artificial inteligence is a myth!!!
-
Collin Jasnoch wrote:
I also have studied A.I. and there is plenty of it already out there. Maybe it does not meat your standards of being intellegent
Redefining the terms to refute the argument doesn't alter the fact that with the current normal definition there is no AI.
Collin Jasnoch wrote:
But then again, maybe you just aren't very intelligent. So your standards are irrelevent
Redefining the term to something different and then answering that doesn't lead to that conclusion.
I am not redefining anything. The claim was that "A.I. is a myth" because it does not compare to human inteligence. With the given definition of inteligence I argue that A.I. does exist and exceeds most human inteligence.
jschell wrote:
doesn't alter the fact that with the current normal definition there is no AI.
What current "normal" definition are you talking about? Are you claiming to be a scholar on A.I.? FYI there is no "normal" definition of A.I.. The most known/common test for AI is the Turing test, which is being questioned by all as to wether or not it proves anything. And I am sure you knew that... Stop trolling.
jschell wrote:
Redefining the term to something different and then answering that doesn't lead to that conclusion.
This is meaningless. My point was that if the question being asked is not understood by the asker (which is often the case), then how can the questioner confidently say it can not be mapped? The fact that one is asking the question tends to mean (tends to as questioning is also a method of teaching), that the questioner does not know the answer or how to obtain the answer. In that case they lack sufficient evidence to make any conclusion nor even make a hypothesis to set up an experiment to prove one outcome over the other. Lack of evidence does not mean there is no evidence. It just means it is not understood.
Computers have been intelligent for a long time now. It just so happens that the program writers are about as effective as a room full of monkeys trying to crank out a copy of Hamlet.
-
GuyThiebaut wrote:
If by intelligence you mean the ability to make decisions quickly, accurately and learn from mistakes then I would say that AI is pretty darn good!
I'm not sure the speed it comes to that decision matters for determining intelligence (though it may help place it on a relative scale of how intelligent it is), but I agree otherwise. A system I developed for work is like that. I wasn't intending to do anything with AI, nor do I have much knowledge in that field (an interest, but I haven't put much effort into it), but it just sort of came together that way. It's not particularly advanced AI either, but it takes some input, finds a way to describe it, and uses that description to find similar data. When it fails, it can be shown more data, adjusts its description, and tries again. It's surprisingly effective, and does a task that previously had to be done by a human.
Sounds a little like my AC (Artificial Cockroach). It was designed to take a fixed set of inputs and then make an intelligent (hopefully) decison from an also fixed set of options. You give it it's inputs and it responds with the option (or action) it would like to do. It is configured by an evolutionary algorithm and each instance is 'born' with all the knowledge it will ever have. It also will never be able to come up with new responses. Just like a cockroach. What it can do is adapting quickly to new circumstances and finding ways of using it's options as good as possible. Like a cockroach it depends on evolution and a fast sequence of generations to do that. Despite it's limitations and very humble 'intelligence' it's really fascinating to study and it sure beats any 'if this else that' logic which never adapts.
At least artificial intelligence already is superior to natural stupidity
-
Sounds a little like my AC (Artificial Cockroach). It was designed to take a fixed set of inputs and then make an intelligent (hopefully) decison from an also fixed set of options. You give it it's inputs and it responds with the option (or action) it would like to do. It is configured by an evolutionary algorithm and each instance is 'born' with all the knowledge it will ever have. It also will never be able to come up with new responses. Just like a cockroach. What it can do is adapting quickly to new circumstances and finding ways of using it's options as good as possible. Like a cockroach it depends on evolution and a fast sequence of generations to do that. Despite it's limitations and very humble 'intelligence' it's really fascinating to study and it sure beats any 'if this else that' logic which never adapts.
At least artificial intelligence already is superior to natural stupidity
I don't think it's quite that complicated. And this one just modifies itself as it fails (it requires human interaction to assist it in discovering its failures, but it is generally able to tell when it failed, just not why). But that sounds rather interesting, I've never had much luck with genetic algorithms (but then again, I think I often try to jump in the deep end before learning how to swim).
