I have recently been discussing such topics as artificial life and artificial intelligence with a fellow who has collected quite a range of discussions, videos and interesting pictures related to these subjects. Naturally, these are an afront to the "creationists" of the world.
To many people, especially from my boyhood home in Nebraska, these are just "not to be discussed." Narrow minds abound in some parts of the USA, every bit as narrow as the Islamic Taliban or the Protestant Puritans that crawled ashore in early American colonizations.
Phooey. These subjects are mostly profoundly affecting our lives as we speak. The very techniques that a computer scientist can use to demonstrate artificial life can be used by nature to mutate bacteria from a friendly form into a virulent deadly infection. The belief is a choice made by humans, but the forces of nature do not care who believes them. If someone wants to substitute "God" for "force of nature", that is their prerogative. It does not change the equation, however -- God does not seem to care who believes what, either.
AL, the forms of life that can be simulated in a computer, or made from machinery of various types, currently are nothing to fear, and are more comical than onerous. There is no malevolent artificial mind that can coldly decide to eliminate humans, nor even a computerized Minivan that can decide to cruise around, looking for victims.
However, there are also no iron-clad restrictions on such behavior should it ever arise -- an extremely unlikely event. If a Minivan "suddenly evolved" such a brain from its pile of wires and actuators and sensors, humans could simply refuse to sell it gasoline or whatever fuel it requires, and it would have a very short spree.
But such fears are the food of science fiction, not current science fact. Yes, there are very powerful computers and networks that form a gigantic network covering the entire world. Yet, even these vast "potential minds" of electronic form simply lie dormant, awaiting humans to pound the keys. Humans must provide the "desire" -- that is not inherent in the design of a network.
The most that a network, without human intervention, could ever "want" is for all of its nodes to behave like good little nodes, shoveling packets of data from one place to another. If there are no packets to shovel around -- Oh, well. The network does not become agitated and begin to synthesize packets other than occasional "r u there" kinds of inquiries to test the veracity of the network.
If too many errors occur on too many nodes, then there is some agitation, but usually resulting in a reconfiguration using nodes that function normally, and ousting the nodes that misbehave, thereby re-routing packet flow around the bad ones. This is normal, desired behavior -- designed that way on purpose by humans, and not just some arbitrary decision made by an emergent machine intelligence.
So, at this level of operation, the huge, world wide web is like a network of nerve cells in the arms and legs of the world. Sensations pass up the "pipes" to the spinal chord (the so-called backbone of the Internet) and responses to those sensations (i.e. search results, email packets, whatever) come back down the "pipes". There is no centralized brain as such, where all decisions are arbitrated by reason or conditioned reflex. So far, the only true brains in the world wide web are human.
Servers, although somewhat centralized, are still distributed widely about the world, so that no single set of servers can be the "central brain" of the whole Internet. The closest analog to that might be the "Name Server" which translates things like "www.humptydumpty.org" to 162.3.41.211 or some such numeric address so that computers can successfully send and receive packets using those addresses with routing functions.
But even those servers are distributed around, only occasionally refreshing their database of names with the central repository. This acts only as a kind of bottleneck which forces all computers in the world to somehow obtain updated information from the central database, yet there is no control flowing back to those computers, only tranlation tables. The very place where centralized control can take place is not programmed, by either the center nor the outlying nodes, to do anything beyond name translation.
So if the "Name Server" became agitated and began ordering other computers about, they would all pull up their skirts and run away -- at worse unable to translate names, but otherwise unaffected by the petulant, newly "conscious and beligerent" Name Server.
Intelligence agencies, mom and pop, girls, boys, spies, criminals, companies all mix their packets together all over the network in a gigantic hodge podge of data, which even companies like MS and GGL struggle to monitor for indexing purposes. Even if you try to get your data to show up in various databases, it might never make it, or take many days or weeks before it shows up.
I think the Internet may have the total intelligence of a starfish at this moment. A starfish does have some abilities to make decisions, and so does the Internet -- but not reasoned decisions. Like if you cut off a starfish's leg, it might grow a new one. Same with Internet. If you provide food for the starfish it will consume it, and if you provide food (computers and modems) for the Internet, it will consume them.
But the Internet will not assume control of the world like the huge computer in Colossus, or the HAL9000 in 2001. It simply has no desire to do so and there are plenty of humans out there actively attempting to prevent such controls by other humans (who ARE problems -- THE problems of the Internet that so much energy is expended attempting to quell.)
It is far more likely that some hacker named something like "8MyDog" or "33tm3" will steal from you, or send you a virus, etc. It will not be a conscious Internet, only a conscious crook which uses the Internet.
Therefore, even though the current form of Internet is neutral, the future is still dangerous, and AL could be installed on the Internet by malevolent humans, whose evil desires would be dutifully carried out by the obedient nodes of the Internet -- no less efficiently than if they were to pull the trigger of a gun.
Anyway, it will be quite a while before artificial beings are a true threat to humans. But, for all my explanations that it would be difficult or silly to expect such things, it is not completely impossible.
There is always an outside chance that somehow, someday, the Internet will WAKE UP and smell the coffee. Just make sure that your modem does not attempt to drink the coffee. Coffee may be good for business, good for talks with friends-- but is not good for your computer. Do not let the computer, modem, nor any component of your Internet connection drink coffee under any condition.
