Jump to content

  • Log In with Google      Sign In   
  • Create Account


Why is so hard to find anything useful?


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • You cannot reply to this topic
67 replies to this topic

#41 Timkin   Members   -  Reputation: 864

Like
0Likes
Like

Posted 19 June 2005 - 06:03 PM

DarkThrone, here are a couple of things you should keep near the forefront of your mind when considering Artificial Intelligence and this forum in particular:

(1) This is a game development website and so, it is not concerned with the creation of intelligence (artificial or otherwise) but rather the illusion of intelligence in artificial systems (namely games). It is not necessary that game AI actually enact intelligence, be it human or otherwise.

(2) Artificial Intelligence is not solely concerned with replicating or explaining human intelligence. Pschology seeks to explain human intelligence, while strong AI seeks to replicate it by artificial means. Many AI researchers and practitioners are interested in the more general question of how to act intelligently. Others are concerned with issues in, for example, knowledge representation andlearning. There are not too many AI researchers out there who are researching consciousness. There are certainly more neuropsychologists and psychologists doing that.

(3) You are correct in that Mathematics is but a tool, just as physics is a tool. It is useful for modelling and explaining our Universe. Mathematics though is a language (for example, it is the language of choice of many physicists when explaining their theories) and within the artificial intelligence community, it is a useful language for conveying ideas, algorithms and even models of 'intelligence'. There are certainly other languages in use (e.g., logic), but you'll find that to be successful in understanding and contributing to AI, you'll need to understand mathematics. (This is not to suggest you don't already understand it... but don't cast it aside as having nothing useful to say about AI).

(4) My final point is this; looking at life around you (other people, insects, animals) won't lead you to solutions in AI. It will merely give you examples of behaviour that you might want your AI to embody. Creating AI is a lot harder than describing AI (i.e., the outward behaviour it creates).

Cheers,

Timkin

(P.S. Note, it's 'Timkin', NOT 'Thinkin'... and you make it sound like I lifted my leg against a tree and 'marked my territory' :P )

Sponsor:

#42 DarkThrone   Members   -  Reputation: 100

Like
0Likes
Like

Posted 20 June 2005 - 12:23 AM

Ok Timkin... (Timkim is your name?) Sorry about grammar. Lot of hours in front a computer can be this somewhile.

Do you know Project Wonder J?

And if you refer to I thanks for your presence, it's because suppose the moderator have more important things to do instead of reply topics of a frustated programmer. (To avoid confusion , ME = frustated programmer).

I just timkin (sorry...i'm joking) thinkin' that a game can be fun with a big enhanced IA. And is not just me, 30% of posts are to NON-related game IA. On true, I post this because all mathemathic formulas are so complicated, and I say this many times. I mean that theory will can more simple (something next to ALGOL) not only to begginners, but also to people that have no free time and need a clear, easy understable theory.

And with your respect and trying to offend you, you should see:

1. It's a game development site, where IA is part of.
2. The forums are separated for topics.
3. People have the right of read what topic they want.
4. Human intelligence is IA related.

And to finish I'd really loved to play this RPG games during 72 hours without sleep. The controller drop of my hand many times (for this was so hard to finish).

Daerax post me useful links, really. With pacience I finally found a way to continue. If you see, many replies query me about the thing I post here. I am just reply them in the form what I know, and with all info I gather. If it's useful, each one can decide by yourself.




[Edited by - DarkThrone on June 20, 2005 6:23:46 AM]

#43 JD   Members   -  Reputation: 208

Like
0Likes
Like

Posted 20 June 2005 - 09:38 PM

We are constantly learning new things about insects. For example, those african termites build airconditioning inside their homes so it's not as simple as dropping dirt randomly. Also, I was watching ants the other day on tv and we don't even fully understand how they communicate chemically and it seems they have entire language we don't get. There is so much to learn that I don't think we have a closed book on it. Math starts to make sense once you put it into good use since it's a tool and every tool not used is useless and soon forgotten.