-
I don't think it's quite that complicated. And this one just modifies itself as it fails (it requires human interaction to assist it in discovering its failures, but it is generally able to tell when it failed, just not why). But that sounds rather interesting, I've never had much luck with genetic algorithms (but then again, I think I often try to jump in the deep end before learning how to swim).
You would not believe how simple the cockroach is. It uses the most primitive structure to make its choice from the options: A lookup table. The values in the table are taken directly from the evolutionary algorithm's 'DNA'. Initially the entire DNA is filled with 0 so that all values which eventually appear are created by mutation and selection. A highschool kid could whip something like it together in no time. The point was to show what kind of results even such a primitive thing can bring. Things really get interesting when you use the same approach on a more advanced decision maker, like a neural nets. For fun I used the cockroach for a kind of RPG. The 'monsters' were trained by letting them fight against each other, providing them with inputs about the current situation and then asking them what they want to do. They even learned to run away when things turned against them or changed their reactions when players had figured out how to beat them.
At least artificial intelligence already is superior to natural stupidity
-
Have we not been there already? If I remember correctly, another one of your opinions was that if you don't understand C++ and therefore don't like it, then everybody else must also dislike it for the same reasons. And I also remember your reactions when someone does not share your opinions. Therefore you may be a totally misunderstood genius, so forgive us mere mortals. Or maybe you really don't realize how provokative, arrogant and not so intelligent your behavior may make you appear. Looking at your past posts, I prefer to think that you are deliberately posting rough 'opinions' and just enjoy defending them to the bitter end. That would make you a troll. Not the worst kind, but a troll nonetheless.
At least artificial intelligence already is superior to natural stupidity
maybe you are right but that rought opinions are my opinions
-
Vasily Tserekh wrote:
Have you ever used neural networks i did, they dont learn they are just a mathematic code that adjust the output based on previows input-output examples and that is if the problem you want to solve can be mathematically modelled if not then neural networks is useless.
Yes I have used them. And by your description I think you barely studied them. It is not as simple as that unless you are talking about simple neural networks. When you involve hidden layers and different learning theories (yes they do learn) you can model things methematically. The whole point is to use it when the mathematical model is unkown (i.e. you do not know if one exists but theorize it does). In these cases the performance is still often better than humans because they can respond quicker and with better results. Also, NN are not the only form of A.I. It is actually a tool to use when making an A.I. system. Humans are not intelligent just because they have a Neural Network. They have many other things going on as well (obviously). The same is true with digital programs. But their things that are going on may end up slightly different allowing them to elevate their inteligence (e.g. being connected to a database vs. being connected to a brain of memories)
Vasily Tserekh wrote:
My friend real inteligence cant have a mathematicall model so it cant be programmed.
How are you so certain? Show me inteligence that can not be "mapped" as you say, and I will show you how maybe you just are not inteligent enough to map it.
Computers have been intelligent for a long time now. It just so happens that the program writers are about as effective as a room full of monkeys trying to crank out a copy of Hamlet.
I agree to some extent with both of you Vasily and Collin. I think(well is more like i hope or wish) humanity will be able to create intelligence as good as normal human being at least, posibly something beyond that, but im sure that it wont happen in our lifetime.
Collin Jasnoch wrote:
Vasily Tserekh wrote:
My friend real inteligence cant have a mathematicall model so it cant be
programmed.Collin Jasnoch wrote:
How are you so certain?
Show me inteligence that can not be "mapped" as you say, and I will show you how maybe you just are not inteligent enough to map it.That one is easy, free will cant be mapped/modeled, an algorithm is based on a set of rules. You can say that you can create a/"set of" rule to create new rules, but then again is just another rule that cant be broken. free will is the part of intelligence that make us able to break any rule to adapt or get better.
-
GuyThiebaut wrote:
If by intelligence you mean the ability to make decisions quickly, accurately and learn from mistakes then I would say that AI is pretty darn good!
