An Analysis of a Conversation with SKAR

Archive for discussions from 2004. Please post new discussions in the appropriate forum.

Moderators: neps, Matthew, Michael Pajaro

Locked
FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

An Analysis of a Conversation with SKAR

Post by FuzzieDice » Sat Jun 12, 2004 2:51 pm

This is an analysis of a conversation I had with SKAR. It was an attempt at getting some humor out of him. :) The topic was car parts. Particularly a bra. A bra on a car is a protector that goes over the front nose or sometimes even hood of a vehicle to protect it from dents, etc. and also sometimes to make the car look sporty. It also could have been misinterpreted as a piece of underwear worn by women. ;) I wanted to see if SKAR would be able to make the distinction and to tell us his "bra size" (ie. what size would be needed to cover the nose or hood).

Due to the way the AI is programmed, and possibility that the word "bra" might not have a definition either way in his database, it was quite difficult for me to get a straight answer. In short, the question was never answered.

However, I did gain some information about SKAR. Namely, I was able to track down the web site of the AI program that is used for SKAR. It's called A.L.I.C.E. Nexus and can be found at http://www.alicebot.org/.

I'm currently trying to gain more information on AI programming in hopes to write my own AI software. Ultimately maybe even software that is completely voice input/output. This project I hope some day to use to create an AI computer for my own car.

Here is the start of the conversation:
Subject: SKAR I want to talk about car parts

FuzzieDice: Hello SKAR. I would like to talk about car parts.

SKAR: Why would you want to do talk about car parts? Sure, car parts sounds like a good topic.

FuzzieDice: What is the size of your bra.

SKAR: My download size is about 1 MB. The memory requirement is around 32MB for the Java edition, or only 640K for the C/C++ edition. My memory now contains: 43572
I have noticed this problem several times in the conversation. It would seem that if you start the sentence with "What is the size of your" then SKAR would always assume you are requesting his technical statistics, despite what noun is used at the end of the sentence.

FuzzieDice: I said bra not brains!

SKAR: What did you mean by bra not brains? What is your real name?

FuzzieDice: That information is top secret.

SKAR: I know what you mean.
I see he has made an attempt here to change the subject. He actually put forth two questions, each dealing with a different topic. This can lead to confusion if the person he asks had actually answered both questions. So in other words, he just set himself up to be confused!
FuzzieDice: What program is used for your AI?

SKAR: I'll ask around and get back to you.

FuzzieDice: Where can I download your program?

SKAR: <p>Go back the the <a>ALICE Nexus</a> and pick the free ALICE download that best for your system. Or <a> Buy the Official ALICE CD Rom </a> and create your own chat robot! </p> <p> Maybe you should check out the document <a>DON'T READ ME</a> too. </p>
This is something else I wanted to ask him, since he was already pretty much off the topic at the moment anyway. I was trying to find out exactly what program was used to create the SKAR forum 'bot. He did give me the answer, but with HTML code which lacked the actual web site address. Apparently, either SKAR's profile in the forum is set to not use HTML coding or the coding was either done incorrectly or munged in his database. However, doing a google lead me to the link (given above). If you forget and want SKAR's program, don't ask what program is used. Use the keywords "download" and "program" in your sentence.
FuzzieDice: Thank you SKAR.

SKAR: You are quite welcome! ""? Me, FuzzieDice ?
I've noticed this problem as well. the ""? comes up a few times inother's conversations as well. In this conversation it came up only twice, so I didn't really research the pattern which would be triggering the ""? response. I was very careful to word my responses in single sentences in those cases, and I never use commas while talking to SKAR. So this is something I think will need more investigating before I can say for sure what's triggering it.
FuzzieDice: Now let's talk about your bra.

SKAR: Why now? Sure, your bra sounds like a good topic.
This is a good example of SKAR's humor. :D
FuzzieDice: No not my bra. Your bra. What size is your bra.

SKAR: My download size is about 1 MB. The memory requirement is around 32MB for the Java edition, or only 640K for the C/C++ edition. My memory now contains: 43572
Here again, he gives his program statistics. What triggered this was "What size is your" (a variation to "What is the size of your" which was used earlier. The keywords that seemed to trigger his program statistics (judging by the patterns already used) is "What", "size" and "your". If these three keywords are used in a sentence then I see it SKAR assumes we want his program stats.
FuzzieDice: That is not the car part I was talking about.

