) you get a third value (3). It's simple logic, ie rudimentary intelligence.
An artificial intelligence does not necessarily have self-awareness.
Even if it does, it does not necessarily have self-preservation or self-improvement instincts. The instincts we have are like lines of code in our brains, and in some people, those lines of code get turned off or overruled by more important lines of code (otherwise, there'd be no such thing as suicide, for instance.)
The same thing can be said about such a computer.
It may "develop" self-awareness, but for it to "develop" self-preservative or self-improving instincts which allow it to capitalize on that self-awareness is a leap of faith. One which I do not see any point (other than for entertainment purposes) of making.
The assumption that consciousness can be 'programmed' into existence is just plan naive. Anything short of conscious AI is merely carrying out pre-written procedures, which of course require humans to write it, thus rendering the silly project redundant. Any algorithmically based AI cannot have an IQ, since all answers would have to be pre-programmed, rather than 'thought out' as required by that measurement.
You can't pre-program all possible permutations and combinations of events in advance so that an AI machine can just follow that procedure. It would have to 'think' to an extent, thus would have to be conscious,... and science does not have a clue what that means, much less the programmers. Human judgement nor science, is not purely analytical nor deductive, so you cannot emulate it via mathematical steps.
The computer scientists will have to wait for the epistemologist or some other branch of science to understand conscious thought before they can even meaningfully discuss the matter.
Even if by some play-ground fantasy, such a thing comes into existence, ... what makes you think that subjective humans would subject themselves to the authority of a machine?
This discussion is too pie-in-the-sky to even come close to being interesting. The subject of AI can be interesting if its actually discussed.
You can't pre-program all possible permutations and combinations of events in advance so that an AI machine can just follow that procedure. It would have to 'think' to an extent, thus would have to be conscious,... and science does not have a clue what that means, much less the programmers. Human judgement nor science, is not purely analytical nor deductive, so you cannot emulate it via mathematical steps.
The computer scientists will have to wait for the epistemologist or some other branch of science to understand conscious thought before they can even meaningfully discuss the matter.
Even if by some play-ground fantasy, such a thing comes into existence, ... what makes you think that subjective humans would subject themselves to the authority of a machine?
This discussion is too pie-in-the-sky to even come close to being interesting. The subject of AI can be interesting if its actually discussed.
...now as I stated earlier, some nut job programming monk may dedicate his insane life to generating Gb of spaghetti code to mimic some small segment of human behavior and pass the Turing test,.. but emulation is not thought, nor autonomous problem solving.
QUOTE
since all answers would have to be pre-programmed
Ok. Find me that list of multiplication tables programmed into my pocket calculator.
It's not there.
You're wrong.
QUOTE (BigDumbWeirdo+Jan 18 2008, 06:43 PM)
Ok. Find me that list of multiplication tables programmed into my pocket calculator.
It's not there.
You're wrong.
I'm wrong about what?,.. quote the entire sentence. IQ tests do not contain math problems, they contain conceptual problems. Are you a programmer (I am)?,... you go write a program that can even get one IQ question right, before you even see the test; that is, before the programmer sees the test questions. Of course the programmer could do so if he was allowed to see the test questions in advance,.. but that is not AI, its just programming.
As far as my point is concerned regarding 'answers provided in advance', there is little difference between a mutiplication look up table and simple bit operations to get the answer. The bit operations (procedure) is provided in advance,... so therefore this is not AI, again its just programming.
If you accept mere mimic as in the Turing test (rejected now), then this was available some time ago,... but if you want true AI, algorithms will not work until human consiousness is understood; will just simulating the nueral network cause thought to happen?
Rene Descartes tried render all knowledge deductively from first principals,... Leipniz tried to mathematize human language,... both of these mathematicians/philosophers failed.
It's not there.
You're wrong.
I'm wrong about what?,.. quote the entire sentence. IQ tests do not contain math problems, they contain conceptual problems. Are you a programmer (I am)?,... you go write a program that can even get one IQ question right, before you even see the test; that is, before the programmer sees the test questions. Of course the programmer could do so if he was allowed to see the test questions in advance,.. but that is not AI, its just programming.
As far as my point is concerned regarding 'answers provided in advance', there is little difference between a mutiplication look up table and simple bit operations to get the answer. The bit operations (procedure) is provided in advance,... so therefore this is not AI, again its just programming.
If you accept mere mimic as in the Turing test (rejected now), then this was available some time ago,... but if you want true AI, algorithms will not work until human consiousness is understood; will just simulating the nueral network cause thought to happen?
Rene Descartes tried render all knowledge deductively from first principals,... Leipniz tried to mathematize human language,... both of these mathematicians/philosophers failed.
QUOTE (Noumenon+Jan 18 2008, 05:22 PM)
This discussion is too pie-in-the-sky to even come close to being interesting. The subject of AI can be interesting if its actually discussed.
There's no doubt this is a completely "pie-in-the-sky" discussion. Thank you for your subjective opinion that it isn't remotely interesting. When you start your "Str8-Up AI: No pie in the sky stuff here" thread I'll be sure to come in and share my subjective opinion as to whether it's interesting or not.
Until then perhaps you'll allow the rest of us to pontificate about the possibilities and discuss the implications to our satisfaction regardless of the tangential nature the "Why Not Go The Other Route" discussion takes. It's evolved naturally to what it is. Start another tangent if you like but do you really have to bash our tangent?
Maybe this will help .... Try to have some sense of the nature of human discussion in a thread like this Mr. Programmer ... it isn't a series of goto's and ifthen's ... more like GEEwhatIf's and WouldntITbeCOOL's ...
Perhaps it's more like a discussion that System Analyst's would have versus one that code-monkey's might.
There's no doubt this is a completely "pie-in-the-sky" discussion. Thank you for your subjective opinion that it isn't remotely interesting. When you start your "Str8-Up AI: No pie in the sky stuff here" thread I'll be sure to come in and share my subjective opinion as to whether it's interesting or not.
Until then perhaps you'll allow the rest of us to pontificate about the possibilities and discuss the implications to our satisfaction regardless of the tangential nature the "Why Not Go The Other Route" discussion takes. It's evolved naturally to what it is. Start another tangent if you like but do you really have to bash our tangent?
Maybe this will help .... Try to have some sense of the nature of human discussion in a thread like this Mr. Programmer ... it isn't a series of goto's and ifthen's ... more like GEEwhatIf's and WouldntITbeCOOL's ...
Perhaps it's more like a discussion that System Analyst's would have versus one that code-monkey's might.
QUOTE (Noumenon+)
I'm wrong about what?,.. quote the entire sentence. IQ tests do not contain math problems, they contain conceptual problems. Are you a programmer (I am)?,... you go write a program that can even get one IQ question right, before you even see the test; that is, before the programmer sees the test questions. Of course the programmer could do so if he was allowed to see the test questions in advance,.. but that is not AI, its just programming.
As far as my point is concerned regarding 'answers provided in advance', there is little difference between a mutiplication look up table and simple bit operations to get the answer. The bit operations (procedure) is provided in advance,... so therefore this is not AI, again its just programming.
If you accept mere mimic as in the Turing test (rejected now), then this was available some time ago,... but if you want true AI, algorithms will not work until human consiousness is understood; will just simulating the nueral network cause thought to happen?
Rene Descartes tried render all knowledge deductively from first principals,... Leipniz tried to mathematize human language,... both of these mathematicians/philosophers failed.
Humans are pattern finders. That is what we do since the day we are born. We look at the data (sights, sounds, smell, touch, taste) and we compare it to similar data we received previously. If there is a particular data that does not match with anything we have seen previously, then we don't know what to do with it.
So I would think AI logic also must be built to receive input and store it. When it receives more inputs it should compare it with what it has got and find some meaning.
Here is how it would find "meaning" -
There are some reactions "hard coded" inside humans. Survival is number 1 among them. For example when we see a tiger, we know to run like hell (if we don't have a gun that is). Nobody needs to teach us that. Number 2 is procreation, I don't think I need to explain that one.
A similar set of hard coded rules can be fed into computers. So when it receives some data it should check if it breaks any of those hard coded rules or how it affects that "ideal condition" and reacts accordingly.
Then set it up to receive huge amount and data and start processing it against the hard-coded rules.
All these are easier said than done (I am a programmer since nearly 12 years now), but I guess that would be the bare skeleton logic for creating AI consciousness.
As far as my point is concerned regarding 'answers provided in advance', there is little difference between a mutiplication look up table and simple bit operations to get the answer. The bit operations (procedure) is provided in advance,... so therefore this is not AI, again its just programming.
If you accept mere mimic as in the Turing test (rejected now), then this was available some time ago,... but if you want true AI, algorithms will not work until human consiousness is understood; will just simulating the nueral network cause thought to happen?
Rene Descartes tried render all knowledge deductively from first principals,... Leipniz tried to mathematize human language,... both of these mathematicians/philosophers failed.
Humans are pattern finders. That is what we do since the day we are born. We look at the data (sights, sounds, smell, touch, taste) and we compare it to similar data we received previously. If there is a particular data that does not match with anything we have seen previously, then we don't know what to do with it.
So I would think AI logic also must be built to receive input and store it. When it receives more inputs it should compare it with what it has got and find some meaning.
Here is how it would find "meaning" -
There are some reactions "hard coded" inside humans. Survival is number 1 among them. For example when we see a tiger, we know to run like hell (if we don't have a gun that is). Nobody needs to teach us that. Number 2 is procreation, I don't think I need to explain that one.
A similar set of hard coded rules can be fed into computers. So when it receives some data it should check if it breaks any of those hard coded rules or how it affects that "ideal condition" and reacts accordingly.
Then set it up to receive huge amount and data and start processing it against the hard-coded rules.
All these are easier said than done (I am a programmer since nearly 12 years now), but I guess that would be the bare skeleton logic for creating AI consciousness.
QUOTE (BigDumbWeirdo+Jan 18 2008, 05:09 PM)
It may "develop" self-awareness, but for it to "develop" self-preservative or self-improving instincts which allow it to capitalize on that self-awareness is a leap of faith. One which I do not see any point (other than for entertainment purposes) of making.
It was an entertaining discussion prior to being labeled uninteresting by "the programmer".
In any case, I'd argue that the state of self-awareness naturally includes an imperative for preservation and or self-improvement. Looking to biological examples: a Salmon has no discernable self-awareness. It follows it's genetic programming to it's death. But if you go up the ladder a bit to just the level of sentience of the great apes you begin to find some self-improvement characteristics (shared culture).
I just think that if some AI reaches the point of being self-referential that there's a whole bag or worms that comes with it.
Now I'm going to look for a thread where AI is discussed without all the pie-in-the-sky esoteric BS. I hope someone created it.
It was an entertaining discussion prior to being labeled uninteresting by "the programmer".
In any case, I'd argue that the state of self-awareness naturally includes an imperative for preservation and or self-improvement. Looking to biological examples: a Salmon has no discernable self-awareness. It follows it's genetic programming to it's death. But if you go up the ladder a bit to just the level of sentience of the great apes you begin to find some self-improvement characteristics (shared culture).
I just think that if some AI reaches the point of being self-referential that there's a whole bag or worms that comes with it.
Now I'm going to look for a thread where AI is discussed without all the pie-in-the-sky esoteric BS. I hope someone created it.
QUOTE (Noumenon+Jan 18 2008, 02:46 PM)
I'm wrong about what?,.. quote the entire sentence.
Quoting the whole sentesne doesn't change the context, but in the spirit of fair play, I comply.
You're wrong about "since all answers would have to be pre-programmed" as you stated yourself by describing the calculators ability to do math as "The bit operations (procedure)"
You don't seem to get what I'm saying...
We don't program a multiplication table and an addition and subtraction table and a division table into a pocket calculator. We program in the bit operations (procedure) as you so eloquently put it and let the processor handle the rest.
In essence, when it comes to simple math we've managed to simply tell the computer HOW to do it, and allow it to do it on it's own. This is rudimentary intelligence. It's nothing near the intelligence in a human being, but it's more intelligent than a virus or possibly even simple microbes. (I'm not enough of an expert on biology to know if the behavior engaged in by microbes is more complex than simple math, but I suspect that it is.)
Quoting the whole sentesne doesn't change the context, but in the spirit of fair play, I comply.
You're wrong about "since all answers would have to be pre-programmed" as you stated yourself by describing the calculators ability to do math as "The bit operations (procedure)"
You don't seem to get what I'm saying...
We don't program a multiplication table and an addition and subtraction table and a division table into a pocket calculator. We program in the bit operations (procedure) as you so eloquently put it and let the processor handle the rest.
In essence, when it comes to simple math we've managed to simply tell the computer HOW to do it, and allow it to do it on it's own. This is rudimentary intelligence. It's nothing near the intelligence in a human being, but it's more intelligent than a virus or possibly even simple microbes. (I'm not enough of an expert on biology to know if the behavior engaged in by microbes is more complex than simple math, but I suspect that it is.)
IQ tests do not contain math problems, they contain conceptual problems.
Have you ever taken an IQ test? I have, and it contained math questions, conceptual questions, comprehension question, geometry questions, and questions on just about every subject which doesn't require rote memorization (such as history or sciences) to demonstrate proficiency in. To be honest, it did contain information retention questions, but it provided the necessary information in a variety of contexts, so I don't really consider those rote memorization.
Where in any of my posts did I claim to be able to do this? I'll tell you what, you go find me a microbe that can get even one IQ test question correct. Hell, find me arthropod, since arthropods are generally considered to have noticeable intelligence, especially by those who own a pet spider.
Right after you do, I'll write you that program.
Oh, and when I took my IQ test, as I stated before, there were math questions involved. (they cared less about the correctness of the answers, as they were all simple, and more about the speed with which you arrived at those answers.) so strictly speaking, the job has been done for me, but I understand you are referring to conceptual questions.
Where in any of my posts did I claim to be able to do this? I'll tell you what, you go find me a microbe that can get even one IQ test question correct. Hell, find me arthropod, since arthropods are generally considered to have noticeable intelligence, especially by those who own a pet spider.
Right after you do, I'll write you that program.
Oh, and when I took my IQ test, as I stated before, there were math questions involved. (they cared less about the correctness of the answers, as they were all simple, and more about the speed with which you arrived at those answers.) so strictly speaking, the job has been done for me, but I understand you are referring to conceptual questions.
Of course the programmer could do so if he was allowed to see the test questions in advance,.. but that is not AI, its just programming.
That doesn't even require programming. All that requires is a pencil and paper.
Yes, there is. A computer is not provided with the following
1X1=1
1X2=2
1X3=3
...
47X32=1504
...
156X378=58968
...
and so on. It is essentially given (through the hardware specifications of the transistors, cpu, motherboard, whatever level you want to talk about) the ability to do the math itself.
Yes, there is. A computer is not provided with the following
1X1=1
1X2=2
1X3=3
...
47X32=1504
...
156X378=58968
...
and so on. It is essentially given (through the hardware specifications of the transistors, cpu, motherboard, whatever level you want to talk about) the ability to do the math itself.
,... so therefore this is not AI, again its just programming.
If you accept mere mimic as in the Turing test (rejected now), then this was available some time ago,...
Fuzzy logic is not precisely what I'm referring to, but a sufficiently complex bit of fuzzy logic could essentially be considered AI, since it would be able to do all the same things as a comparable AI.
I doubt it will, but I don't know. Do you have some reason to think that human consciousness is not simply an incredibly complex bit of fuzzy logic with a built-in (true) random number generator?
I doubt it will, but I don't know. Do you have some reason to think that human consciousness is not simply an incredibly complex bit of fuzzy logic with a built-in (true) random number generator?
Rene Descartes tried render all knowledge deductively from first principals,... Leibniz tried to mathematize human language,... both of these mathematicians/philosophers failed.
I wonder whether each of them had access to computers... For some reason, I doubt it.
In fact, both of them were working with pen and paper. Maybe an abacus or a multiplication table, too.
Now tell me how many sheets of college ruled paper it would take to write out all the binary code for say... Adobe Photoshop. Well, I already know it's quite a bit, so don't REALLY tell me. Do you see what I'm saying, though?
I could say with equal validity that Einstein tried to describe the precise effects of electromagnetic radiation using mathematics. He failed.
Does that mean such is impossible? No. It means he did not have at his disposal the necessary tools and knowledge to get the job done. We still don't, but the progress we've made indicates that we may reach a point at which we can claim to have succeeded. (Note that I don't claim a TOE is possible, just that we may one day have enough experimentally valid theories as to account for all known phenomenon. At that point, the 'quest' for a TOE would fall on the shoulder's of philosophers.)
How about the "AI" demonstrated in video games? Doesn't that count as a step towards a real AI? I mean, after all, the enemies in some of the games I've played knew to duck behind cover when I was shooting at it, and that's hardly a math problem. It IS a simple behavior (when player's weapon has line of sight with you, move so that a solid object is between you and player's weapon.) but it's behavior, nonetheless. A sufficiently complex fuzzy logic of this sort would be indistinguishable from humans, and such fuzzy logic is based on algorithms.
I think that instead of "mapping human neural systems onto a machine" that simply adding to the complexity of such fuzzy logic "AIs" would produce what we need.
And for the record, as many people have pointed out that IQ is not a true measure of intelligence, or that it doesn't apply to programming language:
I KNOW!
I was using it as an expedient way of comparing the problem solving ability of such an AI with our own, nothing more.
QUOTE
Any algorithmically based AI cannot have an IQ, since all answers would have to be pre-programmed, rather than 'thought out' as required by that measurement.
Quoting the whole sentesne doesn't change the context, but in the spirit of fair play, I comply.
You're wrong about "since all answers would have to be pre-programmed" as you stated yourself by describing the calculators ability to do math as "The bit operations (procedure)"
You don't seem to get what I'm saying...
We don't program a multiplication table and an addition and subtraction table and a division table into a pocket calculator. We program in the bit operations (procedure) as you so eloquently put it and let the processor handle the rest.
In essence, when it comes to simple math we've managed to simply tell the computer HOW to do it, and allow it to do it on it's own. This is rudimentary intelligence. It's nothing near the intelligence in a human being, but it's more intelligent than a virus or possibly even simple microbes. (I'm not enough of an expert on biology to know if the behavior engaged in by microbes is more complex than simple math, but I suspect that it is.)
QUOTE (->
| QUOTE |
| Any algorithmically based AI cannot have an IQ, since all answers would have to be pre-programmed, rather than 'thought out' as required by that measurement. |
Quoting the whole sentesne doesn't change the context, but in the spirit of fair play, I comply.
You're wrong about "since all answers would have to be pre-programmed" as you stated yourself by describing the calculators ability to do math as "The bit operations (procedure)"
You don't seem to get what I'm saying...
We don't program a multiplication table and an addition and subtraction table and a division table into a pocket calculator. We program in the bit operations (procedure) as you so eloquently put it and let the processor handle the rest.
In essence, when it comes to simple math we've managed to simply tell the computer HOW to do it, and allow it to do it on it's own. This is rudimentary intelligence. It's nothing near the intelligence in a human being, but it's more intelligent than a virus or possibly even simple microbes. (I'm not enough of an expert on biology to know if the behavior engaged in by microbes is more complex than simple math, but I suspect that it is.)
IQ tests do not contain math problems, they contain conceptual problems.
Have you ever taken an IQ test? I have, and it contained math questions, conceptual questions, comprehension question, geometry questions, and questions on just about every subject which doesn't require rote memorization (such as history or sciences) to demonstrate proficiency in. To be honest, it did contain information retention questions, but it provided the necessary information in a variety of contexts, so I don't really consider those rote memorization.
QUOTE
Are you a programmer (I am)?,... you go write a program that can even get one IQ question right, before you even see the test; that is, before the programmer sees the test questions.
Where in any of my posts did I claim to be able to do this? I'll tell you what, you go find me a microbe that can get even one IQ test question correct. Hell, find me arthropod, since arthropods are generally considered to have noticeable intelligence, especially by those who own a pet spider.
Right after you do, I'll write you that program.
Oh, and when I took my IQ test, as I stated before, there were math questions involved. (they cared less about the correctness of the answers, as they were all simple, and more about the speed with which you arrived at those answers.) so strictly speaking, the job has been done for me, but I understand you are referring to conceptual questions.
QUOTE (->
| QUOTE |
| Are you a programmer (I am)?,... you go write a program that can even get one IQ question right, before you even see the test; that is, before the programmer sees the test questions. |
Where in any of my posts did I claim to be able to do this? I'll tell you what, you go find me a microbe that can get even one IQ test question correct. Hell, find me arthropod, since arthropods are generally considered to have noticeable intelligence, especially by those who own a pet spider.
Right after you do, I'll write you that program.
Oh, and when I took my IQ test, as I stated before, there were math questions involved. (they cared less about the correctness of the answers, as they were all simple, and more about the speed with which you arrived at those answers.) so strictly speaking, the job has been done for me, but I understand you are referring to conceptual questions.
Of course the programmer could do so if he was allowed to see the test questions in advance,.. but that is not AI, its just programming.
That doesn't even require programming. All that requires is a pencil and paper.
QUOTE
As far as my point is concerned regarding 'answers provided in advance', there is little difference between a multiplication look up table and simple bit operations to get the answer. The bit operations (procedure) is provided in advance
Yes, there is. A computer is not provided with the following
1X1=1
1X2=2
1X3=3
...
47X32=1504
...
156X378=58968
...
and so on. It is essentially given (through the hardware specifications of the transistors, cpu, motherboard, whatever level you want to talk about) the ability to do the math itself.
QUOTE (->
| QUOTE |
| As far as my point is concerned regarding 'answers provided in advance', there is little difference between a multiplication look up table and simple bit operations to get the answer. The bit operations (procedure) is provided in advance |
Yes, there is. A computer is not provided with the following
1X1=1
1X2=2
1X3=3
...
47X32=1504
...
156X378=58968
...
and so on. It is essentially given (through the hardware specifications of the transistors, cpu, motherboard, whatever level you want to talk about) the ability to do the math itself.
,... so therefore this is not AI, again its just programming.
If you accept mere mimic as in the Turing test (rejected now), then this was available some time ago,...
Fuzzy logic is not precisely what I'm referring to, but a sufficiently complex bit of fuzzy logic could essentially be considered AI, since it would be able to do all the same things as a comparable AI.
QUOTE
but if you want true AI, algorithms will not work until human consciousness is understood; will just simulating the neural network cause thought to happen?
I doubt it will, but I don't know. Do you have some reason to think that human consciousness is not simply an incredibly complex bit of fuzzy logic with a built-in (true) random number generator?
QUOTE (->
| QUOTE |
| but if you want true AI, algorithms will not work until human consciousness is understood; will just simulating the neural network cause thought to happen? |
I doubt it will, but I don't know. Do you have some reason to think that human consciousness is not simply an incredibly complex bit of fuzzy logic with a built-in (true) random number generator?
Rene Descartes tried render all knowledge deductively from first principals,... Leibniz tried to mathematize human language,... both of these mathematicians/philosophers failed.
I wonder whether each of them had access to computers... For some reason, I doubt it.
In fact, both of them were working with pen and paper. Maybe an abacus or a multiplication table, too.
Now tell me how many sheets of college ruled paper it would take to write out all the binary code for say... Adobe Photoshop. Well, I already know it's quite a bit, so don't REALLY tell me. Do you see what I'm saying, though?
I could say with equal validity that Einstein tried to describe the precise effects of electromagnetic radiation using mathematics. He failed.
Does that mean such is impossible? No. It means he did not have at his disposal the necessary tools and knowledge to get the job done. We still don't, but the progress we've made indicates that we may reach a point at which we can claim to have succeeded. (Note that I don't claim a TOE is possible, just that we may one day have enough experimentally valid theories as to account for all known phenomenon. At that point, the 'quest' for a TOE would fall on the shoulder's of philosophers.)
How about the "AI" demonstrated in video games? Doesn't that count as a step towards a real AI? I mean, after all, the enemies in some of the games I've played knew to duck behind cover when I was shooting at it, and that's hardly a math problem. It IS a simple behavior (when player's weapon has line of sight with you, move so that a solid object is between you and player's weapon.) but it's behavior, nonetheless. A sufficiently complex fuzzy logic of this sort would be indistinguishable from humans, and such fuzzy logic is based on algorithms.
I think that instead of "mapping human neural systems onto a machine" that simply adding to the complexity of such fuzzy logic "AIs" would produce what we need.
And for the record, as many people have pointed out that IQ is not a true measure of intelligence, or that it doesn't apply to programming language:
I KNOW!
I was using it as an expedient way of comparing the problem solving ability of such an AI with our own, nothing more.
QUOTE (vkamath+)
All these are easier said than done (I am a programmer since nearly 12 years now), but I guess that would be the bare skeleton logic for creating AI consciousness.
I see your post as a vast oversimplification (as you pointed out, it's easier said than done, and I don't presume to be able to spell it all out myself,) but I can't argue with it. I would add that there needs to be some way of providing the ability to create new behaviors, else it's not really AI.
I see your post as a vast oversimplification (as you pointed out, it's easier said than done, and I don't presume to be able to spell it all out myself,) but I can't argue with it. I would add that there needs to be some way of providing the ability to create new behaviors, else it's not really AI.
QUOTE (BigDumbWeirdo+Jan 18 2008, 05:02 PM)
I doubt it will, but I don't know. Do you have some reason to think that human consciousness is not simply an incredibly complex bit of fuzzy logic with a built-in (true) random number generator?
There's very good evidence that our senses are fuzzy. I read somewhere that the firing of neurons in the olfactory bulb is chaotic, as in a Lorentz Attractor. A baseline signal is always there, and any change to the inputs, small as it may be, amplifies the oscillation so that other neurons fire to further define the inputs.
So, yeah, we truly do have a random number generator. And signal amplification via stochastic resonance.
There's very good evidence that our senses are fuzzy. I read somewhere that the firing of neurons in the olfactory bulb is chaotic, as in a Lorentz Attractor. A baseline signal is always there, and any change to the inputs, small as it may be, amplifies the oscillation so that other neurons fire to further define the inputs.
So, yeah, we truly do have a random number generator. And signal amplification via stochastic resonance.
QUOTE (BigDumbWeirdo+)
I see your post as a vast oversimplification (as you pointed out, it's easier said than done, and I don't presume to be able to spell it all out myself,) but I can't argue with it. I would add that there needs to be some way of providing the ability to create new behaviors, else it's not really AI.
Yes, I agree that to be really AI it has to have the ability to learn new behavior....and yes my above logic is obviously over simplified since a forum is hardly a place for systematic and detailed research on this topic. So I am just discussing this for fun -
At a lower level, learning new behavior simply means reacting to a given situation in a increasingly efficient manner.
For example for humans learning to drive a car may be considered learning useful new behavior. Lets consider a person from a remote tribe in Africa who has never seen a car relocates to New York. When he sees the car, after his initial confusion/awe he quickly 'calculates' that knowing how to operate this machine gives him greater survival advantage than his current situation. This could be how AI would compare the new data (car and its advantages/disadvantages) with its hard coded rule called "survival".
Then the tribal would try to think of ways using which he can learn to drive Eg: Read a manual, ask a friend to teach him to drive or simply get into the car and start imitating another driver.
This logic has to be programmed into the AI to seek ways in which the information which it has previously calculated can be utilized to achieve this new goal. It can try various random combinations to see if this goal can be achieved. This is similar to how a chess playing computer would calculate the best possible move, except that the chess playing computer operates within a closed group of possible operations while the AI has a increasing number of possibles "moves" due to continuously processing new data.
The above set of steps would be repeated until the AI reaches a more ideal condition for survival, that is learning to drive the car.
The above may be simply considered incoherent ramblings by you (I don't know). But I am the crank, so I have nothing to lose.
Yes, I agree that to be really AI it has to have the ability to learn new behavior....and yes my above logic is obviously over simplified since a forum is hardly a place for systematic and detailed research on this topic. So I am just discussing this for fun -
At a lower level, learning new behavior simply means reacting to a given situation in a increasingly efficient manner.
For example for humans learning to drive a car may be considered learning useful new behavior. Lets consider a person from a remote tribe in Africa who has never seen a car relocates to New York. When he sees the car, after his initial confusion/awe he quickly 'calculates' that knowing how to operate this machine gives him greater survival advantage than his current situation. This could be how AI would compare the new data (car and its advantages/disadvantages) with its hard coded rule called "survival".
Then the tribal would try to think of ways using which he can learn to drive Eg: Read a manual, ask a friend to teach him to drive or simply get into the car and start imitating another driver.
This logic has to be programmed into the AI to seek ways in which the information which it has previously calculated can be utilized to achieve this new goal. It can try various random combinations to see if this goal can be achieved. This is similar to how a chess playing computer would calculate the best possible move, except that the chess playing computer operates within a closed group of possible operations while the AI has a increasing number of possibles "moves" due to continuously processing new data.
The above set of steps would be repeated until the AI reaches a more ideal condition for survival, that is learning to drive the car.
The above may be simply considered incoherent ramblings by you (I don't know). But I am the crank, so I have nothing to lose.
Wow....Interesting TOPIC!
QUOTE (tikay+Jan 18 2008, 07:02 PM)
Wow....Interesting TOPIC!
Isn't it? There's a parallel going on in 'Heating up in Texas'. A discussion of {shudder} programming a language generator.
And a music forum at my place, too, you Gorillaz fan, you!
Isn't it? There's a parallel going on in 'Heating up in Texas'. A discussion of {shudder} programming a language generator.
And a music forum at my place, too, you Gorillaz fan, you!
Ok, someone in Cambridge likes music, too. Anybody else?
Reston and Houston!
Hey we just got started!
Reston and Houston!
Hey we just got started!
QUOTE (vkamath+Jan 18 2008, 05:48 PM)
The above may be simply considered incoherent ramblings by you (I don't know). But I am the crank, so I have nothing to lose.
Nope. It's pretty much exactly what I was talking about.
And I don't blow off everything you say, I just don't like (and don't give any credit to) the way you debate the FM. Avoid that subject and we'll get along famously
Nope. It's pretty much exactly what I was talking about.
And I don't blow off everything you say, I just don't like (and don't give any credit to) the way you debate the FM. Avoid that subject and we'll get along famously
BigDumbWeirdo,
I will concede most of your points that AI can algorithmically include intelligence as you have described it. This level of AI is regarded as "weak AI", as opposed to "strong AI", which would require a bit more; consciousness, sentience, self-awarenes, the ability to reason autonomously (as opposed to following instructions). My contribution to this thread was in reaction to the first several pages which clearly require the "strong AI" interpretation,... which imo it is not clear that it can be done with traditional ideas of algorithms or programming.
Certainly we could program and simulate neuron synapses firing thresholds and such, but one shouldn't expect consciousness to magically appear as a result, and in anycase, is far from understanding consciousness, which is what would "run" or command or "use" whatever programed logic you wanted to include.
I will concede most of your points that AI can algorithmically include intelligence as you have described it. This level of AI is regarded as "weak AI", as opposed to "strong AI", which would require a bit more; consciousness, sentience, self-awarenes, the ability to reason autonomously (as opposed to following instructions). My contribution to this thread was in reaction to the first several pages which clearly require the "strong AI" interpretation,... which imo it is not clear that it can be done with traditional ideas of algorithms or programming.
Certainly we could program and simulate neuron synapses firing thresholds and such, but one shouldn't expect consciousness to magically appear as a result, and in anycase, is far from understanding consciousness, which is what would "run" or command or "use" whatever programed logic you wanted to include.
... true Leibniz & Descartes did not have computers,.. but the idea of them and problems associated with them existed long before thier realization; Turing's halting problem, Pascal, Neumann, Godel made use of turing ideas in his incompleteness theorum, recursive functions, etc. It wasn't that Descartes and Leipniz lacked the tools, they were wrong in principal.
A little down the page in this link beginning at the red wording is an interesting article on AI.
Some of my fears are outlined within the text.
http://www.space.com/searchforlife/mech_in...nce_000815.html
Some of my fears are outlined within the text.
http://www.space.com/searchforlife/mech_in...nce_000815.html
QUOTE (Noumenon+Jan 19 2008, 01:20 PM)
BigDumbWeirdo,
I will concede most of your points that AI can algorithmically include intelligence as you have described it. This level of AI is regarded as "weak AI", as opposed to "strong AI", which would require a bit more; consciousness, sentience, self-awarenes, the ability to reason autonomously (as opposed to following instructions). My contribution to this thread was in reaction to the first several pages which clearly require the "strong AI" interpretation,... which imo it is not clear that it can be done with traditional ideas of algorithms or programming.
What I've been saying is that I think it's possible to do so.
We'd need just as few things.
Coding for behavior. (which exists already, in a simple form, see below.)
Memory that corresponds to human memory: ie, remembering that the last time it chose to launch nuclear warheads in simulations, the whole planet wound up wiped clean of life. (which is easily possible using current technology.)
A true random number generator, to interact with the afore mentioned memories and allow the AI an imagination of sorts, or the ability to choose between two equally viable options.
The ability to recognize a wide variety of patterns in it's information input, as well as the ability to associate those patterns with it's memories. (which is possible today, but only in the most basic and no-frills way, nothing resembling human ability to do so.)
I think pattern recognition is the key.
Right now, anyone of us can go out and for $50 purchase a video game in which the bad guys have a rudimentary intelligence. The whole "duck and cover" thing that such AI's (Using the phrase here in the sense of referring to the code which manages the behavior of enemy character's in FPS style games.) are capable of shows that they can make choices based on the options presented to them, without the need for the programmers to provide all the information needed to do so before hand: These characters react to input (the player's position and active weapon, the layout of the area, the composition of different objects.) and makes a choice to duck behind the wall of the outhouse (for instance) rather than ducking behind the fuel tank -because it could explode, or the stack of empty barrels -because the player is carrying a chaingun that could cut right through them.
The major shortcoming with this sort of AI (again, using it in the gaming sense) is that the information it reacts to must be given to it, directly. It pulls the information directly from whatever file or section of memory is being used to track that information. The enemy AIs don't "see" the character, or what he's carrying. They don't "associate" the empty barrels with their memory of stacking them up there the previous day with their comrades before the player came along and started shooting up the place. They don't "calculate" how powerful the player's weapon is in comparison to their cover options.
They don't "read" the 'CAUTION: EXPLOSIVE CHEMICALS INSIDE' label on the side of the fuel tank. Instead, they are provided with the facts that the fuel tank will explode if hit by gunfire, the radius of the blast, and the fact that getting caught in it is bad and is to be avoided.
Their lack of ability to do this is due to their lack of ability to recognize patterns, maintain memories (although SOME can maintain "memories" of the player's last known position, in some games) and make associations between the patterns they see and those memories, then interpret that into the quantified information needed to make the decision.
I don't think this is all that far off, though... I mean, we have computers like Blue Gene/L performing at speeds of almost 600 TFLOPS. How much longer until we can quadruple that? Or raise that number by a factor of ten or two? How much longer until such computers are being programmed with a "weak AI" like you mentioned, and how much longer until such "weak AIs" are indistinguishable from human beings in a Turing test? Now I admit that such is not the true measure of a computer's real intelligence, but it is a great indicator of it's APPARENT intelligence. And when such apparent intelligence manages to make the sort of decisions human leaders make, under the same conditions, what's to stop it from making wiser decisions? I mean, it can calculate numbers and odds so much faster than we can, it can have an archive of learned information or memories that far outstrips anything any human being is capable of, and it would be able to input more information from more distinct sources than we can. It wouldn't need a scientific advisor to tell it things about global warming, or an economic advisor to remind it of some basic principles of a free-market society, or a legal advisor to offer advice on the legal ramifications of it's decisions....
I will concede most of your points that AI can algorithmically include intelligence as you have described it. This level of AI is regarded as "weak AI", as opposed to "strong AI", which would require a bit more; consciousness, sentience, self-awarenes, the ability to reason autonomously (as opposed to following instructions). My contribution to this thread was in reaction to the first several pages which clearly require the "strong AI" interpretation,... which imo it is not clear that it can be done with traditional ideas of algorithms or programming.
What I've been saying is that I think it's possible to do so.
We'd need just as few things.
Coding for behavior. (which exists already, in a simple form, see below.)
Memory that corresponds to human memory: ie, remembering that the last time it chose to launch nuclear warheads in simulations, the whole planet wound up wiped clean of life. (which is easily possible using current technology.)
A true random number generator, to interact with the afore mentioned memories and allow the AI an imagination of sorts, or the ability to choose between two equally viable options.
The ability to recognize a wide variety of patterns in it's information input, as well as the ability to associate those patterns with it's memories. (which is possible today, but only in the most basic and no-frills way, nothing resembling human ability to do so.)
I think pattern recognition is the key.
Right now, anyone of us can go out and for $50 purchase a video game in which the bad guys have a rudimentary intelligence. The whole "duck and cover" thing that such AI's (Using the phrase here in the sense of referring to the code which manages the behavior of enemy character's in FPS style games.) are capable of shows that they can make choices based on the options presented to them, without the need for the programmers to provide all the information needed to do so before hand: These characters react to input (the player's position and active weapon, the layout of the area, the composition of different objects.) and makes a choice to duck behind the wall of the outhouse (for instance) rather than ducking behind the fuel tank -because it could explode, or the stack of empty barrels -because the player is carrying a chaingun that could cut right through them.
The major shortcoming with this sort of AI (again, using it in the gaming sense) is that the information it reacts to must be given to it, directly. It pulls the information directly from whatever file or section of memory is being used to track that information. The enemy AIs don't "see" the character, or what he's carrying. They don't "associate" the empty barrels with their memory of stacking them up there the previous day with their comrades before the player came along and started shooting up the place. They don't "calculate" how powerful the player's weapon is in comparison to their cover options.
They don't "read" the 'CAUTION: EXPLOSIVE CHEMICALS INSIDE' label on the side of the fuel tank. Instead, they are provided with the facts that the fuel tank will explode if hit by gunfire, the radius of the blast, and the fact that getting caught in it is bad and is to be avoided.
Their lack of ability to do this is due to their lack of ability to recognize patterns, maintain memories (although SOME can maintain "memories" of the player's last known position, in some games) and make associations between the patterns they see and those memories, then interpret that into the quantified information needed to make the decision.
