Artificial Intelligence (aka AI)

[mhc]

has anyone else thought about this?:

extroverts, or perhaps non intuitive types will get a shock and rude awakening if and when true AI is ever invented.

one also has to wonder, why would it be advantages to actually have a true AI system in the first place or to what ends would it be helpful? is the purpose of it to understand non objective language, or to simply mimic human behaviour?

i see artificial intelligence as akin to buying a computer just because it seems novel, without actually having a use for it.

[ACow]

AI is more fantasy/religion than science at the moment, and not only because it is so far beyond our current abilities.

Not only is human "intelligence" a horribly flawed, erroneous, and poorly defined thing that we really shouldn't be trying to emulate in anything, but we pretty much value machines specifically because they do not possess the issues of "humanity" or "intelligence".

I've seen most people talk about AI more like they are talking about god/father/authority figures rather than talking about technology, and i'm yet to have heard a use for it that either isn't absolutely ridiculous (apart from maybe companions for retards/dementia patients?), isn't based on things inherently subjective (i want an AI to tell me how to best run my life/society), or which doesn't run into philosophical/practical problems of verifiability of its output (how do you know an AI's output is valid if it doesn't already align with your own beliefs?...which would then make it useless because we already have one of you and we don't really appreciate that intelligence either...).

[mhc]

i totally agree. even the examples we see of ai in tv shows, such as the computer on star trek (however, maybe by naming the system "computer" they are implying it is not an ai - or it could just be perhaps to help viewers distinguish who the characters are talking to) is not much more than an articulate computer system. the holodeck however, portrays characters that seem to have personality, but even that to is scripted in to the holodeck parameters.

again, i agree with your arguments and they align with much of the problems i see with the concept of ai. the use for retards or dementia patients seems quite practical tho i must say.

i would be interested to hear arguments by people in favour of the concept of ai, and what they think it would be able to accomplish, as i am of the opinion it would just have the effect of the "google syndrome", except it would be much worse. perhaps this is why google is actively pursuing ai, as a means of furthering their dictatorship

[P-O]

An A.I. is a problem solving device. Last time i checked the world still has problems to solve and an A.I. can potentially do it better than we can... Not to mention the technology required to build an A.I. is the same technology that will allow us to artificially augment our own intelligence.


For example:

You might ask the A.I. how to get to mars. While thinking about the problem, the A.I. completely redesigns the spaceship we're using; giving it new propulsion mechanisms and radically altering how we go about space travel.

Ask the A.I. how to stop or reverse global warming.

Ask the A.I. how to send information efficiently over large distances.

Just because we can't come up with the answer ourselves doesn't mean we can't check to see if it's correct.

[mhc]

An A.I. is a problem solving device.

technically im not sure if a problem solving device could constitute an artificial intelligence, but i understand the manner in which you are applying the principle.

Last time i checked the world still has problems to solve and an A.I. can potentially do it better than we can... Not to mention the technology required to build an A.I. is the same technology that will allow us to artificially augment our own intelligence.

hm, i'm not sure i agree with this. i believe creative problem solving to be the most effective at solving problems, and computer programs as they stand, would not be capable of anything close to simulating an imagination in order to behave creatively. logically passing a problem or situation over known parameters, or attempting to apply factual knowledge or information to a problem would be cumbersome at best, not to mention binary information processing is far less than ideal for processing the vast amounts of information that would be needed.

For example:

You might ask the A.I. how to get to mars. While thinking about the problem, the A.I. completely redesigns the spaceship we're using; giving it new propulsion mechanisms and radically altering how we go about space travel.

Ask the A.I. how to stop or reverse global warming.

Ask the A.I. how to send information efficiently over large distances.

Just because we can't come up with the answer ourselves doesn't mean we can't check to see if it's correct.

fine examples where the type of ai you have in mind would be not only practical but highly effective, however i stand by my imagination and creativity prerequisite and therefor problem. without a creative imagination, any ai would only be able to apply previously known information in a previously known manner. you could argue that it could be capable of self learning, but without prior loading of rules and conditions what would it seek out to learn? then would it still qualify as a true ai?

basically, i can see a type of computer system in star trek being possible at some point, but true artificial intelligence no. however, if technology was augmented with biological systems, maybe. by this i mean something like sustaining a brain in some technological containment and using it to process electrical signals from whatever technological device was connected to it. in this sense, the technology could stimulate some part of the brain to release feel good chemicals if a predefined situation was met, to "teach" it what was required or good from bad, thus enabling the brain, and therefor machine to "know" what was good or right. once the brain knew what was good or bad, it could then possibly gain emotions from events or situations and then perhaps display some kind of free will in avoidance or seeking out bad or good situations or outcomes

[ACow]

An A.I. is a problem solving device. Last time i checked the world still has problems to solve and an A.I. can potentially do it better than we can... Not to mention the technology required to build an A.I. is the same technology that will allow us to artificially augment our own intelligence.

