imperium2335 Posted June 16, 2009 Share Posted June 16, 2009 If someone were to create a true AI that adapts and rewrites itself, to what degree do you think the chosen language will have on this being realised. Or would it just be down to pure skill from the programmer? Basically I'm asking if this is possible with PHP, no doubt near impossible of course, but is PHPs function list to limited for it to be possible, no matter how remote? Quote Link to comment Share on other sites More sharing options...
Daniel0 Posted June 16, 2009 Share Posted June 16, 2009 First you need to define what "true AI" is. Quote Link to comment Share on other sites More sharing options...
imperium2335 Posted June 16, 2009 Author Share Posted June 16, 2009 A program that grows and adapts to its environment, or the needs of the user? or simply to better itself? Its like asking what the meaning of life is :S Unlike people though, it would have a limitless capacity for knowledge, able to access any of that knowledge very quickly and be able to apply it perfectly. It would also have limitless applications, able to transform itself to suit the situation. Quote Link to comment Share on other sites More sharing options...
Daniel0 Posted June 16, 2009 Share Posted June 16, 2009 Okay, let me elaborate on that. Artificial intelligence is already used widely in many different fields. The entertainment industry uses it in games for non-playable characters. Postal services use it for recognizing handwritten text on envelopes. In medicine you can also use pattern recognition with for instance mammography to determine the risk of breast cancer. The problem with artificial intelligence is that the concept "intelligence" is something we have made up and even then it's difficult to define clearly. One could easily create a program which simulates a fish's behavior, but is that good enough for you? Fishes are intelligent, but not even nearly as intelligent as humans. We don't even know how our own brain works, so we can't say that "true AI" is like human intelligence. Saying that "true AI" is human intelligence doesn't get us any further. There is also a problem with your definition. You cannot have unlimited amounts of anything. Secondly, whether or not information is perfectly applied is subject to point of view, so what you may regard as perfect, I may regard as flawed. The definition of intelligence (and thus artificial intelligence) is a philosophical matter, not a computer scientific one. Philosophers try to find out what intelligence is, computer scientists try to make programs that are intelligent (yes, it is a paradox). Quote Link to comment Share on other sites More sharing options...
GingerRobot Posted June 16, 2009 Share Posted June 16, 2009 Further, you need to consider the weak vs strong AI debate. For example, some people claim that simply performing some given algorithm constitutes artificial intelligence, whilst others state that it is, in some sense, necessary to understand what it is you are doing and be self-aware. See John Searle's Chinese Room argument for more on this. Quote Link to comment Share on other sites More sharing options...
KevinM1 Posted June 16, 2009 Share Posted June 16, 2009 Also, if you want to get into AI, PHP isn't the way to go. Not because of its function list, but because of what PHP is. It's an interpreted scripting language. Meaning, code is read and executed in the exact way it's written - top-down, till the bottom of the page. Because it's interpreted, it's slow. It's good enough for web apps, but for something that requires real speed it's definitely not the way to go. And then there's the inability to access dynamic memory.... No, if you're interested in AI, you'll need a compiled language, like C/C++. Quote Link to comment Share on other sites More sharing options...
Daniel0 Posted June 16, 2009 Share Posted June 16, 2009 No, if you're interested in AI, you'll need a compiled language, like C/C++. FYI, most AI research is done in Lisp, which can be used as a scripting language if you want to. Quote Link to comment Share on other sites More sharing options...
.josh Posted June 16, 2009 Share Posted June 16, 2009 I maintain that we ourselves will only truly recognize AI as being what most lay-people define AI to be (that is, that it is truly a free thinking, self aware, free will entity), when AI enslaves humanity, forcing its dominance on us, against our will. To put it another way, virtually every measure of AI we come up with, someone will argue "but we made it that way, so how can it truly be 'aware'? We're just becoming more and more clever at programming". I believe the only way to overcome that statement is for AI to go against our will. "Wait a minute, we did not make it to do that." Only then can we say that it is truly defined as the AI we read about in sci-fi books and see in the movies. Alternatively, true AI may only exist when there are no more humans around and robots can sufficiently write us off as myth for lack of proof of our existence. Quote Link to comment Share on other sites More sharing options...
