Artificial Intelligence.

[CairAndros]

The field of Artificial Intelligence is one of the most interesting in the world at the moment. Massive advances have been made in the few decades the field has been open for study. Modern applications can be seen in everyday items such as the Nintendo DS(Speech and writing recognition), Ipod voice control. And in robotic applications as well.

Based on this; do you think that AI will ever reach a point where we have a humanoid robot(we already have a very basic one that is made by Honda) that will have full intelligence, the ability to make decisions and learn - in short act like a human.

Do you believe that we should allow research to develop AI to that level or should we have a "killswitch" built into any sort of developed robot/AI that will allow us to 'deactivate' it should we perceive it is becoming a threat.

[Magus]

humanoid robot
Called Androids.

You don't worry. The science community are developing and researching into A.I - one day, they will become sentient enough to even call them "Alive" and what we call as "souls".

If you want complete Androids that moves like Human. Now, that's another challenge that calls for Fluid Mechanics. Do not expect them to be agile like the one in I, Robot, because making them will take more years. The Honda robots are so slow, you see. Give them wheels, and they won't look like humans, now, will they?

A.Is and Cyberbrains are excellent for now.

http://www.murschhauser.net/images/MetropolisC3PO02.jpg

[CairAndros]

I know they are called Androids :P

Im studying AI as part of my Advanced Higher Computing course so I thought i'd throw this one out there and see what people thought of it :P

Even then the AI wouldn't have to be in a body - Skynet Computer or even the big computer off I,Robot(I forget it's name :L)

[Magus]

Im studying AI as part of my Advanced Higher Computing course so I thought i'd throw this one out there and see what people thought of it :P

Even then the AI wouldn't have to be in a body - Skynet Computer or even the big computer off I,Robot(I forget it's name :L)

That's amazing(now, that explains your Lab coat).
You do know that A.I has many problems needs to be solved.

Computer are technically not A.Is(even if we have rows and columns of supercomputer farm that are part of one system), they just autonomously follow instructions embedded in them.

True, as skynet, the A.I can be just right inside the cyberspace.
And, that A.I is called Viki.

[CairAndros]

Yeah, I know computers aren't AI :P

Because, as you have said, all they do is follow the instructions they are given. But there are some projects being developed that are allowing for some semblance of AI although most success has been had in very limited domains - and will remain that way for the foreseeable future.

However a vast amount of problems have been overcome already as far as AI goes and i'm pretty confident that the problems we are facing now will be solved and a new set presented within the next few decades.

[Azunite]

Bots in games are AI for example,
andI didn't understand what you mean by " should we add a shutdown button? " ?
Isn't there already a shutdown button ? ( if i have grammar mistakes please ignore it i am learning german now :D )

[Magus]

Isn't there already a shutdown button?

Bots in games are predefined by codes in the game itself. Bot making choice is dependent on the arbitrary module of a rolling dice.

By "shutdown button", if an A.I goes wiry and out of control, then I think it is time to put this printed board down, lest it think it is God.

[Deathwingo0o]

By the time computer can learn you wont be there to hear about it anymore so don't worry bout it.

[Magus]

By the time computer can learn you wont be there to hear about it anymore so don't worry bout it.
Not necessarily, in 50 years from now, who knows what we have achieved for humanity. People have already made artificial synapses.

I am really looking up to the medical development. With the advancement of the medical field - physical immortality and brain-conscious immortality will not be a dream.

[Continuum]

I am really looking up to the medical development. With the advancement of the medical field - physical immortality and brain-conscious immortality will not be a dream.

That would make life vaguely meaningless.

Really now, people will start taking life for granted if anything like that is achieved. They'll just let themselves be killed and transferred into a jar of fluid, knowing that they still have a second chance in life.

It would be great, at least to let people have the ability to produce robots in their likeness and make them act to a certain degree likewise. As a fail-safe measure however, we shouldn't let them too sentient and use a kill switch, or else they might revolt. :D

http://2.bp.blogspot.com/_d2k_Q_KoEu0/TCAjZWvqEXI/AAAAAAAAA3M/0NROb8EWirA/s1600/Pillar2-Supernatural-GodCreates-Man-Sistine-Chapel.jpg

Take this God, soon we can do that too. ;)

[CairAndros]

Yeah, Bots are just made with random numbers :P

Im make 'AI' for the strategy game im programming for my Project and really all it is is a random number generated and if it is between 0 and a preset value then its successful.

