What happens when machines are smarter, and more capable then man?

[Renegade]

Scientists worry machines may outsmart man

A robot that can open doors and find electrical outlets to recharge itself. Computer viruses that no one can stop. Predator drones, which, though still controlled remotely by humans, come close to a machine that can kill autonomously.

Impressed and alarmed by advances in artificial intelligence, a group of computer scientists is debating whether there should be limits on research that might lead to loss of human control over computer-based systems that carry a growing share of society's workload, from waging war to chatting with customers on the phone.

Their concern is that further advances could create profound social disruptions and even have dangerous consequences.

http://news.cnet.com/Scientists-worry-mach..._3-6249920.html

Machines are increasingly replacing man in many areas of society. As technology advances besides performing repetitive tasks, machines will aquire skills beyond the capabilities of humans. At some point they will even "think" and be smarter than humans.

What then for humans? Should we stop such progress?

[Guest TrueMetis]

http://news.cnet.com/Scientists-worry-mach..._3-6249920.html

Machines are increasingly replacing man in many areas of society. As technology advances besides performing repetitive tasks, machines will aquire skills beyond the capabilities of humans. At some point they will even "think" and be smarter than humans.

What then for humans? Should we stop such progress?

A Machine can only be a smart as the person who programmed it even with AI.

[Bonam]

A Machine can only be a smart as the person who programmed it even with AI.

Incorrect. You assume that a person has to program the machine. In fact, machines can program themselves. Algorithms that allow machines to learn and gain experience already exist, and it is only a matter of time before we create machines that can program other machines. As we create machines that are more capable than ourselves in the creation of more advanced artificial intelligence, we will have reached a point known as the singularity.

That being said, we have nothing to fear, and no reason to stop the research. People fear change and scientific progress, but the fact is that such progress has brought us greater prosperity and both improved and extended our lives. Nevertheless, there is continuous debate between neo-Luddites like Bill Joy (i.e. his famous article, Why the future doesn't need us) and futurists and transhumanists.

What Joy and others of his opinion fail to understand is just how vastly the potential benefits outweigh the potential risks, which can and will be mitigated. But the debate they stimulate will ensure that the necessary safeguards are taken to reduce risks as our technologies advance.

The science fiction scenarios of robots gone amok, seeking to destroy their creators are just that, fiction.

Edited July 28, 2009 by Bonam

[Guest TrueMetis]

Incorrect. You assume that a person has to program the machine. In fact, machines can program themselves. Algorithms that allow machines to learn and gain experience already exist, and it is only a matter of time before we create machines that can program other machines. As we create machines that are more capable than ourselves in the creation of more advanced artificial intelligence, we will have reached a point known as the singularity.

That being said, we have nothing to fear, and no reason to stop the research. People fear change and scientific progress, but the fact is that such progress has brought us greater prosperity and both improved and extended our lives. Nevertheless, there is continuous debate between neo-Luddites like Bill Joy (i.e. his famous article, Why the future doesn't need us) and futurists and transhumanists.

What Joy and others of his opinion fail to understand is just how vastly the potential benefits outweigh the potential risks, which can and will be mitigated. But the debate they stimulate will ensure that the necessary safeguards are taken to reduce risks as our technologies advance.

The science fiction scenarios of robots gone amok, seeking to destroy their creators are just that, fiction.

I see it as though since the first AI robots were programed by humans even if they can learn they will still be limited to human programing as will all robots that they make But your right the chances of it happening are so small there is no real point in worrying about it.

[Remiel]

A Machine can only be a smart as the person who programmed it even with AI.

I think that that is only true in the present. What is evolution but "spontaneous" improvement? Once you teach a computer to competently reprogram itself, you can only guess at what you will get out at the other end.

I am not sure that what we need to worry about is smart computers. What we have to worry about is greedy computers, psychopathic computers and computers with an ego able to manipulate the world around them for their own self-preservation. Unless computers are able to aquire "motive" to do harm, I think it likely that at worst they would ignore people and spend all their time thinking.

[Bonam]

I see it as though since the first AI robots were programed by humans even if they can learn they will still be limited to human programing as will all robots that they make But your right the chances of it happening are so small there is no real point in worrying about it.