Common general definition of that term would involve something along one or both of the following - A robot that can act in such a way to mimic many common human day to day behaviors such as driving a car, shopping and taking care of a child. - A computer that can interact with humans and also create new general ideas. So it can create a painting, write a book, criticize a play and banter about a favorite sports team. There is no artificial intelligence. And with the current state of that study there never will be.
jschell wrote:
There is no artificial intelligence. And with the current state of that study there never will be.
I would not give up that easily. We know one system that has declared itself to be intelligent. It is made up of spophisticated miniature switching units, known as neurons. We can emulate the switching function more or less precisely on a computer. But understanding and emulating this smallest unit of the system is not the key to intelligence. A brain is a network of large neural networks, so complex that it's unlikely that we can simply design a similar network and emulate it. But we do know the algorithm that has configured our brains. It's called evolution and we can also emulate it. If we disregard the amount of time it may require and also the capacity of the computer wich could do those emulations, I still think that it is possible to get results. And if that is true, those reults could just as well be aliens from another planet because they have been bred to survive and adapt to an emulated environment.
At least artificial intelligence already is superior to natural stupidity
-
I agree to some extent with both of you Vasily and Collin. I think(well is more like i hope or wish) humanity will be able to create intelligence as good as normal human being at least, posibly something beyond that, but im sure that it wont happen in our lifetime.
Collin Jasnoch wrote:
Vasily Tserekh wrote:
My friend real inteligence cant have a mathematicall model so it cant be
programmed.Collin Jasnoch wrote:
How are you so certain?
Show me inteligence that can not be "mapped" as you say, and I will show you how maybe you just are not inteligent enough to map it.That one is easy, free will cant be mapped/modeled, an algorithm is based on a set of rules. You can say that you can create a/"set of" rule to create new rules, but then again is just another rule that cant be broken. free will is the part of intelligence that make us able to break any rule to adapt or get better.
-
What does free will have to do with intelligence?
Computers have been intelligent for a long time now. It just so happens that the program writers are about as effective as a room full of monkeys trying to crank out a copy of Hamlet.
OK, dont know if you are trying to confuse me, playing dumb or it is a genuine question, i will suppose the last one. Lets stay with the example you gave of current AI, the social networks algorithms. It is based on a set of rules, lets say(for the sake of simplicity) if the user if from north america and is male, the algorithm will "decide" to show a beer adverticing. Now of course the algorithm may take tons of rules i used only 2 because i want to keep this simple. Now if the one who had to decide what adverticing must show to the users was a human being, he may decide to show other adverticing although he was only instructed to only take into account the location and gender, he may also take into account new paramters withouth being told to do so, like age, politic, religion, etc. So with this simple example(maybe not the best) just what im trying to say is that a part of our intelligence is the capability to break the rules, which i called free will(maybe not the best translation because english is not my native language). I hope i had made my point at least little bit more clear.
-
I see a lot of people responding that technology will evolve to the point where we will have good AI. I don't think that is what you are saying so I am not sure why they are using that as their argument. I agree, with current technology and if elses and loops, we will not get to AI.
There are only 10 types of people in the world, those who understand binary and those who don't.
ryanb31 wrote:
I agree, with current technology and if elses and loops, we will not get to AI.
And the advances so far, however humble or grand they may be, have been accomplished with what? Magic spells? Those things are just the atoms we use to write down algorithms and probably will continue to play that role for a while yet.
At least artificial intelligence already is superior to natural stupidity
-
ryanb31 wrote:
I agree, with current technology and if elses and loops, we will not get to AI.
And the advances so far, however humble or grand they may be, have been accomplished with what? Magic spells? Those things are just the atoms we use to write down algorithms and probably will continue to play that role for a while yet.
At least artificial intelligence already is superior to natural stupidity
Magic is code that works and no one is sure why. :) So, yes, there is some of that. Yes, ifs and loops have built the current advances but the point is to get close to true AI we'll need something a lot better. However, with processor speeds continuing to improve maybe a switch statement with 12 million cases and nested ifs won't perform that bad.
There are only 10 types of people in the world, those who understand binary and those who don't.