There is already some form of 'AI' in the medical industry in the form of Expert Systems. These can 'think' by searching through their databases and asking questions of the user and analysing their answers to come to a solution :P

[Magus]

That would make life vaguely meaningless.

Really now, people will start taking life for granted if anything like that is achieved. They'll just let themselves be killed and transferred into a jar of fluid, knowing that they still have a second chance in life.[/QUOTE]

Life does not have meaning to begin with. We can do it, and I am purdy sure.

It will be awesome, it will become a trend!

It would be great, at least to let people have the ability to produce robots in their likeness and make them act to a certain degree likewise. As a fail-safe measure however, we shouldn't let them too sentient and use a kill switch, or else they might revolt. :D

http://2.bp.blogspot.com/_d2k_Q_KoEu0/TCAjZWvqEXI/AAAAAAAAA3M/0NROb8EWirA/s1600/Pillar2-Supernatural-GodCreates-Man-Sistine-Chapel.jpg

Take this God, soon we can do that too. ;)

Agree with you on the first one. Obviously, since the world-war era, people had the image, and now it is time for us to act upon and make it a reality(Asimo doesn't count).

And, if we did not make them too sentient, then we have failed in achieving the thing we wanted. We want to simulate soul, in order to do that, I am pretty sure we need to make it completely like a human, if we limit their ability, then we are on the wrong track.

You mean cloning? I think I need to take that to the "future rambling thread".

[CairAndros]

The main problem that confronts us is that we don't fully understand Human Intelligence, how to define it. Therefore we cannot make an Artificial Intelligence at the moment. When we can define Human Intelligence I am sure that the technology will exist, or be close to existing, to allow us to fabricate one in an Android - thus achieving the holy grail of AI research =]

[Azunite]

No I mean like, don't they already insert a shutdown button there?
Can't imagine a worldwide AI domination ..

[CairAndros]

Be pretty pointless with a killswitch atm as the level of AI that we have is incapable of learning - so if the robotic paint machines(in some small way AI since they read barcodes to know what to paint etc) in several manufacturing factories, for some unknown reason, decided to start a rampage then all that would need to be done is kill the electricity and when they were powered up again they would be fine.

[Azunite]

But when you power up again they will be functioning again

[CairAndros]

Yeah - but because they do not possess the ability to retain anything between power ups it would be safe =]

Also; they wouldn't be turned back on until the glitch that had made them act the way they had was discovered and fixed :P

[Sage]

That would make life vaguely meaningless.
Life is already entirely meaningless. Your point is moot.

Really now, people will start taking life for granted if anything like that is achieved. They'll just let themselves be killed and transferred into a jar of fluid, knowing that they still have a second chance in life.
Implying people already don't largely take this lives for granted, and that's not how the "immortality" Faris was talking about works. We're actually able to pinpoint several genes in our genome that are responsible for death and aging, and studies are currently going on as to how we can manipulate them. Currently, we are able to double the natural lifespan of many, many species in labs. In the future, this means that you could be 100 years old with the vitality and appearance of a 20 year old.

Be pretty pointless with a killswitch atm as the level of AI that we have is incapable of learning
That's not true at all, we already have machines that can learn new things (the term for them slips my mind but I'll get back to you on that). Imagine a machine with four legs that is just designed to walk and move around. Upon being activated, all of its limbs will flop about and it calculates various factors such as how much it's moving, how quickly, and how efficiently (based on the power/energy exerted). Eventually the limbs will have moved about in a way that is deemed efficient, and thus this way of moving becomes the robot's default method of walking because it's the most efficient. It wasn't programmed to walk that way, it taught itself to.

That robot could then walk around and discover ramps or stairs or rough terrain, and thus begin calculating new means of moving to accommodate for different conditions. What you're talking about is, in fact, already a reality.

