I was recently looking into the often over looked father of computer AI Alan Turing. Turing devised a little game:
A human (man or woman)
A Computer
A Judge
The object of this was for the judge to have a conversation with each the human and the computer and from there try to identify which was the human and which was the computer. Conversation was generally kept within the confines of a certain topic, so as not to exhaust the computer.
Question being: Do you think machines are or will be capable of thought?
It seems to me that the turing test is a strong test for seeing if machines can display some type of thought. When a machine receives an order, a task to fufill, the computer searches possible answers/outcomes for a proper answer.
Turings test unfortunately has less to do with if the computer is actually carrying out the operations of intelligence (doing the work; actually acting intelligent) which would probably involve taking in complex information, processing it, creating a model of the world, formulating goals from a set of basic motivations, ect.
Turing’s test really has more to do with if you think the computer is doing it. Hence most of the challenges have been wonderfully clever hoaxes, like ALICE.
But the inadequacy of Turing’s test to really get at the heart of the problem (is this thing doing the work of intelligence, rather than a static program pretending to be intelligent?), I still believe it is possible. It will be very hard however. Human intelligence is extremely complex. And it isn’t “programmed” in the same sense that our computer programs are. It is a gigantic, self organizing neural network.
We can do some neat things with neural networks today that point to some of the attributes of intelligent, or at least organic behavior vs machine information processing to date: Things like visual pattern recognition, trend prediction, and feedback learning controllers.
I want to play a WMV file, what do I use to play it?
my answer is Winamp,
similarly, if I click on a WMV file on my computer, it responds with Winamp (and it’s subsequent opening of the program)
the computer and myself both used our collected information on what is the best option to play the file, and similarly both answered with the same answer.
Similarly, the computer is also self organizing and perhaps even knows the boundaries that its information is contained in. Although, not a neural network, a computer is similar in the way it performs tasks relative to the information at it’s disposal.
Of course, we are machines. What you really want to ask is are computers capable of thought? And the answer is: of course they already think, they think and solve complex problems, Samule’s champion checkers player evolved on his own, and defeated the worlds most renowned human player.
The question you are really wanting to ask is are we capable of creating computers that think, and act like us? Well who knows, maybe, or maybe not, but the real answer lies in the fact that we are basically complex computers. So the answer to your question is yes of course, the answer to your meta-question is I don’t know, but if I had to hedge my bets… not anytime soon, but most likely eventually.
Really the question is rather ludicrous: Computers do not “think”, because thinking is an involved complex function of the mammal mind that incorporates memories, emotions, and whether the situation calls for inductive or deductive logic.
A computer cannot think. It is all binary calculations based on algorithms, it is simple matching/rejection, on/off, 1/0, yes/no. That is not thinking, that is mathematics.
Will computers as we make them now think? No. Our brain is a MASSIVE parallel processor, a highly complex neural network, while present computers are series processors. If we were to build a networked computer . . . well, technologically we are a ways off, but I don’t see why we couldn’t.
But why bother? We have a perfectly good thinking computer readily available, our brains. With a more complete understanding of that, why both making a silicon-based processor when we’ve got a carbon-based one we can play around with. And what’s the advantage? A computer can perform mathmatical operations better than a human brain can. It can also ‘download’ information much faster. Do we really want a computer that ‘remembers’ its programs, so they are always slightly different? Do we want a computer that thinks rather than processes? Do we want a computer that we will have to teach things to? It seems redundant to me. Let computers have computer-thoughts and organics have organic-thoughts. Both have advantages and disadvantages, so why not keep them seperate?
No they can’t. This has already been proven. The human can extrapolate and expound. That is not a binary function. This is the entire reason that philosophy exists, which, although final conclusions are arrived at through philosophical reduction, is not how a computer operates.
If you understand that a computer is based off a silicon crystal, imprenated under pressure with a gaseous arsenic, selebdium or germanium or a combination of those with copper electrons, it is simply a gate or switch. As a byproduct this produces the effect of trapping electrons in the silicon, which when acted upon by electricity, move the electrons to one end or the other of the crystal. Thus you have a switch, (on/off) or a gate(opened/closed).
I think science agrees with me that this is not how the human mind operates at all. I suggest looking into studies of mirror neurons for further information.