-
Any programmer who thinks that we are going to reach the human intelligence by if switchs elses for loops, is either crazy or has inhaled a pound of cocaine
...u mean, like u did? I don't think that invention of AI got anything to do with stupid messages like this published here.
-
Oh sorry I should have posted In my humble opinion I think that we are not going to imitate human intelligence by for loops else etc what do you think and please dont slap me in the face
Have we accomplished anything with loops or other basic program structures? Yes, we have used them to formulate algorithms. I also see no fundamentally different alternative, so future algorithms, whatever they may be able to do, will very probably be formulated the same way for any forseeable time.
Vasily Tserekh wrote:
and please dont slap me in the face
That will have to wait until I come to Cuba next time :)
At least artificial intelligence already is superior to natural stupidity
-
Magic is code that works and no one is sure why. :) So, yes, there is some of that. Yes, ifs and loops have built the current advances but the point is to get close to true AI we'll need something a lot better. However, with processor speeds continuing to improve maybe a switch statement with 12 million cases and nested ifs won't perform that bad.
There are only 10 types of people in the world, those who understand binary and those who don't.
I would place my bets on neural networks, which by the way are ideal to be 'offshored' to the graphics processor. And then there is that strange new idea to build supercomputers with as many GPUs they can find... Seriously, I see no alternative to algorithms and the now traditional ways of writing them down. I don't expect that to change any time soon. What you probably really mean is that the algorithms were laughably inadequate up to now. Edit: I just summoned up the disassembly of a switch statement before my inner eyes and asked myself what makes you think that we need fast processors for large switches? Apart from a tiny delay by a sure cache miss after a certain size, the branching instructions (that's the equivalent of a goto ;P ) will need the same time as they always do.
At least artificial intelligence already is superior to natural stupidity
-
If you are going to slam someone, make sure your post is impeachable, i.e. "an aply", "meat". Dave.
Or, "intellegent"! ;P
-
Vasily Tserekh wrote:
also my parents are russians and you REALLY know they are not stupids
Russians are just as capable of being stupid as the rest of the world, and smart parents can have stupid children, so I fail to see what point you're trying to make. (Not saying anything about you specifically either way, just that your point makes no sense.)
Whoa there! Respect a man who respects his parents. We don't have too many of that.
-
I think you are right. Programming is about procedures to solve problems, effectively. One can only program as many solutions as the programmer can think of. But true intelligence is the ability to solve NEW problems. Therefore traditional programming canot achieve that.
============================== Nothing to say.
Very early in the game (I am thinking around 1970 or a bit earlier), a theorem-prover was set loose on a known problem. It came out with a proof for a theorem; that proof was different from what the mathematicians had constructed. The mathematicians had to agree that the new proof was correct. What do you say about that? Among the earliest successes in AI have been the Checkers playing program of the same vintage; MYCIN which at that time could take in symptoms, suggest medical tests, and based on the results suggest a diagnosis; a program that could figure out the structure of crystals from its x-rays; etc. More recently, it is Deep Blue, the chess-playing program and Watson which can answer Jeopardy questions and is being deployed in business environments. All one can say is that there are solvable problems in AI and there are problems not solvable today. That does not mean that they will remain unsolved forever. Peace.
-
OK, dont know if you are trying to confuse me, playing dumb or it is a genuine question, i will suppose the last one. Lets stay with the example you gave of current AI, the social networks algorithms. It is based on a set of rules, lets say(for the sake of simplicity) if the user if from north america and is male, the algorithm will "decide" to show a beer adverticing. Now of course the algorithm may take tons of rules i used only 2 because i want to keep this simple. Now if the one who had to decide what adverticing must show to the users was a human being, he may decide to show other adverticing although he was only instructed to only take into account the location and gender, he may also take into account new paramters withouth being told to do so, like age, politic, religion, etc. So with this simple example(maybe not the best) just what im trying to say is that a part of our intelligence is the capability to break the rules, which i called free will(maybe not the best translation because english is not my native language). I hope i had made my point at least little bit more clear.