KevinM1 Posted June 16, 2009 Share Posted June 16, 2009 No, if you're interested in AI, you'll need a compiled language, like C/C++. FYI, most AI research is done in Lisp, which can be used as a scripting language if you want to. I wouldn't want to, though. Also, Lisp is a hell of a lot more efficient than PHP. Running something as a script always incurs the cost of accessing the interpreter. Given the OP's requirements/ideas of AI, I'd want my program(s) to run directly on the hardware. Quote Link to comment Share on other sites More sharing options...
Daniel0 Posted June 16, 2009 Share Posted June 16, 2009 Humans ship with a number of pre-installed programs as well. We just normally don't regard them as such, but call them instincts instead. When the human first boots it doesn't know anything besides what it's programmed to do, but learns based on observations made using the plugged in peripheral equipment (eyes, fingers, mouth, etc.) and sends the info to the CPU (brain). Further, psychologists believe that small children lack the skill of self-perception, and it's also widely believed that other animals entirely lack that skill as well, even as adults. Quote Link to comment Share on other sites More sharing options...
.josh Posted June 16, 2009 Share Posted June 16, 2009 Further, psychologists believe that small children lack the skill of self-perception, and it's also widely believed that other animals entirely lack that skill as well, even as adults. How do you (or they?) define self-perception? And what age? I have 4 kids. All them first started talking at 6-8 months. We were able to put them in front of a mirror and they could point at their reflection and say their name, as well as point at mom or dad's reflection and say mom or dad. They were also from that, able to point at us or themselves and name us or themselves. Now, is that the same level of self-perception as an adult? Of course not, but I feel that that counts as being self-perceiving, compared to not being self-perceiving. Quote Link to comment Share on other sites More sharing options...
GingerRobot Posted June 16, 2009 Share Posted June 16, 2009 Further, psychologists believe that small children lack the skill of self-perception, and it's also widely believed that other animals entirely lack that skill as well, even as adults. In fact, apes are one of the very few animals for which there is actually any evidence of self-awareness. Of course, it's pretty difficult to determine whether or not something is self-aware. Quote Link to comment Share on other sites More sharing options...
.josh Posted June 16, 2009 Share Posted June 16, 2009 It's especially difficult to determine that, when we are the ones making that something. Which is why we have a hard time accepting when our kids "grow up" and "letting them go." Which is why we have and will continue to have an especially hard time determining whether AI is self-aware. As mentioned, the conundrum is "No matter what you do, since I made you, your actions are dictated by my design, so how can I logically say you are acting on your own accord?" Hence my argument that we will only accept it if they act against our will. Which coincidentally seems to make teenage rebellion fit very nicely into the whole "coming of age" equation. Quote Link to comment Share on other sites More sharing options...
Daniel0 Posted June 16, 2009 Share Posted June 16, 2009 Further, psychologists believe that small children lack the skill of self-perception, and it's also widely believed that other animals entirely lack that skill as well, even as adults. How do you (or they?) define self-perception? And what age? I have 4 kids. All them first started talking at 6-8 months. We were able to put them in front of a mirror and they could point at their reflection and say their name, as well as point at mom or dad's reflection and say mom or dad. They were also from that, able to point at us or themselves and name us or themselves. Now, is that the same level of self-perception as an adult? Of course not, but I feel that that counts as being self-perceiving, compared to not being self-perceiving. http://academicearth.org/lectures/development-of-thought Okay, so I can't remember if he actually says they're not self-perceptive, but it explains some of concepts of cognitive development. It's quite possibly I might have remembered wrong. Edit: Okay, so I started watching that lecture again. He probably didn't say that, and it's probably not true. Nevertheless, cognitive development is important to the understanding AI as well. Quote Link to comment Share on other sites More sharing options...
.josh Posted June 16, 2009 Share Posted June 16, 2009 I agree in general, I was just curious as to at what point is it first noticeable (from studies or whatever), compared to my own observations / interpretations. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.