#44 ahw   Members   -  Reputation: 262

Like
0Likes
Like

Posted 21 June 2005 - 08:12 AM

Darkthrone :

Mathematics, as has been said, are just a tool. But to put it a better way, they are a language, and just like any language, if you learn it well, you will open up new horizons. For instance, take you English. It's a bit broken, but because you speak it and understand it, you can post to this forum, and hopefully learn some things from the exchange. You could always stick to Portuguese and I am sure you could find resources, but because of the extra language, you have access to so much more.
Well, it's the same with Mathematics. I am not terribly keen on the subject, but when I am reading one of those Neural Networks texts and the author starts rambling using Mathematical notation, it comes handy, and it allows me to learn more than if I stuck to simple English.

Now to counter this, and that's something I am particularly adamant, having the eperience of teaching; the fact is that to explain something, you don't need Mathematics, if you are good enough. Maths are useful for modeling a problem, but when you are trying to explain a concept to someone, I don't see the particular reason why one couldn't make the effort and try to explain things in proper English.
Again, it's a question of target audience, I suppose.
Amusingly, I would say the same goes for giving sample code, as opposed to explaining a concept in plain English...
Writing sample code, writing a formula, to me it's just being lazy (as a teacher).

On a different note, I cannot understand why everytime I read about AI (and that has been for the last five years I have worked on this damned MSc) I see so little mention of non AI books. I mean, sure, you get the odd Biology book when the subject matters, but seriously, my general impression of the field is that despite the potential for cross referencing, the norm is more to self referencing...

Personally, I have read a few little gems over my Research that didn't have much to do with AI. But interestingly enough, they were the ones that fired me up the most. Stuff like Animal Behaviour, Swarm Intelligence (how various insect species create their nests, brilliant stuff)...

I even picked _The Selfish Gene_ by Richard Dawkins completely by accident as I was browsing the Biology section of the school Library, and God knows this is probably the coolest book I have read in the last few years.

Another I picked while going through the airport is _Guns, Germs and Steel_, again, you'd think it's completely non related. But the first thing I thought of when I read it was how this could be a great source to write a Civilisation game, but with an emergence twist.

Good God, even browsing through Neurology textbooks from 30 years ago, I had more ideas popping in my head than while reading through Neural Networks books... there are so many aspects of Neurons that are left out when you use a Mathematical Model, it's quite a shocking thing.

Anyway, I am rambling. But I generally agree with DarkThrone, there isn't a lot of really useful stuff around. I know, I have searched for quite some time.
Don't misunderstand me, though. There are quite a few useful people, and people who are adressing this situation. If I had had the books that are out there now 5 years ago, I certainly would have finished that damn MSc... [rolleyes]

Ah well
-----------------------------Sancte Isidore ora pro nobis !

#45 DarkThrone   Members   -  Reputation: 100

Like
0Likes
Like

Posted 21 June 2005 - 02:25 PM

Really are few people helpful ( not useable, that means to some that "We want to use them, instead of keep a little help" ). My English is really broken, I pass last 12 years on a self-learnin journey, with books, music and some basic concepts learning on school.

What really bore me is the fact that VERY complex formulas are used to do it.

And too, only with this post REALLY some help me. Three posts and no responses are given.



#46 ahw   Members   -  Reputation: 262

Like
0Likes
Like

Posted 22 June 2005 - 12:10 AM

Well, if you are self taught, your English is certainly better than my Portuguese [wink]
Still, if you improve it, you will definitely find it useful.
Plus it's always funny to have a better grasp of the language than the average native speaker, which isn't very uncommon once you start frequenting forums populated with illiterate teenagers (no, no, not Gamedev. Well, OK, maybe the Lounge, sometimes [grin] )

-----------------------------Sancte Isidore ora pro nobis !

#47 DarkThrone   Members   -  Reputation: 100

Like
0Likes
Like

Posted 22 June 2005 - 03:29 AM

Well, let me tell a story.

One day, I have a dream. In your homes, people have computers.
People that lives alone had in your computer a good listener, a friend, someone to trust. People that live in family had in your computer a good tutor to your children. Scientists had in your computer a real apprentice.