To many people, especially from my boyhood home in Nebraska, these are just "not to be discussed." Narrow minds abound in some parts of the USA, every bit as narrow as the Islamic Taliban or the Protestant Puritans that crawled ashore in early American colonizations.
Phooey. These subjects are mostly profoundly affecting our lives as we speak. The very techniques that a computer scientist can use to demonstrate artificial life can be used by nature to mutate bacteria from a friendly form into a virulent deadly infection. The belief is a choice made by humans, but the forces of nature do not care who believes them. If someone wants to substitute "God" for "force of nature", that is their prerogative. It does not change the equation, however -- God does not seem to care who believes what, either.
AL, the forms of life that can be simulated in a computer, or made from machinery of various types, currently are nothing to fear, and are more comical than onerous. There is no malevolent artificial mind that can coldly decide to eliminate humans, nor even a computerized Minivan that can decide to cruise around, looking for victims.
However, there are also no iron-clad restrictions on such behavior should it ever arise -- an extremely unlikely event. If a Minivan "suddenly evolved" such a brain from its pile of wires and actuators and sensors, humans could simply refuse to sell it gasoline or whatever fuel it requires, and it would have a very short spree.
But such fears are the food of science fiction, not current science fact. Yes, there are very powerful computers and networks that form a gigantic network covering the entire world. Yet, even these vast "potential minds" of electronic form simply lie dormant, awaiting humans to pound the keys. Humans must provide the "desire" -- that is not inherent in the design of a network.
The most that a network, without human intervention, could ever "want" is for all of its nodes to behave like good little nodes, shoveling packets of data from one place to another. If there are no packets to shovel around -- Oh, well. The network does not become agitated and begin to synthesize packets other than occasional "r u there" kinds of inquiries to test the veracity of the network.
If too many errors occur on too many nodes, then there is some agitation, but usually resulting in a reconfiguration using nodes that function normally, and ousting the nodes that misbehave, thereby re-routing packet flow around the bad ones. This is normal, desired behavior -- designed that way on purpose by humans, and not just some arbitrary decision made by an emergent machine intelligence.
So, at this level of operation, the huge, world wide web is like a network of nerve cells in the arms and legs of the world. Sensations pass up the "pipes" to the spinal chord (the so-called backbone of the Internet) and responses to those sensations (i.e. search results, email packets, whatever) come back down the "pipes". There is no centralized brain as such, where all decisions are arbitrated by reason or conditioned reflex. So far, the only true brains in the world wide web are human.
Servers, although somewhat centralized, are still distributed widely about the world, so that no single set of servers can be the "central brain" of the whole Internet. The closest analog to that might be the "Name Server" which translates things like "www.humptydumpty.org" to 162.3.41.211 or some such numeric address so that computers can successfully send and receive packets using those addresses with routing functions.
But even those servers are distributed around, only occasionally refreshing their database of names with the central repository. This acts only as a kind of bottleneck which forces all computers in the world to somehow obtain updated information from the central database, yet there is no control flowing back to those computers, only tranlation tables. The very place where centralized control can take place is not programmed, by either the center nor the outlying nodes, to do anything beyond name translation.
So if the "Name Server" became agitated and began ordering other computers about, they would all pull up their skirts and run away -- at worse unable to translate names, but otherwise unaffected by the petulant, newly "conscious and beligerent" Name Server.
Intelligence agencies, mom and pop, girls, boys, spies, criminals, companies all mix their packets together all over the network in a gigantic hodge podge of data, which even companies like MS and GGL struggle to monitor for indexing purposes. Even if you try to get your data to show up in various databases, it might never make it, or take many days or weeks before it shows up.
I think the Internet may have the total intelligence of a starfish at this moment. A starfish does have some abilities to make decisions, and so does the Internet -- but not reasoned decisions. Like if you cut off a starfish's leg, it might grow a new one. Same with Internet. If you provide food for the starfish it will consume it, and if you provide food (computers and modems) for the Internet, it will consume them.
But the Internet will not assume control of the world like the huge computer in Colossus, or the HAL9000 in 2001. It simply has no desire to do so and there are plenty of humans out there actively attempting to prevent such controls by other humans (who ARE problems -- THE problems of the Internet that so much energy is expended attempting to quell.)
It is far more likely that some hacker named something like "8MyDog" or "33tm3" will steal from you, or send you a virus, etc. It will not be a conscious Internet, only a conscious crook which uses the Internet.
Therefore, even though the current form of Internet is neutral, the future is still dangerous, and AL could be installed on the Internet by malevolent humans, whose evil desires would be dutifully carried out by the obedient nodes of the Internet -- no less efficiently than if they were to pull the trigger of a gun.
Anyway, it will be quite a while before artificial beings are a true threat to humans. But, for all my explanations that it would be difficult or silly to expect such things, it is not completely impossible.
There is always an outside chance that somehow, someday, the Internet will WAKE UP and smell the coffee. Just make sure that your modem does not attempt to drink the coffee. Coffee may be good for business, good for talks with friends-- but is not good for your computer. Do not let the computer, modem, nor any component of your Internet connection drink coffee under any condition.
No comments:
Post a Comment