MY THOUGHTS OF WHAT I THINK ABOUT AI AND HOW I DEFINE IT: I know there is Neural Networks who communicate with each others, so we all can get the feeling of that they are "thinking", but in my opinion(remember that all this is related to what I know and think, so there's no reason to blame my facts cause it is my picture of it) thinking is do what our brains can achieve not reading through a thousands of different loops or tests (for, while, do if) cause that is to restrict to calling it "thinking". I would like to say that I think the differences between the Computer's way to think and the human's way to think is this statement: "Human cannot count fast, computer can, but human can think, computer can ONLY follow instructions". That means that I computer can only "think" what it is told to think about, so the computer is still not intelligence. But then the problem is, what do we define "intelligence", my brother defines it as "someone or something that make decisions". But should we define intelligence even things that "makes decision out of what it know", of course we do that, but we can twist and throw around our thoughts which I believe computers can't and will never be able to do. Of course they can throw thoughts to each other through Neural Networks, but never understand what they are trying to solve if they don't already know it! Which makes me to believe that so called "AI Trainers" program cannot learn so much because they got no info on what to learning. An example over why I don't believe in that Computer will be able to "think" as us: Simple equation: 5x + 6x, now my calculator cannot count with symbols but it still costs 80$ But of course we could program a program who could count with two int's and two char's but what about this equation: (2y(5x + 5xy)-4yx(34x*5y)^2)^2 Now we can easily solve this equation by looking at it (now can a Computer read a line and understand what should be done first? It must know already! It doesn't think then, only look through what it knows!) first and we follow the rules in how you solve equations and follow the prior orders. Now I know I maybe talked a lot of shit. But one way to prove for me that a Computer can solve this kind of Algebra equations, then program or send a link to a program who can, and who can solve any big equations by just typing example: 5y(5x+5xy)-5(x-y). If it can I will take back anything that had to do with equations-solving. Now just last thing, I did just express my thoughts I dont know much however I can learn, and still better than computers I think. And this were my thoughts about AI, share yours, remember I didn't try to MAKE YOU BELIEVE WHAT I BELIEVE ABOUT IT! Greetings , Max "Beginner" K, Söråker in Sweden.

Check out Matlab:
http://www.mathworks.com/

And a less than 50$ calculator that does symbolic calculus:
http://en.wikipedia.org/wiki/TI-92

Quote: Original post by Maxkarlstedt
(2y(5x + 5xy)-4yx(34x*5y)^2)^2 Now we can easily solve this equation by looking at it

That's not an equation. But this is an equation:

(2y(5x + 5xy)-4yx(34x*5y)^2)^2 = ∂x/∂t + ∂y/∂t

Can you solve it by looking at it? Or, more importantly, why would you try to solve it?

I believe the reason why computer programs are no match for the problem-solving skills of humans is simply because these skills are unacceptable. Humans work through heuristics to reach an almost correct solution for small problem sets that are not too degenerate. On degenerate problem sets, or some that are too large, humans give inaccurate and wrong answers. Who would want a program that can give such answers?

The expectation of always being right is what drives program development. The only reason for which humans appear intelligent, and can do things machines cannot, is because it is acceptable for a human to be wrong (or incapable of answering) in almost all cases, while it is not for a machine.

Which is the question, what is AI? or what is thinking? Many people, I think, take the name artificial intelligence a bit too far, nobody (ok, maybe just almost nobody) claims that current AI is the same as thinking, from a human stand point. So what is AI? As one of my professors teaches: "AI is search". AI is basically better ways of helping computers find solutions to problems that work a little faster than depth-first search. Often, depending on the problem and the method, the solution can be something counter-intuitive (according to humans) and surprising, which is what gives AI it's "magic".

As far as real thinking: Searle (sp?), Ryle, Dreyfus, and all the rest can bite me. I don't care or like what they have to say, it's just too annoying :).

But like you said, this is all just my opinion too. :)