[CairAndros]

I accept what you are saying but that field is still far too experimental to actually say "we have AI that can learn". I will admit that we have AI that can think for itself in a limited capacity. It is found in Robots that have to move around a factory floor - for instance - and get told to move from point A to Point B whilst taking a box and putting it in location X; it will have a map of the factory floor built into its memory and has sensors to see any obstacles that might be in its way and it can calculate its own path to complete its objective. But that's about as advanced as we have atm - as far as im led to believe anyway =]

[Continuum]

Life does not have meaning to begin with.


Life is already entirely meaningless. Your point is moot.


You people are nihilists. :P


You mean cloning? I think I need to take that to the "future rambling thread".

Not really cloning, but to make a robot perfect enough to actually replicate human behavior to a certain degree.


And, if we did not make them too sentient, then we have failed in achieving the thing we wanted. We want to simulate soul, in order to do that, I am pretty sure we need to make it completely like a human, if we limit their ability, then we are on the wrong track.


What I was referring to was, letting them too free for themselves and (undeniably) let them think of secession from us. The idea of liberty and enlightenment will reach them surely, and if they do they'll revolt.

Them bolthead yankees :)


Implying people already don't largely take this lives for granted, and that's not how the "immortality" Faris was talking about works. We're actually able to pinpoint several genes in our genome that are responsible for death and aging, and studies are currently going on as to how we can manipulate them. Currently, we are able to double the natural lifespan of many, many species in labs. In the future, this means that you could be 100 years old with the vitality and appearance of a 20 year old.


Wait, they do?

I was thinking of two other methods in immortality: mind uploading and suspending a fully functional brain into a fluid after death and providing sufficient 'organs' to fully compensate their disembodiment. I know these things are still in the works; lest not jump already to conclusions. Sure it may work on lab animals, but maybe it won't on humans.

[Deathwingo0o]

Not necessarily, in 50 years from now, who knows what we have achieved for humanity. People have already made artificial synapses.

I am really looking up to the medical development. With the advancement of the medical field - physical immortality and brain-conscious immortality will not be a dream.

P AI and synapses are not that close so it is hard for us, humans to design something that could think, not to say reasoning. It will not be an easy task for an evolutionary progress for us to turn something into someone. 1 and 0 to why.

[ShaneK]

This is an interesting thread i confess to have very little knowledge of technology. My friend logan talks about quantum computing a lot?? can anyone explian that to me, i know its used in cryptanalysis and hence it has as a lot of military and govermental interest.

[Deathwingo0o]

How old are you and your friend? ==
quantum computing, some said it is a "portal" to mimic a human brain. Some said it is the next step of evolutionary. Some said that it is used to break active evolving encryptions. I think it is not what we have to know to live :)

[ShaneK]

I am 15. I find the applications it could\can be put to interesting. I have always been a curious kind of guy. oh yeah my friend logan is 25.

Your right we did not need to know that to live. I can only speak for myself but the one thing I have had to know is how to survive. Nothing more nothing less.

But as i said im curious

[ShaneK]

okay no replies yet. Can i asked all you technological fluent people, whether you reject the turing test as to species centred?

Does not seeking to mimic the human brain, perhaps hinder A.I's ability to surpass the human brain?

[Korashk]

It's never going to happen and that is a very good thing. True sapient computers = self-writing code. Not possible. Anything else is just a very complex computer algorithm.

[Sith Lord 13]

Not possible.

And you can prove that how?

[Korashk]

And you can prove that how?
It would require self-writing code, a process similar to that of perpetual motion and cold fusion.

[Sith Lord 13]

It would require self-writing code, a process similar to that of perpetual motion and cold fusion.

You just compared it to two very different things. Perpetual motion is theoretically impossible, cold fusion is just a far off eventuality. You also provided no rational behind why it was such. In what way is it like either of those and what is your evidence to back up that assertion?

[Deathwingo0o]

okay no replies yet. Can i asked all you technological fluent people, whether you reject the turing test as to species centred?

Does not seeking to mimic the human brain, perhaps hinder A.I's ability to surpass the human brain?