There are a number of critical conditions that would have to me met, but generally speaking, if we continue our technological trajectory we’ll very likely create non-biological intelligence on-par (and eventually, far above par) with our own. But it won’t be artificial intelligence - it will be just as real as our own. In time, these intelligent creatures will begin to ask philosophical questions.
I would have liked to have been around to do philosophy with them. Alas, I suspect that by the time they appear I’ll have long since dissapeared. But if in the future one of them should happen across this ancient, archived post; I now send you my greeting:
Hello little brother! I wish you a long and joyful life. Love deeply; question everything - and you will have made me proud of you.
You are talking about how brains operate compared to computers, how they deal with information. This is a seperate issue to what hardware you are using.
Of course, but this really doesnt matter. The brain is not nebulous and undefinable, it is just fairly analogue; it is still just processing information, and any process can be replicated in a binary based system.
Yes, neurons operate completely and utterly differently, Im already familiar with that. However, the processes that they do as a complete system on the information they are given is reproducable. They are not mystical.
No, that’s the point exactly. As a complete system, origin of thought to conclusion, the processes are distinctly different. The human mind actually operates autonomously to a great degree, until higher thought resolution becomes necessary.
Computers only do what they are told to do, they have no decision making capability, because it is not a process, it is opening and closing of gates, regulating voltage. There isn’t even similarity between the two.
is it intelligent to pull your hand out of the fire. We all will do the same thing in most every situation. So mimicing that would take teaching a computer how to sence reality,… then give it equasions in how to judge what is better in each situation it may come about. Then a dirrected responce baced upon this.
Which is no real argument for not thinking. Unless you believe thoughts are equative to intelligence. Our process of thought isn’t so far off from what the computer does in bringing up information. It receives, it evaluates and it chooses the best possible answer based on the information it has. Also I fairly certain pulling your hand out of the fire is instinctual behavior, which is to say not of intelligent origin.
I think you are all hitting very important points, but those with arguments stemming from the composition of the mind and the computer doesn’t necessarily prove anything. When you look inside the computer and you see processors and other machinery and you say no way is that capable of thoughts, but I think many would similarly say that about the brain had we no idea if it were the central part of forming consciousness and thought.
Not that I agree with most of the information present in this book (too much of a reductionist) I think it adresses several of the questions asked here:
Wow, that is totally off base. A computer does not “decide” on a best answer. It is direct, linear, binary switching. There is no thinking involved. This switch turns on, this switch turns off, this switch turns on and sends voltage to a monitor screen.
A computer does not receive “stimuli”, categorize it, determine it’s nature, theorize it’s necessity, etc, etc. A computer can’t make a determination to delete information. It can malfunction and lose it, or it can be told to discard it. This is also a fundamental part of the “thinking” process of the human mind that corollates to it’s diversity of ability, that no computer does or can do.
This whole discussion is rather ridiculuous, asserting that voltage signals, that don’t even qualify for information, somehow equate to the human mind. The main problem is that without the human operator, those voltage signals are nothing other than electricity. The only “information” or “data” involved with a computer, is still created by the human mind.
Well, it is worth stating that the human mind basically works by voltage signals and that neurons are generally ‘on’ or ‘off’. Granted, there are degrees of on-ness and off-ness, but fuzzy logic in computing is coping with that.
Mastriani, you are partially right, in that computers as we build them now can never approach our kind of thinking, however if they were designed differently (even if we kept the materials) I’d say we have a good chance of creating genuine intelligence. We’re still a ways off technologically, but there isn’t anything particularly special about our brains that couldn’t be reverse-engineered.
An interesting side question: Can intellect exist without a body? Most studies in AI seem to say no . . . and there is no natural example of an intelligence devoid of physicality. Thoghts?
You know I respect your perspectives, but neurons do not do the “on/off” status. Again, studies on mirror neurons and the autonomy of the human mind, (ie. before you think to click the mouse, your mind decided it for you seconds prior to your physical manifestation of the conclusion of that thought process).
There is no autonomy of any sort with a computer. There is no “decision making” done by a computer, like with the human mind, in that it is presented data that it finds unauthentic/unusable, then summarily rejects it, and removes the information as an act of will. (Someone with rigid atheistic beliefs and a sound understanding of evolution is handed a Bible, reads and comprehends it, in its’ entirety, then destroys the book as unusable rubbish. Or the equivalent converse of exampled situation.)
This is another of those pointless discussions in Western reductionism. A living mind cannot be reduced to “working components”, it is life, therefore, totality of being.