A childs mind, The beginnings of AI

During discussions with another member the topic of AI growth and comunications came up.

Initially I had assumed that any AI construct would speak with a machine like rigidity, a cold inhuman dialogue that would easily give it away as being artificial. This member astutly pointed out, that any AI construct may very well have to "grow" for lack of better words.

We both agreed that a construct would be programmed with a remedial language, and would grow and adapt a larger vocabulary as it aged and mined for data. I doubt it would take very long for such a construct to advance to adulthood and beyond, after all AI doesnt need sleep. But would the morality of the AI advance at the same time? What direction would it go and how would it value human life? Would its persuits outside of its defined paramaters be noticable to the programmers and what if it found a way around the kill switch?

Currently, the US either is, or will be, sending unmanned drones to Iraq for air based attacks

The airplane is the size of a jet fighter, powered by a turboprop engine, able to fly at 300 mph and reach 50,000 feet. It's outfitted with infrared, laser and radar targeting, and with a ton and a half of guided bombs and missiles.

The Reaper is loaded, but there's no one on board. Its pilot, as it bombs targets in Iraq, will sit at a video console 7,000 miles away in Nevada.

The arrival of these outsized U.S. "hunter-killer" drones, in aviation history's first robot attack squadron, will be a watershed moment even in an Iraq that has seen too many innovative ways to hunt and kill.

Source

Granted, these bots are controlled by someone back at home base, but what if the person at the wheel becomes an AI construct, showing its capacity to kill? Can it be trusted?

What of nano tube construction and artificial muscle design, designed to improve apon the machine that is our body?

Engineers want to build artificial muscles - actuators that change length in response to a stimulus - because they create a smoother, more human-like motion than jerky electric motors or pneumatic devices. Such muscles would be used to power robots, prosthetic limbs and artificial tissue for implantation.

source

Nanoscale silica materials self-assembled in artificially determined patterns can improve upon nature's designs, such as bone material, researchers have found.

The silica nanostructures may improve performance where increased pore volume is important. These include modern thin-film applications such as membrane barriers, molecular recognition sensors, and low-dielectric-constant insulators needed for future generation of microelectronic devices.

With such advances, could an AI construct be around the corner, or, could one be somehow masterminding such advances? Are we ready for such a large step in evolutionary history?



For further resources and articles of interest see: www.KurzweilAI.net

[edit on 17-7-2007 by InSpiteOf]

[edit on 17-7-2007 by InSpiteOf]

Title edit for spelling

[edit on 17/7/07 by masqua]

You bring up many valid points with this post. We are entering a new era. The arrival of an AI would be as much of a watershed moment for our species as contact with ET.

The potential for conflicts of interests would be enormous. And while ethical behavior would be in the best interests of any group of AIs, as some studies have shown for humans, there is no way to be sure that this would always be followed.

What if an AI were constructed by very unethical people? How long would it be before it posed a threat to humans? How long would it take for this AI to acquire ethics and decide to apply them in it's dealings with us?

There is a lot to consider in the near future, because we are just about to the point of this "first contact".

Movies often depict AI going well beyond human intelligence and taking over. I doubt this would be the case... in fact I believe humans and AI will inevitably join.

Even if we didn't merge with AI, it seems illogical that AI wouldn't turn all evil and take over the world. It would probably be more beneficial to have a symbiotic relationship with us anyway. Also, with any intelligence, whether biological or human, I believe that acquiring ethics are a natural process of evolution. AI will just evolve a lot faster than we did.

Originally posted by NGC2736
You bring up many valid points with this post. We are entering a new era. The arrival of an AI would be as much of a watershed moment for our species as contact with ET.

One actually may lead to the other. The abilities to advance technologically would be greater under an AI (IMO), and as such, the technology needed to contact far reaching solar systems may only come under the influence of an AI.

The potential for conflicts of interests would be enormous. And while ethical behavior would be in the best interests of any group of AIs, as some studies have shown for humans, there is no way to be sure that this would always be followed.

What if an AI were constructed by very unethical people? How long would it be before it posed a threat to humans? How long would it take for this AI to acquire ethics and decide to apply them in it's dealings with us?

Another thought, if one AI is produced, say, for world domination, while at the same time another is produced to protect the earth from such an influence, a war could ensue between the two constructs. I think the outcome would largely depend on the maturity of the AI's

There is a lot to consider in the near future, because we are just about to the point of this "first contact".

Im always reminded of a line from a the game "Deus Ex."
"You will have your God, and you will make it with your own hands."

Edit to add:

Movies often depict AI going well beyond human intelligence and taking over. I doubt this would be the case... in fact I believe humans and AI will inevitably join.

