Can Computers Think?

This is a question which I've often thought about and considered - is it at all possible that anything that a computer will do today, or even 100 years from now be considered thinking? Can AI ever truly exist, or is a computer always going to just be that box at your feet?

One view is that AI can never truly exist and this viewpoint is quite well argued by the "Chinese Room Argument".

The way I see it, however, humans don't really do anything specific to prove that they can "think" any better than a computer might be able to: we have a sort of base programming and evolve and learn based on it. We can't stray from what goes on in our mind, so we are stuck with whatever is there. How are we any different from a piece of software?

Really interesting question, and I figured I'd get a little input from the tech community and see what you guys think

The thing about computers is that all they really do is process information. So if you are wondering about computers as we know them today, there is no possible way they can truly think and make decisions for themselves. They are given instructions and merely process the information for us which has absolutely nothing to do with thought.

When comparing us to a computer, is we have the ability to absorb and learn from what goes on in the world around us. Although computers can seem to do this with certain peripherals and software its highly technical and all they are really doing is having a code to react a certain way when a certain event takes place.

Now it may be possible to develop this to a point where they will be highly sophisticated and be able to do a lot more things, as a result of "knowing" what to do in a certain event, but they can never truly think about it and make a decision. I highly doubt it is possible to make AI especially as in seen in movies. One last thing is how can true intelligence be artificial. It is something where a person must first have the capability to retain information, and they also must learn it (which computers can both do, yes) but they also have the basic instincts that every human already has that is not something that can be programmed into a machine.

Just my quick opinion.

The closest thing we have for computer making decisions is artificial intelligence (AI) or expert systems. This is basically a decision tree that may be very complex but all the outcomes have been thought of in advance. I imagine that a lot of call centers use either or printed or computer version of this.

Another limitation computers have is vision. Computer may be able to calculate much faster than humans but they are slow at recognizing what is in front of them. Some systems use fast Fourier transforms (FFT) for digital signal processing to speed up computer vision for decision making. The military is very interested in computer vision because in a crowded battlefield it would be valuable for a missile to be able to pick out an enemy vehicle from the clutter around it.

As bball4ever39 has mentioned learning is one thing that computers are generally not able to do. True intelligence requires this. That alone is not enough however. I do not know the right word but people have background or cultural information to draw on to shape what they say and understand while communicating with another person. That is why it is so hard for computers to translate one language to another. They can do a literal translation but that often produces a lot of nonsense. Computers lack the human "background or cultural information" necessary to make the proper translation.

To see why this is so difficult look at the free language translator called Babel Fish Translation. You can enter text to translate or translate whole web pages sometimes. It is not the best but sometimes you can make out enough to sort of understand what is being said.

It's not all about the AI. Comparing humans to computers has to do with consciousness.

First of all, in the field of cognitive robotics and artificial intelligence fields there's a term that is very similar to human consciousness, it's the "Artificial consciousness" or "machine/synthetic consciousness."

(I'll call Artificial Consciousness as AC later on.)

Second, keep in mind that AC is in close relation with neurology, genetics, cognitive sciences and psychology. The technology currently advances in quantum leaps, everyday new inventions and achievements are reached. As every day passes we realize that we're moving closer to understand "ourselves," human beings. Information processing is very important as we send/receive information, communicate and process the gained information(s). The way we think, we're aware of specific situation(s), living in the "now," and thinking in the future-- planning, setting goals, having desires. All of these, all together form the consciousness.

Consciousness it's one of the most crucial attribute of human being's minds, it's in close relationship with subjectivity, self-awareness, sapience (acting considering the end-results, being aware of consequences-- judgmental thoughts). Basically philosophical areas are very much involved here too and in every other cognitive sciences, too.

To develop AC up to a state where it matches exactly our consciousness it takes a while and first we should at least discover and realize the way we think. Consciousness is very ambiguous and even thought it has traces in almost every science it still it gets much more complex. It's sort of "the more you learn you realize the less you know," because "stuff" get more complex and it gets extremely hard until scientists will get that "A-ha!" feeling and get the whole picture. It was the same with every great invention. It takes time. When that will happen? I'm certainly not Nostradamus and can't know that for sure. In the near future? Not likely.