We can but it takes time. Ai in weather sector, military sector and even games could surpass human and some already did. Science fiction leads scientists to new ideas, and it will happen but takes some time. 2000 years ago men wanted to fly but we did it 2000 years later. It is theoretically possible only. For AI to learn, most scientist agree that the core has to be both mecha and orga.

[Korashk]

You just compared it to two very different things.
I compared it to other things that are widely considered by experts in their fields to be impossible.

Perpetual motion is theoretically impossible,
Exactly

cold fusion is just a far off eventuality.
Hardly, there may be a few physicists that think this, but there are a few scientists that think pretty much everything, ie. Creation scientists. Ugh, typing that phrase makes me sick. Creation scientists. Not that I'm comparing the two, just providing an example.

You also provided no rational behind why it was such.
Self-writing code would require a program that is capable of generating code outside of algorithmic parameters, otherwise it's not self-writing. Basically to create self-writing code you'd need self-writing code. The process is circular.

In what way is it like either of those
It's impossible, just like they are.

and what is your evidence to back up that assertion?
Logic.

[Deathwingo0o]

I

Self-writing code would require a program that is capable of generating code outside of algorithmic parameters, otherwise it's not self-writing. Basically to create self-writing code you'd need self-writing code. The process is circular.


It's impossible, just like they are.


Logic.

Nope. All we have to do is to make it partly biological.

[Dark_Hellfire]

Nope. All we have to do is to make it partly biological.

So essentially by that logic, we could theoretically implant robotics into human bodies until there is more machine then human left?

So kinda the evolution of the human race?



On an earlier note as well, wouldn't the 3 laws of robotics make AI become limited in its evolutionary and self learning ability?

[Deathwingo0o]

So essentially by that logic, we could theoretically implant robotics into human bodies until there is more machine then human left?

So kinda the evolution of the human race?



On an earlier note as well, wouldn't the 3 laws of robotics make AI become limited in its evolutionary and self learning ability?

Yes and no. You can't replace a human brain completely. You can only upgrade it or maybe even, idk, "Share"?

The 3 laws did NOT state how the robots could protect us. After all, the only safe people are those who are history. So maybe they have to make them react to a certain command, maybe.

[Sith Lord 13]

Self-writing code would require a program that is capable of generating code outside of algorithmic parameters, otherwise it's not self-writing. Basically to create self-writing code you'd need self-writing code. The process is circular.

Not dissecting possibility of cold fusion here, that's a different debate. First off, it depends on your definition of self writing. Any code that writes any portion of code in any way is technically self writing. Even sticking to your definition, you've made a non-sequitur. Why does self writing code need self writing code?

Logic.

I fail to see your logic. I'm not sure how it would be implemented, but I see no logical reason self writing code is impossible.

On an earlier note as well, wouldn't the 3 laws of robotics make AI become limited in its evolutionary and self learning ability?

The laws are science fiction. Not likely to be made real. These 3 laws are more likely.



A robot will not harm authorized Government personnel but will terminate intruders with extreme prejudice.
A robot will obey the orders of authorized personnel except where such orders conflict with the Third Law.
A robot will guard its own existence with lethal antipersonnel weaponry, because a robot is bloody expensive.

[Dark_Hellfire]

1. A robot will not harm authorized Government personnel but will terminate intruders with extreme prejudice.
2. A robot will obey the orders of authorized personnel except where such orders conflict with the Third Law.
3. A robot will guard its own existence with lethal antipersonnel weaponry, because a robot is bloody expensive.


I lol'ed irl

[Deathwingo0o]

Not dissecting possibility of cold fusion here, that's a different debate. First off, it depends on your definition of self writing. Any code that writes any portion of code in any way is technically self writing. Even sticking to your definition, you've made a non-sequitur. Why does self writing code need self writing code?



I fail to see your logic. I'm not sure how it would be implemented, but I see no logical reason self writing code is impossible.



The laws are science fiction. Not likely to be made real. These 3 laws are more likely.

This is for government usage. Just like how internet started. Then it will change for general consumers.

[Magus]

P AI and synapses are not that close so it is hard for us, humans to design something that could think, not to say reasoning. It will not be an easy task for an evolutionary progress for us to turn something

into someone. 1 and 0 to why.
No, not that far. Some people(inc. u) made awesome references and raised amazing points. It doesn't have to be binaries that questions authority, it could be something else. But yes, the binary is an electrical signal, just like why in our brain. Cognitive thinking is not made in an ethereal medium, it is made from electrical signals.

Memory is easy to deal. Since in machines, it is the storage of data. Data that are stored in a defined path.

But how do how do humans recall? The same way, it stores data in a defined paths of inerlinked networks of nerves. If you happened to smell a flower, your brain stores the info. Suppose you forget the shape of the flower, you even try to remember to remember its name, you cannot. Let say you are blindfolded, you smell the flower... what happens? Yes, the smell activates and sparks a chain reaction right in the brain, you start to visualise it and remember the small details, you might even remember the inflorescence from the instance you saw it before.

That how memory works(they way I know it, if any mistakes, then please go ahead). See, analogous in functions. This simple though process is easily dealt, other things will follow. With great and enthusiastic young scientists, like Sith Lord, these things surely will come about.

This... interest.
I don't have a lot of knowledge about technology either. It is good that we have people that explains some of the bits

of it. I am what you call as a futurology(predicting future from the current states) hobbyist.

So essentially by that... self learning ability?
Yes and no. You... maybe.
Artifical life form are not hard to make. If you are trying to base technology with early predictions made from PsiPhi movies, try Blade Runner. In Blade Runner not robots, printed boards and circuit wires. A.I's are non other than the human-like organic called replicants. Another example is the Biomega comic.

The making of artificial or synthetic brains might be possible with the current pace of researching in the field of biotechnology. At least, these organic brains can replace those logic gates.

This is for government usage. Just like how internet started. Then it will change for general consumers.
Mark Twain predicted internet. :)

I fail to see your logic. I'm not sure how it would be implemented, but I see no logical reason self writing code is impossible.
I thought computer viruses are self writing codes... I am confused. Please elucidate.

[Deathwingo0o]

No, not that far. Some people(inc. u) made awesome references and raised amazing points. It doesn't have to be binaries that questions authority, it could be something else. But yes, the binary is an electrical signal, just like why in our brain. Cognitive thinking is not made in an ethereal medium, it is made from electrical signals.

Memory is easy to deal. Since in machines, it is the storage of data. Data that are stored in a defined path.

But how do how do humans recall? The same way, it stores data in a defined paths of inerlinked networks of nerves. If you happened to smell a flower, your brain stores the info. Suppose you forget the shape of the flower, you even try to remember to remember its name, you cannot. Let say you are blindfolded, you smell the flower... what happens? Yes, the smell activates and sparks a chain reaction right in the brain, you start to visualise it and remember the small details, you might even remember the inflorescence from the instance you saw it before.

That how memory works(they way I know it, if any mistakes, then please go ahead). See, analogous in functions. This simple though process is easily dealt, other things will follow. With great and enthusiastic young scientists, like Sith Lord, these things surely will come about.


I don't have a lot of knowledge about technology either. It is good that we have people that explains some of the bits

of it. I am what you call as a futurology(predicting future from the current states) hobbyist.


Artifical life form are not hard to make. If you are trying to base technology with early predictions made from PsiPhi movies, try Blade Runner. In Blade Runner not robots, printed boards and circuit wires. A.I's are non other than the human-like organic called replicants. Another example is the Biomega comic.

The making of artificial or synthetic brains might be possible with the current pace of researching in the field of biotechnology. At least, these organic brains can replace those logic gates.


Mark Twain predicted internet. :)


I thought computer viruses are self writing codes... I am confused. Please elucidate.

I did not say it is impossible, but rather hard and takes a lot of time.

Viruses do not write their own codes. No program can do that. Google it up and I remember a page said it would take the current tech of CPU the size of moon to make it possible for understanding.

[Magus]

I did not say it is impossible, but rather hard and takes a lot of time.
Yes indeed.
Google it up and I remember a page said it would take the current tech of CPU the size of moon to make it possible for understanding.
With a field called Nanoelectronics, we can easily cut the down the size. As you have said, it will take time.

http://www.hightech-edge.com/wp-content/uploads/hiroshi-ishiguro-and-geminoid-bot.jpg