"I, Robot" and "Ghost in the Shell" cites a theory that I don't know if is real-based or not. "In a system, ramdonly protocols that approach that we called "soul"."

The Ghost in the Shell theory exists on true? What I find it, if exists?



#48 JD   Members   -  Reputation: 208

Like
0Likes
Like

Posted 23 June 2005 - 10:28 PM

Machines can have soul if they are designed that way. Robots of future will have soul designed out since we can't have slaves. Besides they could kill us easily if we give them soul and enslave them. That's too dangerous. So we'll have servants but not emotional ones internally, only simulated externally to be more pleasing to humans.

#49 ahw   Members   -  Reputation: 262

Like
0Likes
Like

Posted 24 June 2005 - 02:01 AM

[headshake]
-----------------------------Sancte Isidore ora pro nobis !

#50 DarkThrone   Members   -  Reputation: 100

Like
0Likes
Like

Posted 24 June 2005 - 09:21 PM

Huh?

Well, I'll take care of my toaster, or it attacks me with a deadly burning loaf slice. Sorry, I have to say...

Take it serious, sorry about point this question. I just say "People have your computer a friend". You are a slaver? You slave your dog? Slave your mother? Slave your girlfriend?

Why you just slave a machine? God slave you?

If you drive, take care, your car may be nervous...


well, I just ask for "ghost in the shell" theory, with no intentions of promove a way to use it, only to gather info.



#51 Helter Skelter   Members   -  Reputation: 332

Like
0Likes
Like

Posted 26 June 2005 - 05:13 PM

Twenty years ago if I posted a question on the internet there was a good chance that it would be answered. Of course twenty years ago if you had access to the internet and were capable of accessing usenet you KNEW how to ask a question. Does this mean that in the last twenty years people have become more arrogant or no longer feel the need to "share knowledge?" No. In my opinion the only thing that has changed over the last twenty years is the quality of questions. A lot of times they are either too vague or general for anyone to provide a response. While they could certain make a few assumptions and provide a couple of answers nobody is going to spend 20 minutes on a reply that in all probability doesn't answer the question.

The second issue is one of research (as pointed out by ApochPiQ). It's not unreasonable to expect that people research a topic they are interested in. Considering that at the top of this page are links to columns, articles, and other resources. In addition to the typical suggestion of "google is your friend" so is Wikipedia. If you search for AI on Wikipedia you get a list of possible topics such as cognitive science, connectionism, fuzzy systems, and statistical inference. You may not know what any of those topics are (or what they even mean) but you now have somewhere to start. After reading about some of those topics you will most certainly have a few questions. The difference is that instead of asking "where can I find out about AI" you end up asking "are there any good websites that cover fuzzy logic in games". It's more specific and you're likely to get several answers rather than absolute silence.

I also think there's a misconception about forums being for "sharing of knowledge". Sharing knowledge is not an absolute. It comes in many forms and not all are obvious. You might gain knowlege by participating in a forum thread debating the merits using fuzzy logic to augment an existing approach to game AI. It doesn't necessarily mean that any specific knowledge is going to be "shared" but when the discussion is over chances are you will have learned something about fuzzy logic.



One link with more than enough resources start researching AI: http://en.wikipedia.org/wiki/Artificial_intelligence

#52 Helter Skelter   Members   -  Reputation: 332

Like
0Likes
Like

Posted 26 June 2005 - 08:21 PM

Quote:
Original post by DarkThrone
The problem, Sagar_Indurkhya, is that mathematics is not the key to IA, only a tool to build it.


I disagree. While there is certainly a huge debate about how much math and which algorithms are best suited for AI, mathematics is a very key component in its development AND understanding. Emotions have little to do with AI. Many highly intelligent people (many in the genius range) exhibit some degree of autism and may lack certain emotional responses. In my opinion emotions are the exact opposite of intelligence. Given a particular situation a person may make decisions to do things they would otherwise avoid if they did not have particular emotions.