For example:

You might ask the A.I. how to get to mars. While thinking about the problem, the A.I. completely redesigns the spaceship we're using; giving it new propulsion mechanisms and radically altering how we go about space travel.

Ask the A.I. how to stop or reverse global warming.

Ask the A.I. how to send information efficiently over large distances.

Just because we can't come up with the answer ourselves doesn't mean we can't check to see if it's correct.

That's more or less what I mean when I say fantasy.

AI becomes a sort of powerful wizard creature that acts as a popular framework onto which we project magical hopes and dreams for our technology/future and to save us from ourselves and our problems of today, even though we have no basis to believe such a thing can exist, how we would get there, or that its solutions wouldn't be absolutely banal (i.e. I have a solution for global warming right now, it doesn't mean you'll like it...)

/SF-plot-device: A.I, design for me a better A.I than you...repeat ad infinitum...

[mhc]

to save us from ourselves

thats what i meant by shock or rude awakening in the original post. most people dont realize their problems are self derived

[P-O]

That's more or less what I mean when I say fantasy.

Speculating about what a higher intelligence could invent is always going to be fantasy. If I'm a mouse thinking about how a human would go about getting the cheese it's not wrong to fantasize that the answer is outside of the normal reality that I'm used to.
If we can create an A.I. we can give it as much processing power as we want. The limits of the technology are unknown.

I'm not saying that our lives depend on A.I. It just seems to me that those would be reasonable uses for such a technology if it were developed.

and computer programs as they stand, would not be capable of anything close to simulating an imagination in order to behave creatively.

Right, but computer programs as they stand don't mimic a mind. If it can't think and theorize creatively, it's not useful as an A.I. That's the challenge.

Ultimately we want to mimic how the human brain works and up the processing power. That doesn't sound unreasonable to me. Brains exist; they operate according to normal physical principles.... I should be able to simulate it.

[P-O]

by this i mean something like sustaining a brain in some technological containment and using it to process electrical signals from whatever technological device was connected to it. in this sense, the technology could stimulate some part of the brain to release feel good chemicals if a predefined situation was met, to "teach" it what was required or good from bad, thus enabling the brain, and therefor machine to "know" what was good or right. once the brain knew what was good or bad, it could then possibly gain emotions from events or situations and then perhaps display some kind of free will in avoidance or seeking out bad or good situations or outcomes

You know, there's nothing magical about chemicals in the brain. These chemicals just inhibit or excite particular neurons. There's all these complicated on/off gates in the brain. These effects can be simulated as well, in principle.

[Architect]

extroverts, or perhaps non intuitive types will get a shock and rude awakening if and when true AI is ever invented.

No, why would that be? It does seem that ES types are the least interested in and believing of hard AI, but I believe that's just due to having an intuitive inferior function.

AI is more fantasy/religion than science at the moment, and not only because it is so far beyond our current abilities.

No it's not. We're having a hard time getting computers to act like humans exactly, but they have intelligence far exceeding ours already. Consider Watson which was able to read and use millions of page of the internet for it's answers. And these are subtle, inference type questions too, consider the answer “A long, tiresome speech delivered by a frothy pie topping,” to which Watson answered “What is a meringue harangue?”. Watson went on to easily beat the best Jeopardy humans in the world.

Not only is human "intelligence" a horribly flawed, erroneous, and poorly defined thing that we really shouldn't be trying to emulate in anything, but we pretty much value machines specifically because they do not possess the issues of "humanity" or "intelligence".

This seems a full non-sequetor from your previous statement, but I couldn't agree less. Human intelligence (why quote it? An overused tendency) brought us out of the mud slime to the modern world where we're able to communicate like this on the internet, in addition to knowing about the stars, the origin of the universe, and the DNA that makes us up. I personally wouldn't knock that.

The rest of your post seems to dribble off, but no we're specifically working on AI systems for important reasons. Such as medical diagnosis. Or semantic understanding of your searches. Presently searching is purely syntactic, and bringing semantic understanding, which seems to be at the center of Hard AI, would greatly enhance our lives.