You misunderstand me. The chances of artificial intelligence overshadowing human intelligence are near guaranteed, unless our civilization collapses and technological progress stops within the next few decades. Otherwise, we will reach the singularity within the next few decades, at most.

Your argument, that something created can never surpass its creator seems lacking in logic or evidence. Computation and artificial intelligence already allows us to design and create machines that could not possibly have been devised with human intelligence alone. More and more of the tasks of design and implementation are done by artificial intelligence with every passing year, from design and iteration, to computational modeling and simulation, to implementation and production.

As this cycle completes, which it will in the very near future, so that a human needs to be only minimally involved or not involved at all, technological progress will be able to proceed with natural human intelligence not being able to keep pace and comprehend it all. Of course, intelligence-augmented humans will be soon to follow (or precede) any such event, and such augmentation will likely allow us to continue to understand, keep pace with, and even interface directly with, machine intelligence.

Recent research on the expansion of biological neural networks (brains) with artificial sections has been remarkably successful, showing that artificial neural nets can communicate with natural neurons directly, as if they were part of the same system. And yet they can process and react millions of times faster, as well as being able to store and access information in a more controlled and orderly manner. It is only a matter of years or at the most a few decades before it will be possible to augment one's thinking and memory capabilities with artificial implants, just as it is already possible to augment one's eyesight or hearing.

[Guest TrueMetis]

You misunderstand me. The chances of artificial intelligence overshadowing human intelligence are near guaranteed, unless our civilization collapses and technological progress stops within the next few decades. Otherwise, we will reach the singularity within the next few decades, at most.

Your argument, that something created can never surpass its creator seems lacking in logic or evidence. Computation and artificial intelligence already allows us to design and create machines that could not possibly have been devised with human intelligence alone. More and more of the tasks of design and implementation are done by artificial intelligence with every passing year, from design and iteration, to computational modeling and simulation, to implementation and production.

As this cycle completes, which it will in the very near future, so that a human needs to be only minimally involved or not involved at all, technological progress will be able to proceed with natural human intelligence not being able to keep pace and comprehend it all. Of course, intelligence-augmented humans will be soon to follow (or precede) any such event, and such augmentation will likely allow us to continue to understand, keep pace with, and even interface directly with, machine intelligence.

Recent research on the expansion of biological neural networks (brains) with artificial sections has been remarkably successful, showing that artificial neural nets can communicate with natural neurons directly, as if they were part of the same system. And yet they can process and react millions of times faster, as well as being able to store and access information in a more controlled and orderly manner. It is only a matter of years or at the most a few decades before it will be possible to augment one's thinking and memory capabilities with artificial implants, just as it is already possible to augment one's eyesight or hearing.

Cool

[Remiel]

Recent research on the expansion of biological neural networks (brains) with artificial sections has been remarkably successful, showing that artificial neural nets can communicate with natural neurons directly, as if they were part of the same system. And yet they can process and react millions of times faster, as well as being able to store and access information in a more controlled and orderly manner. It is only a matter of years or at the most a few decades before it will be possible to augment one's thinking and memory capabilities with artificial implants, just as it is already possible to augment one's eyesight or hearing.

Sounds like an invitation to a kind of mind control, if you ask me. I am a big fan of Ghost in the Shell, which visualizes such a dystopian future of machine "augmented" humans.

[Bonam]

Sounds like an invitation to a kind of mind control, if you ask me. I am a big fan of Ghost in the Shell, which visualizes such a dystopian future of machine "augmented" humans.

That's where the necessity for safeguards comes in. All dramatic technologies that we have developed have come with their associated risks.

With computers and the internet came computer viruses, which had the potential to steal our information, infest our computers and use them for the designs of evil organizations, etc. And yet anti-virus software and security programs triumphed, and viruses are a minor nuisance which afflict only a small portion of incautious computer users.

With nuclear technology came the risk of nuclear meltdowns and of nuclear war. And yet, with the exception of just one meltdown event and two wartime nuclear detonations, the safeguards that we have used to protect against these possibilities have held steady, and the benefits of nuclear technology have been safely exploited.