We're in the informational age currently. The next upcoming era --in my opinion-- is going to be the post-informational and the quantum physics dominated, power of mind capabilities and subconsciousness. In the last 20-40 years amazing inventions and achievements happened regarding subconsciousness and to realize 'how our mind works.' Accelerated learning is the result of those achievements. NLP (Neuro Linguistic Programming) is the same. Also, basically every other science can be "re-traced" to psychics and we develop their meanings on a quantum-level. "Everything all around us is energy."-- one of the most important invention in the quantum physics field.

To get back on topic, in close relationship with AI and AC there is image processing-- the ability to perceive 3D-like images, to encode/decode it, transfer it, analyze it and react depending on its meaning. This is image processing in a nutshell. It's true that currently digital computers perceive and implement data but all of this happens on a different level and it is a bit far from being "self-aware" in a situation.

Awareness being another required factor. Now that seriously is mind boggling. Why? How can a machine, robot be self-aware. It can be... but the problem is that the thesis itself as "awareness" is a very complex and ambiguous (this one too) process. This again gets drained into the philosophy. With awareness there is in close relationship the anticipation-- that is also a very must-needed crucial factor for total consciousness. Also, we can't forget that we, humans have got intuition (again a subject that's debatable and must be proven).

To achieve total consciousness basically the "robot" must be instructed with a model which must be similar with the real-life. A model, an algorithm, series of processes and images that's going to be effectuated and run whilst powered on. The machine would compare and analyze perceived images with the ones that are currently on its storage then compared with its very own model and ultimately calculates an anticipation feeling. After all of these the machine could be somewhat consciousness, up to a specific degree-- let's say 80%. These robots would get bug-fixed on a daily basis and in a matter of time, the process would get almost perfect-- close to 99.99%.

I seriously don't think that AC can be acquired on a full, total, 100% level. We can get close, perhaps in the near future. Until then a helluva (sp?) new inventions must be achieved and first we need to understand ourselves.

Thanks for reading. Keep the discussing flowing and come up with new ideas.

it is nice to talk about consciousness but that something that is very nebulous and ambiguous. The problem is that scientists are a long way from truly understanding how our brains work and process information. When they figure that out then they have a better chance at mimicing that with some kind of computer intelligence. Consciousness or awarness is something that probably will not happen in any of our lifetimes because right now they are not something that anybody knows enough about to create a computer that can or will develop it. To do this they have to modeled mathematically. The problem is how do you model consciousness or awarness?

Yes, that's exactly on what I've tried to emphatize.

Another important factor is emotion. We, humans tend to think, judge, react, perceive information according to our emotions.

When we're happy, we view situations and events from a different angle, basically we've got a different point of a view, an euphoria-like state (which gets emulated by 'getting high' via drugs). When we're angry, we tend to react angry and pissed of even towards people we love; we refuse opportunities, we eventually react in such way that we later on regret. When we're in a love, it's a certain possibility that "mental distortion" happens; we end up viewing, noticing and seeing our loved ones differently-- others say that love "blinds," whilst the process is mental distortion that happens due to our emotions, our high frequency of feelings and we "do" that to ourselves, we consciously (or not) "fall" in love.

Emotions have a great role in influencing the way as humans react and live their life. We're emotional beings, our most important desire is to 'be happy.' Happiness is a complex feeling that is built up by a few other feelings and emotions as well: to love and be loved--affectionate feelings, to be safe (that includes financial independence--money that is often over-rated but it can't 'bring' happiness) and the need to succeed. First one is obvious (love), second one is again clear (money; shelter & food-- the basics), third is a bit complex. It's that we yearn to succeed in an area, to be famous and successful. No matter how introvert or shy a person is, he/she has desires to succeed and to be the best in his/her specific area.

Thinking about these it's quite hard to model them. It somewhat boggles the mind and I'm so close to state that this can't be possible but I won't say it. Everything is possible. But the big question is when?

I do agree with your statement that it won't happen in neither your or my lifetime. Those subjects are too vague and their development, research and modeling is "far away" located in the future (if ever).

