Can a machine know?

[katuu]

Hey guys!!

Umm i'm doing an essay for THeory of Knowledge.. and my topic is "Can a machine know?" I was wondering if anyone could help me with some ideas?
I have to discuss the problems of knowledge, and ways of knowing. I might discuss things like, what is a machine and what is knowing. So yea any ideas or comments would be greatly appreciated !

THanks a lot
Rob

Hey, i'm doing the same essay and so i was wondering if you had any information I could use- i'd appreciate it. Thanks

[spidergoat]

Read about Daoism. An ancient Japanese religion which presupposes all matter to be living. Rocks and rivers. You name it. The spirit supposedly inhabits all things like a ghostly second universe.

What would a Rock think about?

I think you are talking about either the Chinese philosophy of Daoism, or the Japanese religion of Shinto.

[KitNyx]

Read up on connectivism.

- KitNyx

[ibnerd]

Hi. I'm in the IB programme and I too am doing that essay. I would suggest reading the novel 'Code of the lifemaker' by James P Hogan. It has a very interesting veiw point on the subject.

[lightgigantic]

Even by the use of algorythms you can get a person who is totally unfamiliar with chinese writing to give a chinese answer to a chinese question - completely bereft of the knowledge of the signifigance of the question and the signifigance of the answer - this is essentially the distinction between processing and knowing

[non-possession]

you guys are mainly thinking about whether a COMPUTER can KNOW something or not, whereas you can relate the question to different areas as well. "Open the window and look around, and not just at the thing that is in front of you!"
since everyone is saying his/her opinions I would like to put my opinion as well.
I am doing the same essay as well (Can a machine know?{IB TOK}), and I was actually thinking of relating this knowledge to 'pain'...so emotion..'Can a machine know what a pain is?' Have you guys watched the film called I Robot with W.Smith? In that film they talk about the machines that are controlled by themselves.
I don't know whether I am saying the right things, so I stop here..haha
What I want to say is that everyone thinks differently, and therefore there is always more than one answer. So don't just look at the computer area.

[cato]

Daniel Dennett.

[mackmack]

edited.

[mackmack]

type out on yahoo: human level artificial intelligence and you got all the answers you are looking for.

robots that think, reason, act, and behave like a human being.

[cato]

yahoo? are you twelve?

[wesmorris]

A machine cannot know, because knowing is to integrate information into ego (self). A machine (as designed by a human) has no ego, no sense of self and thus, cannot know - at least not yet.

While I believe the above to be objectively factual, I do think that machines could eventually be constructed to create ego. I don't think the term "computer" would quite do it justice however. Nor do i think that current computer design (arrays of transisters, resisters, etc.) could allow for it. I don't think ego is simply an algorithm, but is evidence of physics that is as of yet unknown.

[Kron]

How do you know that HUMANS 'know'? Our 'ego' could just be an illusion. How would you define integration into ego?

According to me, all human knowledge is simply information that has been learnt, cataloged and extrapolated upon.

(P.S: What's with the burst of new users in this thread???)

[wesmorris]

EGO is self. It is not an illusion. Of course you can argue that it is if you like, but in that case, there is absolutely no point in discussing it, as you don't really exist to do so. So unless you take that position, EGO is real.

"integration into ego" is for instance, acceptance into one's "world view" (the arrangment of concepts that exist in their mind) that some piece of information is real to the point that there's no reason to doubt it, like for instance "I know I exist".

It is interesting to note that all knowledge is ultimately faith based... or if you prefer... based on the assumption of self. "I think therefore I am" is redundant. "I am" suffices, and it is an assumption that IMO, is pointless to refute. If "I am not", what is the difference? I think I am, and any attempts to think otherwise negates itself. If "I am not", what is it that came to that conclusion? Something did, or there is no conclusion. As such, the "I" in "I am" is a sufficient label to describe what it is that reached the conclusion, unless of course you choose not to be, and then you can just not think or identify with anything... which on second thought, I doubt many people could ever do.

[mackmack]

"A machine cannot know, because knowing is to integrate information into ego (self). A machine (as designed by a human) has no ego, no sense of self and thus, cannot know - at least not yet."--wes

have you read some of my stuff?? i have an entire website on this subject matter and i won't repeat it here. but to point out the simple things, a human being is a machine. we where built by our creator and designed by him.

now, lets get on with the details, i'm trying to build a machine very similar to a human being. this means the machine will think, reason, feel, sense, act, and behave like a human being. this is why they call it human-level artificial intelligence. Alan turing made it up (one of the founders of AI). he devised a test to make sure that the machine has human intelligence.

like i said i already devoted a website to this subject matter but if you wanna check out more about HLAI go to the search engines and type it out. there are lots of sites on this subject matter.

[eburacum45]

The 'Chinese Room' argument doesn't convince me at all. Searle thinks that a person can be trained to carry out instructions which process questions and casual conversation in Chinese so that answers can be generated which imply understanding and mimic the replies of a Chinese speaker, all without that person understanding the language at all.


If the person inside the 'Chinese Room' is able to produce results which consistently convince an observer that the system (Chinese Room plus occupant) really does 'understand' Chinese, then the set of instructions themselves (that is to say the program he or she is following) actually does understand the language.

We have already produced programs which can answer questions and carry out conversation with a human, but none of them are capable of mimicking the responses of a human with true understanding for very long. The reason for ithis is obvious; thse programs simply regurgitate previously written responses in a number of limited iterations; they don't even come close to understanding the questions. To achieve understanding would need true AI, and this could not be achieved using any simplistic system of stimulus and response. Once a system is developed which can reproduce the process of understanding, and demonstrate that understanding consistently, then that system really can be said to display understanding.

I would expect such a system to be very complex, perhaps more complex than a human brain (at least at first); if Searle thinks that such a complex system could be reduced to a Chinese Room arrangement, then that may be correct. To achieve a simulation of understanding in a reasonable timescale would probably require a great number of operators; in this case none of the operators would understand the language individually, any more than the individual cells of your brain understand English; but the system as a whole would.

[wesmorris]

"A machine cannot know, because knowing is to integrate information into ego (self). A machine (as designed by a human) has no ego, no sense of self and thus, cannot know - at least not yet."--wes

have you read some of my stuff?? i have an entire website on this subject matter and i won't repeat it here. but to point out the simple things, a human being is a machine. we where built by our creator and designed by him.

But this is an assumption that is simply unreasonable from the perspective of pursuing comprehension. To presume a creator and its intent is well, far too presumptuous for my taste, but thanks for playing.

[Kron]

But this is an assumption that is simply unreasonable from the perspective of pursuing comprehension. To presume a creator and its intent is well, far too presumptuous for my taste, but thanks for playing.

Looks like there's something we agree on

The human mind is a system. A system of biochemistry, electricity, some mechanics and maybe a bit of quantum mechanics, but a system nonetheless. If you take any individual part of this system (you break the system down into smaller ones), none of these parts will understand anything. They're simply exchanging information with different parts according to set rules. 'Ego' is just an emergent phenomenon.

[wesmorris]

Looks like there's something we agree on

Yah.

The human mind is a system.

Tru dat, and ultimately, thought (like all else that is or will be) is necessarily a reaction from first principles.

A system of biochemistry, electricity, some mechanics and maybe a bit of quantum mechanics, but a system nonetheless.

Tru dat.

If you take any individual part of this system (you break the system down into smaller ones), none of these parts will understand anything.

Hmm... mostly, but applying the term "understand" outside of a system capable of consciousness (which a subsystem alone would not be) is missapplication of the term.

They're simply exchanging information with different parts according to set rules. 'Ego' is just an emergent phenomenon.

This is where I cannot help but to split hairs, as I haven't personally been able to move beyond the exception I see in the statement. Information. Information seems to me to imply some sort of meaning, which again, taken in the context of a subsystem of mind, rather than mind itself seems meaningless. Perhaps you mean what I think of as "data". Information, to me, seems to imply "meaning" as in "impact on a mind"... whereas data is meaningless until analyzed. Nitpicking, possibly entirely incorrect bastard that I am.

Further I think "just an emergent phenomenon" is also quite presumptuous. It might be, certainly. It also could be that the phenomenon is indicative of a missing component to the universal model, or rather that this emergance takes place because of a previously ignored degree of freedom: the abstract, which as far as my rather limited mind can comprehend - has no place to exist in classical "space-time". Certainly space-time seems to accomodate material quite nicely, but any observation or "meaning" associated with [i]relating[i] or modeling the materials therein.. well, where does it fit into space-time. Point to a set of coordinates and I'll buy that it can be occupied with material, but the coordinate system does not allow for the meaning that material might contain (like for instance if there is a functioning mind there).

*sigh*

But it could just be that I'm a bit thick-skulled.

Maybe it just seems to me that when describing the universe, given that the description or model thereof necessarily exists within the system, the meaning (which only exists subjectively) must be accounted for by the model. Rather perhaps, that the model itself (or maybe just that there can exist a model) is part of the model, since modeller of the system is utterly steeped in, and part of the system he attempts to model.

[mackmack]

"The 'Chinese Room' argument doesn't convince me at all. Searle thinks that a person can be trained"--eb

my porgram don't use the chinese room method. its completely new. something i designed for the past 7 years.

first of all, people when they look at my post they think i'm making all this up. then that forces me to give proof. i have posted program designs on how this whole "human intelligence works". i define the consious and what it is. how it works and how to implement the software. go to the site and check it out. (even non AI researchers should know what i'm trying to do). if you still don't beleive in the design codes then try the AI demonstrations i posted. see if the AI programs are intelligent.

there are lots of websites claiming that human intelligence is just too complex and can't be simulated or mimiced. thats a false statement. the human mind is the most powerful computer in the world and we as intelligent species will try to mimmic that type of technology someday. if people are smart enough to create something as complex as an internet or a personal computer then i think they will be smart enough to build a machine with human intelligence someday.

[wesmorris]

link your site then why don't you.