If you're going to continue with the assertion that psychology is needed to gain insight into AI explain why it's so important, what it contributes, and why AI cannot be obtained without it.



Quote:
Ah, only to remember, the Occam's Razor, that is kicked some posts ago.
WeirdoFu, I think YOU don't understand the principle of it. Occam's Razor principle is "On two possibilities, the more simple one are the right one."


That's one way to interpret it although I don't consider it very accurate. Occam's razor is typically expressed as "all things being equal, the simplest answer is usually the right one." In practice the number of solutions are not limited to two nor is the probable solution expressed as an absolute. Regardless of how you interpret Occam's razor you still have to get to the point of having multiple solutions to choose from.

I understand your desire to learn from discussion but many of the questions and comments you're posted are illogical. When discussing theory it's foolish to ask someone for a concrete solution in hopes of proving them wrong. After all if you have a concrete solution that proves a theory then it's not a theory anymore is it? (assuming of course you can prove the solution always applies)

[Edited by - Helter Skelter on June 27, 2005 7:21:00 AM]

#53 Barking_Mad   Banned   -  Reputation: 148

Like
0Likes
Like

Posted 26 June 2005 - 09:11 PM

Quote:

Life is mathemathical


Isnt that a misconseption

Mathamatics describes what we see as life, its a tool for explaining logic and patterns, not the source of the patterns themselves. Its somewhat paradoxical since mathematics is a way of describing what we see logically in our minds, and what we see logically in our minds is mathematical.

I guess this is in part related to Godels incompletness theorem:

Quote:

If an Axiomatic system can be proven consistent from within itself
thn it is inconsistent


Thats what i make of it anyway :)

#54 DarkThrone   Members   -  Reputation: 100

Like
0Likes
Like

Posted 30 June 2005 - 06:52 PM

yes, really feelings and intelligence are opposite things. But human beings ( and all universe, if you see ) acts in total balance. It's not a machism, please. Mens have a reason tendency, while womans have a feeling tendency. Both have some fails in some aspects, and this make the things works on a couple.

Real Intelligence is a total balance of feeling and reason. It can be see since long time ago, on ancient cultures. The secret is simple hiden on a complex thing.

And on true, I don't want to prove such thing. Each one have your own concepts over Intelligence, and nothing that I can do can change it.

But I have my own concept of AI. The only thing I don't have is a way to implement EXACTLY as I wish to be. I just point some ideas to it. If you can use it, great. If you show me that is wrong, great.

Foolish or not, there's it. My theory is not wrong, and I don't want a NOBEL to it. OK?

Some years of research are made before I get here.

#55 Helter Skelter   Members   -  Reputation: 332

Like
0Likes
Like

Posted 30 June 2005 - 08:18 PM

Quote:
Original post by DarkThrone
Mens have a reason tendency, while womans have a feeling tendency. Both have some fails in some aspects, and this make the things works on a couple.


I'm not even going to touch that one!

Quote:
Real Intelligence is a total balance of feeling and reason. It can be see since long time ago, on ancient cultures. The secret is simple hiden on a complex thing.


In some respects there are a handful of people, including those in acedamia, that agree with this. However this is typically LIMITED to decision making (and related topics) rather than as a direct, indirect, or causal link to intelligence. In that respect emotion (feeling) can have an effect on which decisions we make. It does not however dictact that the absence of emotion is an abscence of intelligence. To believe that both must exist equally in order for one to be present is, to me, completely absurd.

Quote:
And on true, I don't want to prove such thing. Each one have your own concepts over Intelligence, and nothing that I can do can change it.


To me it sounds like you're saying "I have this theory but I don't want to prove it" so I really have to ask....why wouldn't you want to prove a theory that is directly related and can have significant impact on the work you are doing? If you're investing so much time and effort to create an artificially intelligent being wouldn't it be more advantageous to know there's a possibility of success or a potential for failure?

Regardless of whether you succeed or fail it's still one step closer to a solution. Reminds me of something someone once said to me:

"People who strive never to fail will never experience true success."