Also, great addition, pointing out that modeling must happen in a mathematical way.

If computers could think IT guys would not be needed.

Remember a computer is only as dumb as the user that uses it.

madhyena hit the nail on the head. a machine can only do what it is programmed to do. you can give a computer an extremely complex program to cover heaps of pre-thought conditions, ccd's can even be used to attempt de-coding body language, microphones can be used to analyse pitch & amplitude of the voice, infrared heat sensors to detect body temperature, a pc can do a good job like this at analysing and detecting a humans current state, if it's stressed, happy, etc... but....

the accuracy of the analysis depends on how well the program has been written PLUS some people react differently to things, we aren't robots, we are all unique!

this is just the input stage, as for the output...the computer can only say or do what is has been programmed to do and what peripherals it has set up to do it. again, the quality of the program written needs to be pretty damn good to have a pre-calculated response for every situation.

this is where being conscious comes in. a human be faced with a new situation and figure something out. show a computer something new, which it has not been programmed for or have the hardware to support and you get BSOD

In a word, No.

that turing test supports my argument that a machine can only do what it is programmed to, it has no consciousness or intelligence & cannot make decisions. only input, calculations and output. not sure where you got the no from?

you think "random" on a cd player is truly random? that the cd player just picks a track at random? it's a calculation called a random number generator it's just a clever use of maths and the calculation is programmed via software or even hardware.

you can have heaps of various sensors for inputs and heaps of programming to handle heaps of situations and heaps of output devices, but ai is just that - artificial/not real/does not exist.

Human body is also just a machine that can do only what it is programed to do. It is geneticaly programed. This "program" changes with mutations and "better" version survive. I can see no theoretical reason why machine could not be programmed that way, of course there are many practical reasons (complexity, hardware capability, programming techniques etc.).
Problem is that it is almost imposible to tell how computers will look, act and how will they be programed 100 years from now.

lol no. computers have been used to invent things but a computer has never invented something. a pc simply cannot have ideas. a pc has never had a debate with forum members a pc has never lost it's purse or bloody handbag. a pc doesnt know how to be cute to get what it wants. a pc has never cracked a joke. it cannot do anything but exactly what it has been programmed to do.

us humans however can do all of the above as we can have ideas, we can think of things, we have imagination. ive seen a quote from einstein that imagination is more important than knowledge.

I strongly disagree.
One of the main reason that AI is not primary target in IT is market. There is not enough money in that.
Let me give you example. Chess. Chess is highly creative game, especialy at high level. It needs to invent new things (moves) and it is realy not all about calculation at highest level. Today computers can play at that level. They stil are not good as humans are in chess but they are getting closer and closer and it is just a metter of time (years) when humans will be not able to confront machines at this highly creative game.
You don't believe me?
What about art?
T1here are computers that are creating art too. Take fractals for example. OK, you can say that they need imput for that and you are right BUT, humans also needs input to create something.
I am telling you it i9s a metter of time when computers will be programmed to take input on their own will (in a metter of fact it is happenning right now).
There is other inputs on that question and that is antropomorphism.

All in all no matter what bible and Red book fromm Mao say, we are not the only possible way of existanse.
(suggested reading A.C.Clarke "Crusede", this is hw forum after all)

playing chess is not using ideas or inventing. it's understanding the rules and being able to calculate as many possible future moves as possible and this is where the pc has a huge advantage over a human, as we can miss a possible future move but a pc cant forget - it's programmed to calculate each possible move and to retaliate using the move with calculates to have the highest probability of success

You are terribly wrong.
If you are right how is it possible that humans still playing chess better than computers? (I know you, now you will strt talking about Kasparov playing against IBM but that was totaly comertial thin and certainly was not a real match)
Also I am really suspecting that yolu are aware how deep this game is (no offens intended), it is simply not just calculation. There is a lot of "instict" in that. Take for example 6th game in match Fischer-Petrossian. That winning move simply can not be calculated it has to be "anticipated". Chess is not just to calculate mate in 4 moves that is what makes difference between average players and end extremly good players. Compurters are getting close to that.