What you have described is not free will. You just added more choices. The algorithm can do that as well. I think what you maybe want to say is you believe a human can adapt to an unknown option (this really has nothing to do with free will). Your assumption is wrong. Various studies have shown that we can have a network readjust itself even after being taught. It's simply new parameters. In your example age etc. The AI does learn that age is relevant for beer. And maybe even more effectively. A human would continually think "oh they are 21" while the AI will quickly pick up on low response (why...because beer is an acquired taste...but neither AI or human care). The AI then starts given rum ads while the human keeps failing with beer ads. -------- This has nothing to do with free will. Free will is about being able to choose A over B even though something guides us to A. One could argue that because AI does not have free will they wiould consistently pick the more intelligent and logical choice. Then again one can also argue that free will is an illusion. The choice we pick under the belief of "free will" was in fact a guide yet highly complex calculation that we simply do not comprehend, but use every moment of our lives.
Computers have been intelligent for a long time now. It just so happens that the program writers are about as effective as a room full of monkeys trying to crank out a copy of Hamlet.
-
Any programmer who thinks that we are going to reach the human intelligence by if switchs elses for loops, is either crazy or has inhaled a pound of cocaine
Vasily Tserekh wrote:
Any programmer who thinks that we are going to reach the human intelligence by if switchs elses for loops, is either crazy
Perhaps. My personal opinion is that AI systems and the current state of hardware can't come close to the computational and reasoning prowess of the human brain. But what does that have to with AI being a myth? Perhaps you don't realize that what was once considered state-of-the-art AI (image processing, expert systems, robotics, natural language translation, neural networks) is being used by you and I in our daily lives. /ravi
My new year resolution: 2048 x 1536 Home | Articles | My .NET bits | Freeware ravib(at)ravib(dot)com
-
What you have described is not free will. You just added more choices. The algorithm can do that as well. I think what you maybe want to say is you believe a human can adapt to an unknown option (this really has nothing to do with free will). Your assumption is wrong. Various studies have shown that we can have a network readjust itself even after being taught. It's simply new parameters. In your example age etc. The AI does learn that age is relevant for beer. And maybe even more effectively. A human would continually think "oh they are 21" while the AI will quickly pick up on low response (why...because beer is an acquired taste...but neither AI or human care). The AI then starts given rum ads while the human keeps failing with beer ads. -------- This has nothing to do with free will. Free will is about being able to choose A over B even though something guides us to A. One could argue that because AI does not have free will they wiould consistently pick the more intelligent and logical choice. Then again one can also argue that free will is an illusion. The choice we pick under the belief of "free will" was in fact a guide yet highly complex calculation that we simply do not comprehend, but use every moment of our lives.
Computers have been intelligent for a long time now. It just so happens that the program writers are about as effective as a room full of monkeys trying to crank out a copy of Hamlet.
OK i think i understand what you are saying, and yes maybe the term free will is not correct, but to me it seems the best way to call it. Im not an expert in Neural networks, but i have used them several times at college and yes they can learn, but unless you are talking about a new type i dont know or heard about, the structure of a neural network is static, what changes with the training is the weight of the path the impulse runs through between 2 neurons, and yes after they are trained they can be trained again to "learn" something diferent. But as far as i know what they cannot do is to change the input, i mean if you create a huge neural network that works with an input of lets say 25 parameters, it will always take into consideration 25 parameters, it cant grow bigger than that, in the example the NN takes 2 parameters into consideration, if it was designed that way, doesnt matter how many layers it has it wont take into account any other parameter, for that being posible you would need a new NN. So what im trying to say is that what i dont see posible in AI as it is right now is the adaptability our human intelligence has to get new parameters into account(at least talking about NN, im sure there are plenty of other models/theories in AI that i havent heard about). Well then all this was on my assumption that there is no NN that has the ability to mutate/evolve(with this i mean that the same NN, can change to be able to receive a new input it wasnt designed for) without being redesigned, if exists a model or something i dont know please just tell me the name of that model or theory so i can read about it and agree completely with you.