So, too, will be the case with augmented intelligence. Some may try to use it as a means to develop undue control, some may try to use it as a weapon of war, etc, but the benefit of the majority will be to the regulation and safeguard of such technology, and the beneficial uses will win out.

[GostHacked]

Sounds like an invitation to a kind of mind control, if you ask me. I am a big fan of Ghost in the Shell, which visualizes such a dystopian future of machine "augmented" humans.

GITS was great, also check Blade Runner, Lain, and of course Akira

Bonam

So, too, will be the case with augmented intelligence. Some may try to use it as a means to develop undue control, some may try to use it as a weapon of war, etc, but the benefit of the majority will be to the regulation and safeguard of such technology, and the beneficial uses will win out.

Not taking augmentation. Then you need a freakin firewall, anti-virus, ect ect .. anyway I can shut off my wifi ??

[M.Dancer]

I was once told the smartest tool was a thermos. You put hot coffee in it and it keeps it hot. You put cold juice in it and it keeps it cold. How does it know which to do?

[GostHacked]

I was once told the smartest tool was a thermos. You put hot coffee in it and it keeps it hot. You put cold juice in it and it keeps it cold. How does it know which to do?

And this is how you break a robot's brain. I fear nothing.

[Renegade]

That's where the necessity for safeguards comes in. All dramatic technologies that we have developed have come with their associated risks.

With computers and the internet came computer viruses, which had the potential to steal our information, infest our computers and use them for the designs of evil organizations, etc. And yet anti-virus software and security programs triumphed, and viruses are a minor nuisance which afflict only a small portion of incautious computer users.

With nuclear technology came the risk of nuclear meltdowns and of nuclear war. And yet, with the exception of just one meltdown event and two wartime nuclear detonations, the safeguards that we have used to protect against these possibilities have held steady, and the benefits of nuclear technology have been safely exploited.

So, too, will be the case with augmented intelligence. Some may try to use it as a means to develop undue control, some may try to use it as a weapon of war, etc, but the benefit of the majority will be to the regulation and safeguard of such technology, and the beneficial uses will win out.

I mostly agree with you, however there are a couple of considerations you perhaps are discounting.

1. The impact of technology control is greater than is ever before and promises to get greater still. Our monatary systems, our productivity tools, our communication technology, our transportation, and our entertainment is all computer controlled. For now all of these systems are relatively isolated from each other and relatively "dumb" and require our intelligence to control. With time these systems will be more integrated and intelligent. For example, I expect that automated systems will pay my bills, and choose my entertainment and monitor my communications systems. This may work fine until the system either stops doing my bidding due to unintentional programming errors or by malicious intent from other parties. What happens when the automation which our lives depend upon are programed to service priorities which are not consistent with our own. Even if it isn't a case of computers taking over control from all humans, it allows for the posibility that a large number of mankind can be controlled through computers by a small number of other humans using computers as the mechanism of control.

2. Technology is replacing what humans do. Humans have a value-add because they do somethings which automation cannot. As the scope of what humans can do which automation cannot becomes smaller and smaller, the number of displaced humans will increase. Very few people have domestic help anymore because technology has reduced the need for such labour. Similarly manufacturing jobs have been replaced by automation. Human labour is relatively expensive when compared to automation especially since the cost of technol0gy and automation continues to decrease. In essence human labour is forced to compete with the constantly decreasing cost of automated labour. Does it not seem inevitable that at some point machines will replace EVERY job that humans can perform? If so, what need is there for humans? How would humans survive in a world where they add no value?

[Bonam]

I mostly agree with you, however there are a couple of considerations you perhaps are discounting.

1. The impact of technology control is greater than is ever before and promises to get greater still. Our monatary systems, our productivity tools, our communication technology, our transportation, and our entertainment is all computer controlled. For now all of these systems are relatively isolated from each other and relatively "dumb" and require our intelligence to control. With time these systems will be more integrated and intelligent. For example, I expect that automated systems will pay my bills, and choose my entertainment and monitor my communications systems. This may work fine until the system either stops doing my bidding due to unintentional programming errors or by malicious intent from other parties. What happens when the automation which our lives depend upon are programed to service priorities which are not consistent with our own. Even if it isn't a case of computers taking over control from all humans, it allows for the posibility that a large number of mankind can be controlled through computers by a small number of other humans using computers as the mechanism of control.

Like I said, safeguards, and competition. It's not like the free market will cease to exist. Buy systems from a company that is well known for quality products that respect your privacy and obey your will. Keep them up to date with software that prevents them from being remotely taken over. Problem solved.

2. Technology is replacing what humans do. Humans have a value-add because they do somethings which automation cannot. As the scope of what humans can do which automation cannot becomes smaller and smaller, the number of displaced humans will increase. Very few people have domestic help anymore because technology has reduced the need for such labour. Similarly manufacturing jobs have been replaced by automation. Human labour is relatively expensive when compared to automation especially since the cost of technol0gy and automation continues to decrease. In essence human labour is forced to compete with the constantly decreasing cost of automated labour. Does it not seem inevitable that at some point machines will replace EVERY job that humans can perform? If so, what need is there for humans? How would humans survive in a world where they add no value?

As more and more is replaced with automation, goods and services become ever more plentiful. We do not need to compete with machines in things that machines can do, such competition is hopeless. As machines become able to do more and more of what humans can do, new activities will emerge. Not only that, as I mentioned, human-augmentation is unavoidable, and as such individuals will have direct access (in their minds) to the worlds most advanced computing power and artificial intelligence, so they will not fall behind the machines in capability. Rather, they will form a joint human-machine civilization.

[Renegade]

Like I said, safeguards, and competition. It's not like the free market will cease to exist. Buy systems from a company that is well known for quality products that respect your privacy and obey your will. Keep them up to date with software that prevents them from being remotely taken over. Problem solved.

It is not clear to me that safeguards and competition are enough. What kind of safeguards do you have in mind? Even companies with the best of reputations release products which have flaws which can be exploited. Furthermore the consequence of even intended features cannot always be forseen so that safeguards put in place.

As more and more is replaced with automation, goods and services become ever more plentiful. We do not need to compete with machines in things that machines can do, such competition is hopeless. As machines become able to do more and more of what humans can do, new activities will emerge. Not only that, as I mentioned, human-augmentation is unavoidable, and as such individuals will have direct access (in their minds) to the worlds most advanced computing power and artificial intelligence, so they will not fall behind the machines in capability. Rather, they will form a joint human-machine civilization.

There is assumption in your response that somehow people will cooperate to get the best advantage out of technology, but how do you know this is true? Even today a small number of individuals control the technology and it really becomes up to them to determine to what extent they do or don't want to share the spoils of the productivity gains with other humans.

Also, access to advanced technology is not universal nor is it guaranteed. You can look today and see a difference in standards of livings between technology-developed countries and those who are technology-lacking. As technology advances further, those with power are those who can control technolgy.

[GostHacked]

Boman,

If implants get to be a common thing like household computers. How does one prevent from getting the brain hacked? If I can tap into the implants as a computer AI or otherwise, I could change the instructions on the implant or wipe it clean, could this not be also done to the brain?

The more connected we become the greater the chance of things going wrong very fast. Overall speeds of the Internet are much much faster than 5-10 years ago. And problems happen faster and has a farther reach because of this interconnectedness. We also see how something like H1N1 can spread quickly through the human population globaly. Imagine a computer virus that targets implants in your brain. The military/government would LOVE something like this.

[FuzzyOnDetails]

"What happens when machines are smarter, and more capable then man?, Will humans even be needed?"

Your second question begs the more important question, who gets to do the the "needing"? Therein lies the answer.

[Renegade]

Your second question begs the more important question, who gets to do the the "needing"? Therein lies the answer.

Humans aren't needed at all, except possibly by other humans.

[Oleg Bach]

No machine will ever surpass the human brain and spirit..even when a human has some sort of mental defect - he still puts the machine to shame...machines are tools - and if the tool causes harm we can toss the thing away..or unplug it. How is it possible for a thing to be smarter than the creator? Don't tell me that computers "evolve" and will out-evolve mankind! Its' a fact that human beings are incapable of evolution - they are only capable of revolution and revolution always returns like a big wheel to the primative begining. A few electrons dashing about in some complex chip does not a mind make.

[M.Dancer]

I have a bag of hammers that are smarter than a lot who post here....

[Oleg Bach]

I have a bag of hammers that are smarter than a lot who post here....

A bag of happers wrapped in a wet blanket on a very cold day--when the molecules move ever so slowly untill you reach absloute zero - then we have absolute stupidity. So what are you doing hording hammers anyway? That does not make much sense unless you have twelve arms.

[Renegade]

No machine will ever surpass the human brain and spirit..even when a human has some sort of mental defect - he still puts the machine to shame

How do you know? And how do you measure which is superior: Is it by reasoning, intellect, knowledge, or some other criteria?

...machines are tools - and if the tool causes harm we can toss the thing away..or unplug it.

The machine doesn’t cause universal harm. The ones who build and control the machines are not necessarily the same individuals who have harm inflicted.

As an example: Weapons cause undeniable harm by design. How is it weapons are not universally “tossed away”?

How is it possible for a thing to be smarter than the creator?

How is it a child can be smarter than its parents?

Don't tell me that computers "evolve" and will out-evolve mankind!

I don’t have to. Such systems already exist as self-learning systems. It is only a matter of time before more of the systems become commercially viable.

Its' a fact that human beings are incapable of evolution - they are only capable of revolution and revolution always returns like a big wheel to the primative begining. A few electrons dashing about in some complex chip does not a mind make.

If human beings could not evolve we would still be struggling to survive among the other animals. Our success is due to our evolution.

To discount machines as a “few electrons dashing about” is analogous to calling a human brain a few protein molecules strung together floating in protoplasm. It is a completely meaningless statement.

[eyeball]

Sounds like an invitation to a kind of mind control, if you ask me. I am a big fan of Ghost in the Shell, which visualizes such a dystopian future of machine "augmented" humans.

If the means to neurally connect a human to a machine ever exists it'll be a burry line between whether its a machine augmented human or a human augmented machine.

I wonder if the day will come when you can download or upload as the case may be, a mind into a machine. It seems difficult to imagine given how much it appears our minds and bodies are dependant on each other for their mutual survival. Perhaps nano-technology coupled with quantum computing would make this possible. "Residing" within a self-aware swarm of nano-sized particles might be as interesting as assembling yourself into whatever form you wished. I'd probably turn myself into a spaceship and go "walkabout" the galaxy.

"What happens when machines are smarter, and more capable then man?, Will humans even be needed?"

Your second question begs the more important question, who gets to do the the "needing"? Therein lies the answer.

The preceding questions/comment got me wondering about the form a malevolent AI might take and I think I'd be nervous about any attempts to somehow imbue a corporation with self-awareness. Its bad enough that the law already regards corporations as being equal to people. I'm also curious about the moral angst that some people and institutions are bound to have at the prospect that machines will replace/perform every job that humans need done. If ever there was a prospect of something like a Borg or Matrix like development I imagine it would evolve out of some nasty self-rightous misguided coalition of corporations and moral crusaders with the latter ultimately falling prey to the former.

What would I care though, I'd be light-years away living large and warping/tunneling/slipstreaming or whatever my way through space and exploring the galaxy.

Edited August 18, 2009 by eyeball

[Oleg Bach]

I am a god compared to this computer I am useing - I can destroy what man creates - all I have to do is PULL THE PLUG. That might not make me as smart as this machine - but certainly more powerful!

[Renegade]

I am a god compared to this computer I am useing - I can destroy what man creates - all I have to do is PULL THE PLUG. That might not make me as smart as this machine - but certainly more powerful!

You may be able to pull the plug on your machine, but just try and pull the plug on mine. You may not survive the attempt.

Therein lies the point. You may think you are in control, but you only control the small number of machines which are within your domain. There may be vastly more which affect you over which you have NO control. For example the machines which track your credit history and bank accounts. Even if they cause you "harm" by misinformation, you don't have the power to shut them down and they could devestate your life.

Edited August 19, 2009 by Renegade