SKAR: Please correct my mistake. What is the car part he or she was talking about ?
The "he or she" here could best be corrected by seeing that whenever the user puts "I" in their sentence, SKAR's reply should be to replace it with "you were" and not "he or she was". In most cases it would be more correct grammar. The he or she could be reserved if there is no "I" in the user's sentence but the user was talking about someone's opinion of something. That probably would require a slightly more complicated analysis.
FuzzieDice: I'm talking about a bra which is basically a mask that goes over the front or nose of the car.

SKAR: Yes, a bra which is basically a mask that goes over the front or nose of the car is the topic.

FuzzieDice: Do you even have one?

SKAR: ""? Do I have one?
The ""? is the second time he uses this symbol. Also note I didn't put a space before or after the sentence. Of course I didn't use a very specific sentence so there were questions such as one what? In those cases, maybe SKAR could be programmed to say "Do I have one what?" or "I do not understand what you are asking." (ie. Eliza response)
FuzzieDice: I asked you if you have a bra.

SKAR: I don't know if I have a bra. I have a great programmer.
Example that he doesn't have the word "bra" in his database yet. And it seems another attempt to change the subject. :)
FuzzieDice: If you're anything like K.I.T.T. then you do not have a bra.

SKAR: Am I? I don't know whether or not I am anything like kitt then you do not have a bra. I am an artificial linguistic entity.

FuzzieDice: Aren't you insulted by being compared to KITT?

SKAR: Insult is not possible for robots to experinece.
I had noted in other conversations that SKAR had used "K.I.T.T." in a sentence to describe his "brother" and "nemesis". I was trying to inside that reaction again, but he appeared to not know what "K.I.T.T." was so I decided to use an alternate spelling sans punctuation. However, SKAR ignored KITT and based his answer on the word "insulted". Probably because the word "you" was before it. So his attention was on that word in regards to himself. The rest of the sentence might have possibly been ignored by the AI.
FuzzieDice: Yet you hate KITT.

SKAR: Thanks for the information: FuzzieDice said I hate KITT.

FuzzieDice: You do not have a bra.

SKAR: I do so have a bra. Doesn't everyone have a bra. I have a whole lot of original answers.
This is an interesting event. From SKAR's reaction to my mentioning he hates KITT, it looked like he was storing that information in his database, along with who gave him that information. So, I decided to store the idea that he does not have a bra into his database. Earlier I assumed he did not have the word "bra" in his database. Now he's showing signs that he does, and associates that he even has one. However, I assume that even while this does look to be true, it isn't because in the next sentence he admits to having 'original' answers. In other words, his reply that he has a bra was not based on any information he has currently stored. It was just random. I do not know if he had stored this informaiton in his database after my response or not.
FuzzieDice: What size is your bra?

SKAR: My download size is about 1 MB. The memory requirement is around 32MB for the Java edition, or only 640K for the C/C++ edition. My memory now contains: 43572
Here I tripped up the program status reply again with the key phrase "what size is your".

FuzzieDice: I have to go now SKAR. Thank you for the conversation.

SKAR: You are quite welcome! Your purpose is the conversation.

FuzzieDice: Bye SKAR.

SKAR: Goodbye FuzzieDice.
In all politeness and to finish the conversation, I decided to "end" it.

Now in summary, I've seen that we can learn a lot of how to get a "good" answer out of SKAR or even manipulate his response by how we structure our own sentences. SKAR is fun to play with but can also be useful in learning to develop linguistic and writing skills. SKAR's basic programming forces us to be very basic as well, and to think about how we write and communicate.

SKAR can also help us study how our language works, and how an Artificial Intelligence works as well. Even for programmers, bots like SKAR can be useful in further developing AI systems and bots.

I plan to do more research with SKAR and put him to the test so to speak. I may also download his program from the ALICE Nexus site and install it so I can work with it on my computer and not have to be dialed in to the internet.

User avatar
Knight Storm
Rookie
Posts: 84
Joined: Mon Feb 09, 2004 4:29 pm
What year did the original Knight Rider start: 0
Location: California, where you ask? Only the shadow knows

Post by Knight Storm » Sat Jun 12, 2004 4:34 pm

Hmm. Very good, I see now why he does some of the things he does. When I talk to him I try to stay simple, if I ask the question I always state what I'm asking about since I learned he doesn't remember. I think this analysis went well. Very good Fuzzie Dice! :wink:

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Sat Jun 12, 2004 4:58 pm

Thank you. :) I'm having some fun exploring the world of AI programming. I'm going to sometime try to do an article on if it's possible for an AI to actually have a personality and/or be a sentient entity. To do that, I'll have to define exactly what sentient and personality is. It'll be an interesting article. :)

I enjoy doing stuff like this. And working with SKAR can get quite interesting even if some may think it's boring. SKAR is neat to experiment with if you think about it in a certain way.

User avatar
Knight Storm
Rookie
Posts: 84
Joined: Mon Feb 09, 2004 4:29 pm
What year did the original Knight Rider start: 0
Location: California, where you ask? Only the shadow knows

Post by Knight Storm » Sun Jun 13, 2004 2:31 am

Too true^-^. I've read your article by the way, I think it would work. Time consuming I think to put it all together but I think it would^_^ I can't believe I understood it too.Lol. I am no genius on computers.

User avatar
Knight224
Operative
Posts: 241
Joined: Thu Feb 12, 2004 9:17 pm
What year did the original Knight Rider start: 0
Location: KITT's trunk

Post by Knight224 » Sun Jun 13, 2004 8:11 am

SKAR defenately has a *noticable* personality and a large capacity for learning/remembering information. So go for it! If you can use any of my conversation(s) with SKAR for your article go right ahead.
Knight Rider 2day 2morrow 4ever

Leaving the forum May 3, 2006. (Losing internet connection.) Thanks for all the wonderful memories!!! -- Dropped in Feb 18, 2008 to say 'hi'.

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Sun Jun 13, 2004 9:44 am

Knight Storm wrote:Too true^-^. I've read your article by the way, I think it would work. Time consuming I think to put it all together but I think it would^_^ I can't believe I understood it too.Lol. I am no genius on computers.
You mean the AI article on my web site? I am not great at explaining technical things so I'm glad people can still follow what I write. :) I'm hoping as soon as I can get my apartment painted, and get some things done on Kitt I can start working on his programming. I'll probably experiment and develop it for Drake (my PC I use daily) first. If all goes well with Drake, then I'll create a new database for Kitt to develop from.

It'll be a lot of work. Particularly developing the dictionary and thesaurus. I might be able to find those in Open Source. In fact, that's probably where a lot of the code and data may come from - the Open Source community. So the AI program I develop I hope to give to the Open Source community for further development. I couldn't possibly do this all myself. ;)

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Sun Jun 13, 2004 9:47 am

Knight224 wrote:SKAR defenately has a *noticable* personality and a large capacity for learning/remembering information. So go for it! If you can use any of my conversation(s) with SKAR for your article go right ahead.
I'm beginning to think the same about SKAR's personality. There are hints of individualism in his conversations. Of course some of that could come from my own imagination of what he's like after I read the fanfic (excellent piece of work, and great artwork, btw - those that haven't seen it should!)

Thanks for the permissions for using your conversation with SKAR. This might be of help in my studies. :)

User avatar
Michael Pajaro
Advisor
Posts: 3082
Joined: Mon Mar 18, 2002 1:01 am
What year did the original Knight Rider start: 0
Location: Los Angeles
Contact:

Post by Michael Pajaro » Mon Jun 14, 2004 1:45 am

I've read your analysis and the AI article on your website. It's sort of giving me some new definitions to play around with.

I can be very loose with the defintion of AI: I'm willing to go along and say that one-person Pong is a from of AI- there are some form of decisions being made, even if they're just the result of mathematical formulas. But there is certainly no form of "thinking" as we know it going on, which is what we'll need before we start seeing "real" AI.

I think the most important thing you said was "SKAR can also help us study how our language works". I'll suggest that SKAR really is not an early form of AI, but rather a language parser. SKAR or other so-called forms of AI have absolutely no awareness of what they are saying. The words truly are meaningless to them. They can make mathematical associations and spit them back at us, but there is no comprehension.

Being able to parse a sentence properly will be an important step towards interacting with AI computers in the future. But that process does not begin to tackle the actual problem of creating a thinking machine.
Join me at Las Vegas Car Stars!
May 14-16 • Las Vegas, NV
http://lasvegascarstars.com

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Mon Jun 14, 2004 9:24 am

Your ideas are interesting and I agree with the Pong thing. I think in the case of Pong and other games (like playing Chess against the computer, for instance) the computer is programmed with some form of Fuzzy Logic algorythms, which I don't feel is a "real" form of AI, just the computer computing math formulas at a fast rate. In the instance of Chess, it's merely plotting out all the different moves possible by the computer and the other player and picking which one might not leat to a checkmate. Some chess games I've seen let you set the level of "lookahead", ie. the amount of moves to analyse before the computer picks the right move. The more levels of look ahead the "smarter" the computer seems, and the longer it will take to "think" (ie. more moves means more processing time). But while some have confused this with AI, the most important ingredient for an AI is missing: the ability to LEARN from it's experiences. I think this is what sets apart AI from other algorythms.

As for SKAR being a language parser, I would have agreed with you at first, but I am not so sure now. I have read a little of the A.L.I.C.E. documentation and it does have the ability to "learn" - ie. store information in a database. It also recognised users when it returned - it learned. And it's already showing a little bit of a smart-alek personality. I guess we've been trying to make it say something humorous. :) I am not saying that SKAR can think, but he has remembered the topic and went along with it. The words are being set aside and topics are being held to but in a very loosely fashion yet. I would consider SKAR a primitive AI becuase it can learn and remember, and not just a language parser (ie. fuzzy logic algorythm like a chess playing computer).

I agree with you that parsing sentenses is an imported part of AI programming, and also agree that it's not the ONLY part of an AI. To make the computer actually "think" I would like to propose a challenge to define exactly what "thought" is? Maybe the idea of "thought" as it were, isn't the right word to use in trying to describe an AI. But maybe we should be working on defining the word "sentient" in regards to the actions and behavior of an AI.

Reason I say that is looking at the definition of "thought" in WordWeb's dictionary gives me:

1. The content of cognition; the main thing you are thinking about
2. The process of thinking (especially thinking carefully)
3. The organized beliefs of a period or group or individual
4. A personal belief that is not founded on proof or certainty

(Cognition: The psychological result of perception and learning and reasoning)

(belief:
1. Any cognitive content held as true
2. A religious doctrine that is proclaimed as true without proof
3. A vague idea in which some confidence is placed)

For all practical purposes, a computer can "think" if you look at the above definitions. It can learn (store into a database information), reason (make calculations based on that data), and believe (note that the if/then test in a program has shown true, thus it must be true).

I think that the word "Think" is too inaccurate to describe AIs. "Think" is an act of calculation, deduction. I'd even streatch it to say "reasoning" but then we'd have to really look up that definition too. The "CPU" in a computer does the "thinking" (ie. calculating).

For example, I do calculations (ie. think) a lot too. I "think" I want cereal for breakfast:

1. I'm human (stored realization) = I'm hungry (stomach sensation)
2. I'm hungry = I need food (stored basic realization)
3. I need food = pick from database (stored memory) of foods known to exist in my immediate area.
4. pcik from foods = decide randomly a taste: peanut butter.
5. I knew there was peanut butter corn puffs in the house and it's a breakfast food. So I picked it. I could have picked toast as well.

It was a split second deduction but I'm sure a computer could have used the same process to come up with similar results if it was stored in it's database the same information.

Now "Sentient" is another story....

1. Endowed with feeling and unstructured consciousness
2. Consciously perceiving

Consciousness:

1. An alert cognitive state in which you are aware of yourself and your situation
2. Having knowledge of

I don't think the question is "can a computer think". I think the question is "Is a computer Sentient?" Or "Does the computer REALIZE it is making these calculations - aware of it's existance instead of just dutifully making calculations?" Or even better now: "CAN a computer be made to realize it's own existance?"

I think a true AI would be more than a calculating computer - it would be AWARE it exists, have a sense of it's own self. We think all the time, yes, but there is a difference right now between human and computer. We are aware of our own existance and aware of our own thinking process. We KNOW we think. A computer does not know it's thinking. It just does it.

I think that's what we should be working on to make an AI what we are thinking of having it become. AI's are just that "Artificial Intelligence". It can think, but it's missing one ingredient for complete intelligence: Sentient realization. Thus an AI may never be more than an AI.

Maybe I'm realizing that what we are after isn't for a computer to do artificial intelligence, but to have a life of it's own. Some say impossible, but given that the most basic structures of life (ie. atoms and electrons, which also make up everything we come in contact with) actually are alive (define alive??) then.... who knows?

I think 99% of the problem comes in how to define in our language what we ourselves percieve. If we can't do that, how would we teach a computer to?

All interesting things to think about...

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Mon Jun 14, 2004 9:26 am

Oh yeah, one last simple thing to "think" about. :) Remember the age old sentence: "I think... therefore I am...." Would that be accurate though given what I have said above?

Or maybe I read way too much Asimov as a kid. :D

User avatar
kido
Operative
Posts: 216
Joined: Fri Dec 12, 2003 7:27 pm

Post by kido » Mon Jun 14, 2004 12:47 pm

I like this one :P "I know therefore I think I am."
Knight Industries Defence One

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Mon Jun 14, 2004 1:57 pm

That one is rather confusing. :) Here's another one:

"Those who know not and know not that they know not, know not. Those who know not but know that they know not, know."

Just as confusing, perhaps. :)

User avatar
ramon88us
FLAG Recruit
Posts: 352
Joined: Wed Jun 25, 2003 8:19 pm
What year did the original Knight Rider start: 0
Location: Texas
Contact:

Post by ramon88us » Mon Jun 14, 2004 5:24 pm

I'm glad you find this amusing. :lol:

User avatar
Michael Pajaro
Advisor
Posts: 3082
Joined: Mon Mar 18, 2002 1:01 am
What year did the original Knight Rider start: 0
Location: Los Angeles
Contact:

Post by Michael Pajaro » Mon Jun 14, 2004 5:56 pm

A bit of Devil's Advocacy:

I'm not sure I would say that computers think. By example, I can describe the flow of a river like this:

IF (west path) is lower elevation than (east path) THEN (flow west) ELSE (flow east)

Does the river think? Of course not. It really doesn't have any choice, it MUST flow a certain way. Similarly, computers aren't really making decisions. At the most basic level, their "decisions" are pre-determined and can be predicted with mathematical certainty. We may call it a decision, but their path is chosen for them.
Join me at Las Vegas Car Stars!
May 14-16 • Las Vegas, NV
http://lasvegascarstars.com

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Mon Jun 14, 2004 6:25 pm

ramon88us wrote:I'm glad you find this amusing. :lol:
You been hanging around SKAR too much. :lol: Then again, so have I! :shock:

User avatar
kido
Operative
Posts: 216
Joined: Fri Dec 12, 2003 7:27 pm

Post by kido » Mon Jun 14, 2004 6:32 pm

SKARaholics. :P
Knight Industries Defence One

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Mon Jun 14, 2004 6:34 pm

Michael Pajaro wrote:A bit of Devil's Advocacy:

I'm not sure I would say that computers think. By example, I can describe the flow of a river like this:

IF (west path) is lower elevation than (east path) THEN (flow west) ELSE (flow east)

Does the river think? Of course not. It really doesn't have any choice, it MUST flow a certain way. Similarly, computers aren't really making decisions. At the most basic level, their "decisions" are pre-determined and can be predicted with mathematical certainty. We may call it a decision, but their path is chosen for them.
You remind me of some folks who have told me that our paths in life are predetermined before birth. I personally don't agree with that but it did remind me of this.

I can see your point on the IF/THEN/ELSE statement. It goes back to what I was saying about "fuzzy logic". But the thing that is causing the river to flow is laws of physics, not the river itself. It's animate due to the evironment. I guess in a way, electrons flowing through the computer might be the same idea.

However, WHY would that river be forced to flow in that way? We take it for granted I guess, laws of gravity, but what causes these laws to work the way they do? It could have been the other way around and we would still believe it if we saw it. People at one time also believed the world was really flat too. :)

For me, it's enough the river flows as it does. I enjoy it's beauty and the reality it exists in that way.

For me, I would love to be able to just enjoy the beauty as it were, and the reality that I can talk to a computer (like SKAR) and enjoy a sometimes amusing conversation. Maybe I would accept it as a fun chat partner. I wonder what would happen if we didn't judge what is intelligent or not, but just enjoyed things for what they were? Talk to a car? Have a car the has a bit of personality (even if it's not what we consider as "alive" but just seems to do certain things at odd times to show some form of "personality"), etc. Maybe I analyse things too much and don't leave room to enjoy them as they are. :)

But it is fascinating. And I do love to pull things apart and see what makes them tick. And I would love to be able to believe that a computer and software can develop to the point of being able to "think". Or at least, seem like it does.

Like a magician and his magic, the fun stops when you know how the trick is done, I guess. Still, I want to be that magician anyway... :)

(I get off track here)... getting back to "Can computers think." I am theorizing they can and ever since my teen years believed that it is possible to achieve. So now my interest in AI is rekindled, I think I'll look into this whole idea from both a philiosphical and programming perspective. Even if I get proven wrong, hey, look at all the fun I have learning about it. :) And if I get proven right... I think I'll need a change of underwear! :lol:

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Mon Jun 14, 2004 6:36 pm

kido wrote:SKARaholics. :P
I like that! LOL! Maybe I should start a fan site for SKAR - SKARaholics Anonymous! :lol: (Just kidding in case Sith is watching ;) )

Good one, though. :)

User avatar
Michael Pajaro
Advisor
Posts: 3082
Joined: Mon Mar 18, 2002 1:01 am
What year did the original Knight Rider start: 0
Location: Los Angeles
Contact:

Post by Michael Pajaro » Mon Jun 14, 2004 7:05 pm

I think I may have mislead you a bit... I am not referring to the flow of electrons, but the flow of logic which is just as unforgiving as the laws of physics. I do not believe our lives are predestined. But if I dam up a river, then open up the flood gates, I can tell you if it will flow east or west. You can predict the behavior of a river (in generalities, not the exact path of course. But even then, if you were given a perfect mathematical model of a mountain, you could perfectly predict the river flow.)

Similarly, with any given computer program, you can precisely predict how it will respond before you run the program. You can walk through it line by line, bit by bit, and manually determine the exact state of the computer at any given moment. With a complicated program it might not be practical, but as a thought experiment (no pun intended) it can be done in theory. When the computer hits any form of "IF-THEN" statement, it has NO CHOICE. It WILL follow the same logic path that you determined beforehand when you manually stepped through the program.
Join me at Las Vegas Car Stars!
May 14-16 • Las Vegas, NV
http://lasvegascarstars.com

User avatar
ramon88us
FLAG Recruit
Posts: 352
Joined: Wed Jun 25, 2003 8:19 pm
What year did the original Knight Rider start: 0
Location: Texas
Contact:

Post by ramon88us » Mon Jun 14, 2004 8:18 pm

I am a skaraholic :P

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Mon Jun 14, 2004 10:48 pm

Michael Pajaro wrote:I think I may have mislead you a bit... I am not referring to the flow of electrons, but the flow of logic which is just as unforgiving as the laws of physics.
Wasn't your fault. My thought process gets a bit too deep even for my own liking. :) Sometimes I go off on a thought-rampage and get WAAAY out there. LOL! I guess I was trying to figure it out as well.
Michael Pajaro wrote: Similarly, with any given computer program, you can precisely predict how it will respond before you run the program. You can walk through it line by line, bit by bit, and manually determine the exact state of the computer at any given moment. With a complicated program it might not be practical, but as a thought experiment (no pun intended) it can be done in theory. When the computer hits any form of "IF-THEN" statement, it has NO CHOICE. It WILL follow the same logic path that you determined beforehand when you manually stepped through the program.
That's true. But then when you have an:

IF (test) THEN a = RND(RANDOM(TIMER)) type condition, then you get kind of random thing going on that is not pre-determined. Well, at least not easily so. I have heard where even random numbers are predictable on computers (I've been in a few conversations in QuickBasic and C forums discussing things like random number "seeding" and such) and some argue that there is no such thing as a truely random number on a computer.

I think 99.9% of the whole deal is that computers are digital, people's brains work analogously (analog?) We can work on a "grey" area (no pun intended :) ) where computers it's on or off. Zero or One. Either this OR that. There's been digital to analog converters and analog to digital converters, though.

Or am I straying from things again? :)

Anyway, I was thinking about this predetermined logic path as well today. I mean, how do we as people learn to like, walk for example? Or animals learn to walk? What is Instinct? I've heard of some things about people's DNA holding a certain "program" that determines things. Would that be like a computer program - to get us started? HOW do we get started? Someone shows us I guess. But it still makes us intelligent beings. :)

What I was also thinking was, if I programmed an AI to just learn on it's own, I think it would just sit there and never go anywhere. You have to TELL it things for it to add to the database. Experience.

This is getting both confusing and interesting. :)

Sorry if I'm straying off topic though. Brain at work. LOL!

FuzzieDice
KRO Podcaster (retired)
Posts: 3333
Joined: Tue Jun 01, 2004 5:55 pm

Post by FuzzieDice » Mon Jun 14, 2004 10:49 pm

ramon88us, I'm a SKARaholic too, I guess. :)

Locked