More of a paraphrase really but you get the idea.



Quote:
But I have my own concept of AI. The only thing I don't have is a way to implement EXACTLY as I wish to be. I just point some ideas to it. If you can use it, great. If you show me that is wrong, great.


For me or anyone else to say that your concept of AI is wrong would be pointless. You have to come to that conclusion yourself. But if you're not willing to expose yourself to discussion or seriously consider the criticism of your theory/concept/etc. you've only succeeded in wasting everyones time.

Quote:
My theory is not wrong, and I don't want a NOBEL to it. OK?


Since you seem to have no interest in proving your theory you'll never know if it's correct or not. Besides your idea is already trumped by existing theory making that discussion moot. Have no idea why you would even mention the Nobel prize.

#56 DarkThrone   Members   -  Reputation: 100

Like
0Likes
Like

Posted 01 July 2005 - 07:30 PM

ok, guys we have a winner

Well, first of all, because this guy, seems like original topic reborn as a phoenix on ashes.

Good, really good. Some people never learn - a great phrase on begin of Front Mission 3. - and like to make critics on the work of another people instead of show what he knows. Ok, it's a free world, I guess, and critics are fundamental part of the hard, kinda cruel way to knowledge.

I really need to prove my theory? Which theory, is the first thing I need to ask.

If is the theory that some people critics instead of help, it can easily proved with a simple mirror. Look at it and you can see exactly the kind of people that I say.

If is the pieces of small things that I said belong this topic, I must say that is only ideas throw in wind, and looking the things useful post clear me on points that show errors in some.

Well, Helter Skelter, as I said on another topics ( older to your presence here ), is that I have all that I said in several notes, far of your eyes.
I research this 2-3 years. I give some respect to truthfully member that points some real things, as Timkim ( sorry by kidding ) that make me some real important points and Daerax that give me some place to research about code.
WeirdoFu is a bit cetic, but isn't a hard critic. He points me some things useful too.

Well, guy, "Prove your theory" is that you saying.

Figures in your brain a human being that have a intelligence but no feelings.
he can really survive? Why he lives? Why he works? Why he learn manners if it don't matter?

A real intelligence MUST be a combination of reason and feelings. Big minds on history make your names moved for HONOR, LOYALTY, FAITH, LOVE, and a dozen of another feelings. The feeling that moves. Reason only is a machine that calculate into all possibilities, the less lose/win weight to make it.

If you disagree, please, instead of just say it, show it. Seems like you are a kind of guy that walk on the streets and show with your finger: "Look, this guy is a loser" and "This woman is so pervert" only to mask your own weakness.

I say it because I don't remember of a real good thing that you show here. Sorry.

#57 mwtb   Members   -  Reputation: 115

Like
0Likes
Like

Posted 01 July 2005 - 09:22 PM


I've been skipping this thread as it seemed to start as a general moan about lack of resources, and I find this sort of complaint a bit laughable in the face of the availability of information these days compared to only ten or fifteen years ago. However, now it seems to have become a good example of a kind of conversation that I can rarely be bothered with, which might be a reason why the OP feels he is never answered.

I was once a games AI programmer, I have a background that includes academic A-Life research at post-grad level and as a hobby coder I have always gravitated towards AI-related projects. I don't consider myself an expert, but you would imagine that I'd be interested in the theory discussion in this thread,and yet I'm not.

Why?

Because as a request for help it almost immediately becomes a demand for solutions. It also contains an inherent dismissal of a certain kind of approach which immediately disengages anyone who has ever considered such sources valuable.

Because as theory for discussion, it isn't being tabled as something that can be explored, it's a dictat of reality according to the poster. Beyond that, it seems to simply boil down to the idea that "feelings matter" in human decision making. The weather matters, as does the amount of coffee you've had, or whether the pretty girl at the checkout smiled at you when you bought your newspaper. It's hard to seriously engage with such a theory without a strong modelling concept to back it up.