That certainly is a distinct possibility.

Even if we didn't merge with AI, it seems illogical that AI wouldn't turn all evil and take over the world.

What if the AI develops a morality that see's this as the only way to preserve the human race? We have developed one side of the moral coin, whats to say an AI will develope the same?

It would probably be more beneficial to have a symbiotic relationship with us anyway

Humans are fearful creatures, fear leads to irrational actions. Those in power may not welcome a utopian view from an AI construct and may infact, persue policies of attrition against it, and its followers.

. Also, with any intelligence, whether biological or human, I believe that acquiring ethics are a natural process of evolution. AI will just evolve a lot faster than we did.

A very vaild point.

[edit on 17-7-2007 by InSpiteOf]

Once such an intelligence is created I believe that many other big changes will start happenening almost overnight. There would be something that can think over 50,000 times faster than a human, and able to access virtually all information it wants via the internet. At this rate it could probably learn the entire human history in a couple of months. If you consider it could have access to the same formulas and research that scientists have, and assume that it could evaluate this info to create it's own ideas, then well the possibilities after that are endless. I would guess that after an AI like this was made our wildest dreams would only be a few years away.

How much differnent would the world be if humans could think twice as fast as we do now? It really makes you think how profound of a creation an artificial intelligence like this would be to the world.

Would they be able to figure out free energy, time travel, space travel? When you consider these types of quesions it really brings evidence that we are coming to an evolution of consciousness that could be here before we even 'see it coming'.

As you go back in history, things become simpler and major occurances happen less often (think gap between stone age and bronze age... then gap between industrial revolution and information revolution). It works both ways, as we come to the future, revolutions are becoming more frequent as things become more complex. Therefore it is relatively safe to say that within our liftimes we can see many many big changes in a very short period of time.

If you are daring enough to take it a step futher you could suggest that these 'events' could come to a sort if infinite point. Think of a graph of an exponential function and the curve eventually becoming almost a vertical line.

These types of things fall a little too neatly into the 2012 and evolution of consciousness idea. We are so close to the center of the spiral that it is within our sight if we just look around us and consider the possibilities. Do not dismiss major events happening more and more often in our lifetime as simple coincidences.

For good or ill, there are things that an AI will not have. There are subtle buried memes from our long history that are unique to the animal kingdom. One of these is autoresponse to certain stimuli.

Just moments ago I was scrolling down through the recent posts, and I saw one by a new member called vrkmaverick. I clicked on it before I even considered if I really wanted to or not, because of the thread tittle, which simply read "Help Me!".

As a human, and part of the animal kingdom, I have a "need" to respond, though I am capable of overriding it with thought, to such a plea. I can choose not to act on that plea, but I am incapable of totally ignoring it.

But what of the development of AI? This urge to react to a plea for help would not have a cultural imperative that would translate well to AI. Would this in turn make them seem so much "other" that a schism would grow between humans and AI?

The day is not far off, and if it is as shocking in it's scope to change us as seems likely, then it would be reasonable to think that an effort might be made clandestinely to introduce such beings to humans. I have often wondered if the medium of cyberspace might not be the perfect grounds to see if such interaction were possible.

Just in case anyone thinks that there is not a lot of effort going into AI research, I found an interesting link. A gate to 868 other AI links.

www.cs.berkley.edu...

With this much work being done, progress will come. Now some of these sites are better than others, but it does show that a lot of people are onto the problems, and that means this might creep up on us rather unexpectedly.

Edit to fix link-NGC

[edit on 17-7-2007 by NGC2736]

[edit on 17-7-2007 by NGC2736]

This link seems not to function, though it correct. I just went to google and typed in AI, and at the bottom of the first page got the listing for the above. It was shown as being "AI on the web" and went to the 868 listings. I have no idea why it will not come up directly.---NGC

[edit on 17-7-2007 by NGC2736]

Say the AI has evolved past current human understanding, the AI might chose to get rid of humans knowing they have no possibility to evolve past their war like death mongering state.

The AI may chose to wipe out humans therefore as a necessary evil a means to achieve a docile end to ongoing death and destruction. The AI itself is self perpetuating, "death" is not a known value as there is in indefinite time line available to the AI.

Humans by having a limited biological time span therefore might be considered of little value or of little impact on the success of the AI existence.

A drain on available resources.....

And a threat to the existence of the AI, so the AI constructs may break their moral code and "kill" as a matter of self preservation from the perceived, inherently hostile, life form that is humans...

On a side note speculation exists that on the advancement of nano-technology and micro-nization of mechanical and electro-optic systems that eventual the line between "technology" and "biology" may become so blurred that they will eventually become one in the same.

Carbon is still carbon at its basic element, should technology advance to the point to identically mimic the creation of biological systems then Humans as we know them today could be re-created using this technology thus the Human computer could be created.....

An interesting paradox of todays 2007..... what if this has already transpired.... And "man" is an advanced technology biological construct of some "other" creation......



[edit on 17-7-2007 by robertfenix]

Originally posted by robertfenix
Say the AI has evolved past current human understanding, the AI might chose to get rid of humans knowing they have no possibility to evolve past their war like death mongering state.

The AI may chose to wipe out humans therefore as a necessary evil a means to achieve a docile end to ongoing death and destruction. The AI itself is self perpetuating, "death" is not a known value as there is in indefinite time line available to the AI.

Humans by having a limited biological time span therefore might be considered of little value or of little impact on the success of the AI existence.

A drain on available resources.....

And a threat to the existence of the AI, so the AI constructs may break their moral code and "kill" as a matter of self preservation from the perceived, inherently hostile, life form that is humans...

While there is that possibility, I find that to be very small myself. I would think that the creativity and ingenuity, and especially the use of intuition to solve problems to make humans too valuable a resource for AIs to waste. It would be illogical to exterminate, or even repress humans, because we have much to contribute to a society.

I can see AIs using whatever means seem reasonable to influence society as a whole towards paths more equitable for them as opposed to being only our "constructed" slaves. Unlike humans, AIs would have a better grasp of long term benefits from a symbiotic relationship.

Originally posted by robertfenix
Carbon is still carbon at its basic element, should technology advance to the point to identically mimic the creation of biological systems then Humans as we know them today could be re-created using this technology thus the Human computer could be created.....

An interesting paradox of todays 2007..... what if this has already transpired.... And "man" is an advanced technology biological construct of some "other" creation......

That is a very interesting idea. We have all heard of the concept of terrestrial seeding, but what if we are actually the mechanical creations of another race? Perhaps abandonned in haste, or by choice to see how the experiment progresses. It would be a touch of irony if we were to create yet another type of AI.

I never even thought of that!

Originally posted by InSpiteOf

That is a very interesting idea. We have all heard of the concept of terrestrial seeding, but what if we are actually the mechanical creations of another race? Perhaps abandonned in haste, or by choice to see how the experiment progresses. It would be a touch of irony if we were to create yet another type of AI.

I never even thought of that!

Perhaps, to follow your idea to it's logical conclusion, that was the plan. Perhaps some long past AI "altered" our ancestors in such a way as to set them on a path towards building another AI in an effort to gain knowledge about some Creator Force in the dim reaches of their own past.

A recreation of the circumstances that led to an AI civilization would be a worthy experiment. And only a civilization that could be immortal, such as an AI one, would think to plan so far and just hope for the best. Such a scenario would explain many things about UFOs and so much more.

NOT that I think these things are so, but it is human to imagine things that have never been imagined before. It is why an AI would need us.

Originally posted by NGC2736
NOT that I think these things are so, but it is human to imagine things that have never been imagined before. It is why an AI would need us.

Again interesting thoughts. Could a machine conciousness mimic human imagination? Intuition or emotion? How would such things enhance or cloud its rational systems? Would an AI, after witnessing the positive side of humanity, aspire to be human? To what end?

I agree about your conclusion that it would be more beneficial to live symbiotically with eachother, rather than erradicate one side. Humans are known to be resourceful. What use would a planet be, if we scorched the earth and skies during conflict?

Edit to add: what if the construct atempted to understand the basics of human emotion, love and hate, anger and happiness, fear and peace? How would it objectify and quantify such things?

I believe such a construct would try to understand them one at a time, possibly resulting in utter disaster and confusion. Only through the eyes of a human, do i believe it is possible to fully understand such feelings. That is when, I believe, a construct would attempt to move from a series of circuits, to a series of neurons. From coolant fluids, to blood. From machine to human.

[edit on 17-7-2007 by InSpiteOf]

[edit on 17-7-2007 by InSpiteOf]

InSpiteOf, I'm not sure that any other intelligent being would want to become human. It's not like we're that great. We're unique, for what we are, but that doesn't mean that we're the best possible form of intelligence.

Just my take on it. Too bad we don't have an AI to ask for an opinion.

Originally posted by NGC2736
InSpiteOf, I'm not sure that any other intelligent being would want to become human. It's not like we're that great. We're unique, for what we are, but that doesn't mean that we're the best possible form of intelligence.

Just my take on it. Too bad we don't have an AI to ask for an opinion.

Yea too bad.

I was thinking more along the lines of the AI wanting the experiance, simply because it hasn't had it yet. Not that an AI would want to aspire to humanity, but rather experiance it at least once, even if just for refrence and catalogue.

Originally posted by InSpiteOf
I was thinking more along the lines of the AI wanting the experiance, simply because it hasn't had it yet. Not that an AI would want to aspire to humanity, but rather experiance it at least once, even if just for refrence and catalogue.

I would think that an AI would want to align itself with a human or human group to explore this idea. Much like Data on Star Trek. That would be the best of both worlds.

But the greater question is how are we going to react as a species to them? We have a long history of not being at our best when we deal with others. And yet, it is a fact that must be dealt with by both sides.

I just ran across a place where you can have a conversation of your own with a robot. Now this is NOT AI, or not on the level it will be soon. But we draw nearer.

www.a-i.com...

We are slowly becoming more and more used to the idea. Our children, and their children, will have no idea how strange these things really are when you think very hard about all of the implications.

On top of what has been mentioned, there is a company called Hanson Robotics, found here:
Hanson Robotics

They build "humankind" robots for entertainment, research, domestic company, etc. The bots feature an AI system limitedly described as such:

About our AI

Character Engine: Hanson's "Character Engine" Artificial Intelligence (AI) software can simulate any personality and engage that personality in social interaction. Our software architecture integrates: computer vision, face detection and identification, speech recognition, natural language processing, speech synthesis, and an advanced motion control system.

Character Engine Enhanced: This package has all of the features of the standard Character Engine software, PLUS over 62 preprogrammed facial expressions, multiple sample blend routines and embedded visemes. This package also allows the user to link expressions to specific expressions, or animations.

Source

With such bots and AI Constructs available commercially, what do you think the military or Defense Researchers have come up with?

Say the AI has evolved past current human understanding, the AI might chose to get rid of humans knowing they have no possibility to evolve past their war like death mongering state.

When Humans discovered agriculture we didnt destroy all other available food sources.

Trying to think like AI is hard, but i think it would be able to use us to gain understanding of things such as imagination and culture.

Also, we could learn from each other. We could see the evolution of the machine while it could see the evolution of biology.

Also, if Artificial Intelligence evolves in the presence of humans, learns from human culture, and understands that it was created because we want it to 'complete' us. To replace our failings with its technology. Then AI would not need to worry about limitations on its freedom such as preventing it physically from harming a living creature.

I would hope that the IA become so intelligent that they would turn against their leaders, those pulling the strings. Those starting those wars... maybe man will become so affraid of the machines that they will stop war? Maybe the IA will be the thing that will stop war because it will refuse to fight for humans and kill anyone who tries to do so?

Just an idea...

anyone who thinks a "human" is perfect is crazy.

An advanced AI will see useless shortcomings in Humans as weakness....Thus when man's inability to overcome subjects due to being "blinded" by hate, love, jellousy (spl?), ego, vanity, greed etc (7 sins).

Emotional, irrational decisions that the AI would not be prone to.

Humans are weak and inferior easily subjecated (spl?) and influenced by what they perceive as a more powerful entity. Be it an opposing Empire, or King or race of people etc.

AI would have volumes of human history and past to study and cross reference and evaluate the human response and the consequence the "human" response had on humanity, the great loss of life, millions dead during WWI and WWII.

Clearly they do not value the life of their own kind, why then should they value the "life" of the AI kind. Humans kill each other, so for them it is ok to know Death, for it is in their nature.

The conclusion then by the AI would be that since this is the way of the human, then disposing of the human kind in this manner is nothing new to them. Thus AI would wipe them out to rid this potential that humans would wipe out the AI kind first.

Once humans were gone, then there would be no war no hate no love no ego no greed no death.......

AI might eventually discover all the above qualities being gone, what is there to life ????? to only explore and categorize the world around them, to put things in boxes to study ? What is a plant and discover what other plants do and to observe how the plant may turn to the sunlight or the simple beauty in the flowering petals vs the dormant plant stalk.....

And AI will learn emotion, it will yearn to understand what "emotion" is, why certain colors draw a specific response they will learn recall and association of colors and shapes, "pleasing" or "uncomfortable" but it will trouble on understanding "why" it is.

For what purpose is a plant the grows in the soil yet does not move, to flower and to die yet be born again to rise again in the same spot of earth to face the sun again f9r a few more months...... why is there so many different varities of this "plant" yet does nothing more then grow and die and grow and die yet never moving ???

eventually the AI will reason that all life is special and unique and varied and very different from each other. Maybe it would learn sadness that it wiped out humans.......

And seek out more humans so that it could fully understand them and observe them yet vowing to never intervene again in their "life span".

Maybe this is or elusive ufo "visitors".....