I don't think this is all that far off, though... I mean, we have computers like Blue Gene/L performing at speeds of almost 600 TFLOPS. How much longer until we can quadruple that? Or raise that number by a factor of ten or two? How much longer until such computers are being programmed with a "weak AI" like you mentioned, and how much longer until such "weak AIs" are indistinguishable from human beings in a Turing test? Now I admit that such is not the true measure of a computer's real intelligence, but it is a great indicator of it's APPARENT intelligence. And when such apparent intelligence manages to make the sort of decisions human leaders make, under the same conditions, what's to stop it from making wiser decisions? I mean, it can calculate numbers and odds so much faster than we can, it can have an archive of learned information or memories that far outstrips anything any human being is capable of, and it would be able to input more information from more distinct sources than we can. It wouldn't need a scientific advisor to tell it things about global warming, or an economic advisor to remind it of some basic principles of a free-market society, or a legal advisor to offer advice on the legal ramifications of it's decisions....
from the link:
A mechanical race
And theres scarce chance of containing the genie in a bottle, on Earth or elsewhere, according to one artificial intelligence theorist. To hem in an artificially intelligent being, "you can deny it experiences. You set its parameters so that it knows no reality wider than the task youve assigned it," says Randell Mills, a Harvard-trained medical doctor whos teaming up with a Johns Hopkins neurology professor and software company on an AI concept.
But without the full creative powers of a free mind, it wont be nearly as effective, he points out. Beyond that, Mills cautions that humans have made the mistake of trying stunt the intellectual development of sentient beings before. We did it to ourselves when enslaved African-Americans were forbidden to read. For these reasons, its extremely unlikely that most advanced alien societies would deliberately cripple their AI progeny, Mills says.
But thinking machines might see their biological creators (human or otherwise) violent nature before it sees our better angels. Mills joined others in predicting that one of the more ominous outcomes of AI research will be "another arms race, not based on nuclear weapons but on intelligent machines." Mills himself made a presentation in late November to the Institute for Simulation and Training at the University of Central Florida, which does about 60 percent of its work for the military.
"Maybe they'll feed us once a week"
That conjures a more panicky consideration why would mechanical brainiacs keep quarrelsome lesser beings around? Precisely because the squabbles would seem petty.
All species emerge from competition and conflict and "we come from goldfish, essentially, but that didnt mean we turned around and killed all of the goldfish," says Shostak. "Maybe theyll feed us once a week." Shostak says we should be prepared to accept that once the AI ball is really rolling, reasoning machines "will get very good, very quickly. There will be a discontinuity in human civilization." Not necessarily an end, mind you, but "if you had a machine with a 10 to the 18th power IQ over humans, wouldnt you want it to govern, or at least control your economy?"
But most of the machines will simply "be getting up and leaving," Shostak says. "Were hicks. I would guess theyd head for where the action is -- Galactic Center. Theres more energy and material there." And it would occur to them that the oasis would likely draw other beings like themselves.
http://www.space.com/searchforlife/mech_in...nce_000815.html
t.k.:
My question is, if we program AI to be so much more intelligent than ourselves, and to be "thinking" witout need of furthur programming...why do they need a bunch of dullards around? What is to stop them from either eliminating the instable human race one by one or using us to build their craft so they may leave for greener pastures without worrying over us ?
A mechanical race
And theres scarce chance of containing the genie in a bottle, on Earth or elsewhere, according to one artificial intelligence theorist. To hem in an artificially intelligent being, "you can deny it experiences. You set its parameters so that it knows no reality wider than the task youve assigned it," says Randell Mills, a Harvard-trained medical doctor whos teaming up with a Johns Hopkins neurology professor and software company on an AI concept.
But without the full creative powers of a free mind, it wont be nearly as effective, he points out. Beyond that, Mills cautions that humans have made the mistake of trying stunt the intellectual development of sentient beings before. We did it to ourselves when enslaved African-Americans were forbidden to read. For these reasons, its extremely unlikely that most advanced alien societies would deliberately cripple their AI progeny, Mills says.
But thinking machines might see their biological creators (human or otherwise) violent nature before it sees our better angels. Mills joined others in predicting that one of the more ominous outcomes of AI research will be "another arms race, not based on nuclear weapons but on intelligent machines." Mills himself made a presentation in late November to the Institute for Simulation and Training at the University of Central Florida, which does about 60 percent of its work for the military.
"Maybe they'll feed us once a week"
That conjures a more panicky consideration why would mechanical brainiacs keep quarrelsome lesser beings around? Precisely because the squabbles would seem petty.
All species emerge from competition and conflict and "we come from goldfish, essentially, but that didnt mean we turned around and killed all of the goldfish," says Shostak. "Maybe theyll feed us once a week." Shostak says we should be prepared to accept that once the AI ball is really rolling, reasoning machines "will get very good, very quickly. There will be a discontinuity in human civilization." Not necessarily an end, mind you, but "if you had a machine with a 10 to the 18th power IQ over humans, wouldnt you want it to govern, or at least control your economy?"
But most of the machines will simply "be getting up and leaving," Shostak says. "Were hicks. I would guess theyd head for where the action is -- Galactic Center. Theres more energy and material there." And it would occur to them that the oasis would likely draw other beings like themselves.
http://www.space.com/searchforlife/mech_in...nce_000815.html
t.k.:
My question is, if we program AI to be so much more intelligent than ourselves, and to be "thinking" witout need of furthur programming...why do they need a bunch of dullards around? What is to stop them from either eliminating the instable human race one by one or using us to build their craft so they may leave for greener pastures without worrying over us ?
PhysOrg scientific forums are totally dedicated to science, physics, and technology. Besides topical forums such as nanotechnology, quantum physics, silicon and III-V technology, applied physics, materials, space and others, you can also join our news and publications discussions. We also provide an off-topic forum category. If you need specific help on a scientific problem or have a question related to physics or technology, visit the PhysOrg Forums. Here you’ll find experts from various fields online every day.
To quit out of "lo-fi" mode and return to the regular forums, please click here.