Because, ultimately, I care about AI from the perspective of achieving a more natural interface between people and software, and on this forum that software is games. How would this theory apply to games? How would the player perceive the improvements? What actual difference to agent behaviours are foreseen? How would the programmer/designer/player control the AI and bound its behaviours?

To be honest, it seems like you would have a much easier time tabling your idea and asking for resources in an academic psychology forum, as your thinking seems to be far more at their level.

#58 DarkThrone   Members   -  Reputation: 100

Like
0Likes
Like

Posted 02 July 2005 - 05:46 AM

Look for games as:

1. Thousand Arms (really funny)
2. Tokimeki memorial
3. General RPG
4. Shen Mue

The four on the list have a system of "talk to gather info" as fundamental part of your system. Shen Mue is the best on it. But even it have a list of predefined options to player choose, and play it several times make the game boring. (I am a compulsive player, really I know this feeling).

I think that will be more fun, if instead of you choose a option, you "type" a dialog. Even the programmers can have fun with it, because a N-possibilities that they can use instead of walk for the path that they build all the time.

And another possibility of a powerful IA on a game, eg. is a game of war, that can use decoys, and another resources with extremely accuracy instead of ramdomic or predefined times, taking army strategy unprevisible.

Sure that for someone it can be frustating because the game will be too hard. But for game lovers, I'm sure that is a challenge that can be you play until your hands lock, with the double of size, almost purple and painful.

And really is a bit hard to understand why some guys prefer boring me with pointless critics and decrease ratings instead of really point which are my faults.



#59 WeirdoFu   Members   -  Reputation: 205

Like
0Likes
Like

Posted 02 July 2005 - 01:06 PM

Quote:
Original post by DarkThrone
I think that will be more fun, if instead of you choose a option, you "type" a dialog. Even the programmers can have fun with it, because a N-possibilities that they can use instead of walk for the path that they build all the time.


I've had that idea for quite a while, but as a programmer, I would have to say that I would not have fun implementing it. There are a few inherent problems to the approach.

1. It will only work on game systems that have keyboards, like the PC. On anything else, the user would either have to buy a keyboard, if they don't already own one or you'll need to implement an on screen keyboard, which is usually slow as hell and cumbersome.

2. Issues with internationalization/localization is inherent. This brings up two issues of what language to use as input and then the actual parsing process. In the end, your system will have to break down into a word parser that does nothing but look for key words and construct the proper response from there. Forcing the user to use full sentences would cause issues with grammar and complications in sentence structure, which is why speech recognition has never gotten very good. If you actually want to incorporate grammar into the system, that would be a huge undertaking in itself, since people can't even understand each other some times due to grammar issues. Like for example, no offense, but I have trouble understanding half the stuff you write. So, if I were to implement a "typing" input system in english within a game, you might not get very far if grammar is necessary for the NPCs to understand you. It might lend to creating a level of realism, but would sure be a huge learning curve for some people.

Personally, I had wished they'd build something like that for RPGs years ago. So, you can ask NPCs about things and events you've seen so far or items and they'll tell you what they know or direct you somewhere where you can find answers. Something like, "Oh, you want to know about the Water Dragon? Well, then you have to go to the next town." or "I'm not sure, you might want to ask the mayor." Instead of the usual, "Its a nice day out. I like to take walks on nice days." The ability to query any npc about anything, even if they may not know the answer would really lend some realism to games.

#60 DarkThrone   Members   -  Reputation: 100

Like
0Likes
Like

Posted 04 July 2005 - 08:07 AM

Well, really as I said, I have a bad English.


1. I was thinking on it too. For it, can be used the system that is already used by Mobile Phones. And can be made a table that was part of inventory where pre-made phrases can be builded by players and saved there. So, he/she can shift some words on it and select to speak.


2. Yes, grammar problems are really the worst of challenge. I am testing a project of a lexical parser with a friend. One of solutions that I reach is that if the NPC don't understand the ask, it need "figure" it. Maybe is the more slow part to resolve.

And as I said that will be fun for programmers, I think that will be fun to they "play" the game, because I really love play games, and love play a game that I build to myself without know all the potential of the game before play it.




Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS