Creating a Powerful AI

Pages: 123456
you are a fool for disregarding my expertise in more subjects rather than programming (such as religion and philosophy, ha)
I didn't know you could become an expert at religion. Not a religion, mind you, but religion in general. And I'm sure it has a deep impact in AI research, too.

The most intelligent computer will be created by one with the most succinct understanding of the universe.
??? Are we talking about intelligence in terms of speed or in terms of computability (i.e. hypercomputation)?
Is snail "food"?
Is escargot "food"?
Is cow "food"?
Is beef "food"?
Is electricity "food"? (if you are talking computers? does the context of who is answering the question matter?)

Just training for "food" would take some work.

What are the rules you are going to use to train your AI?
Are you going to hand-code those rules or will the AI somehow generate those rules on its own?

Have you researched what others have researched so you don't waste time doing something that has already been done (which may or may not work given your requirements and expectations)?
Last edited on
closed account (S6k9GNh0)
I'm afraid that although I like your tenacity to study AI, I would never want to work with you in any region of the field if I were to ever move into it. Your generally an overconfident fool and laughably enough, you even prove it in a majority of your posts. You obviously overestimate yourself and you have an elitist mind set. I say good luck, and take my leave of this topic. I would truly suggest everyone else leave it alone as well and let him say what he wants.
Just looked at the website...

Viblicent wrote:
UGE, the Ultimate Game-development Evironment


No need to spell check when you're a 15 year old expert programmer/physicist/priest/philosopher?

Maybe he just likes being p-words.

Prat.
Oh thanks for the typo notice!

I'm afraid that although I like your tenacity to study AI, I would never want to work with you in any region of the field if I were to ever move into it. Your generally an overconfident fool and laughably enough, you even prove it in a majority of your posts. You obviously overestimate yourself and you have an elitist mind set. I say good luck, and take my leave of this topic. I would truly suggest everyone else leave it alone as well and let him say what he wants.


I know exactly how you feel, you people don't have to keep telling me that. However, I've noticed that over my small idiocies, I learn a lot from them. That's why I continue towards my goals with confidence.

And I'm not what anyone would call an expert, but to myself, I am an expert in my own way. Now this isn't just meaning that no matter how bad I suck, I think I'm an expert. No no, I believe that my "expertise" comes from the array of things I know, which all happen to correlate to Artificial Intelligence. I think that my Artificial Intelligence will be able to do what others cannot do with their AI machines.

Are you going to hand-code those rules or will the AI somehow generate those rules on its own?


First of all, I want to program a bot which will brute force and bake through the internet to understand words. This bot will do at-least 2 informational texts which contain a good variety of words, then it will go onto wiktionary.org to search for those words. But obviously a bot that can actually do that with the AI's comprehension would take a large amount of additional research. So it will obviously be years before I will have anything significant happening with the English language.

Second of all, I'm making the whole brain of this machine based upon how I believe the universe works... I have taken the greatest works from physicists, religious beliefs that have proved most effective, and philosophies that don't act destructive to these effective religious beliefs; combining them into one central rule-basis. The AI should be adaptable to any question or function with a response that is exceptional. This way, it will be able to brute force it's way to the grand keys of the universe (ah, "secret" knowledge).

Now I'm not exactly delusional, I would prefer you say that I'm just hoping for too much; but hey, I think it might just work!

You may have noticed that I have stopped talking about the actual AI as I have planned it. It's because I think I've built my theories to a deep-enough level, making my resulting decision be so I don't want to share it until I have US copyrights over them. I could give you a much more comprehensive, and sensible view of what I'm imagining about how my AI will work, but only if decided to.

Edit:
By the way, I'm not talking about religion as in Gods and Goddesses. The AI has to decide about that for itself. I'm talking about religion by ethics and morals. The AI will know whats right and what is wrong etc.
Last edited on
I don't want to share it until I have US copyrights over them
Well, according to US copyright law, a creative work is protected by it upon creation ipso facto. Of course, you can't copyright something that doesn't exist. Oh, and you can only copyright an implementation, not the principles behind the implementation. For that, you'd need patents.

I'm talking about religion by ethics and morals.
Yeah, that's called "ethics", it's a branch of philosophy.

Now I'm not exactly delusional
I prefer to think you're just ignorant, like the (many, many) people who think they've cracked infinite lossless compression.
So far you've made several outlandish claims which all basically boil down to "the AI will be omniscient" (some of the claims even imply that it will be able to decide on undecidable problems). Do you mind if I ask what makes you think think that you, who started this thread with "how would I store a bunch of bits where I could apply algorithms with ease?", are in a better condition to achieve any of this than any of the many AI researchers that have been working on the problem for the past fifty years? The system you're proposing, if it existed, would be decades if not centuries ahead of the most advanced existing ones. You're saying that you've managed to figure out the mechanism in a year?
I've been studying this for a long while (about a year now)
There's a principle called "Occam's razor" that states that, given two hypotheses that explain the same phenomenon, the one that makes fewer assumptions should be preferred. Wouldn't you making a mistake in your reasoning be more likely? Did you ask the opinion of someone with some background on the subject (basically, a small scale peer review)?
I prefer to think you're just ignorant
You're saying that you've managed to figure out the mechanism in a year?


Here, I have a better way of saying this:

I'm zealous. In fact, it's part of my brain chemistry, I am supposed to take medication everyday for it. It's called OCD, (ha-ha I know that explains everything).

Fortunately, my brain condition accelerates my desire to learn. But I guess, too often I happen to overlook things.

Now that I've found that I needed to come across this, I will relax and allow you to tell me that I'm an idiot. Thank you for trying to help my zealousness, your emotions and furry have done me good. By the way, I'm not really working on this as much as I was, I'm focusing most of my extra time on UGE.

Cheers,
- Ben
What religion are you to be ordained in?

You may have noticed that I have stopped talking about the actual AI as I have planned it. It's because I think I've built my theories to a deep-enough level, making my resulting decision be so I don't want to share it until I have US copyrights over them. I could give you a much more comprehensive, and sensible view of what I'm imagining about how my AI will work, but only if decided to.


There hasn't been a point in this thread where you have talked about the AI as a planned project. All we've heard is it'll be the best thing since sliced donkey scrotum. The wide, sweeping claims you're making are reducing your argument to the point of it being ridiculous. You obviously have no real knowledge of the problem domain, you're just day-dreaming a feature list you'll never be able to implement.

If you want to be taken seriously, start small. If you think that your AI should be able to understand English, develop a flawless language parser first and them build on that. The "classification" thing you were talking about before ("apple" is "fruit" is "food") isn't some miraculous idea you've introduced to computer science, it's the base premise for object oriented design! Also, what you described wasn't "intelligence", it was just basic pattern matching. It has nothing to do with physics, religion, philosophy or the number 42.

By the way, I'm not talking about religion as in Gods and Goddesses. The AI has to decide about that for itself. I'm talking about religion by ethics and morals. The AI will know whats right and what is wrong etc.


Ah, an "expert" in religion and philosophy, that doesn't know the difference between the two. How will your AI decide whether or not to believe in a deity? How will it realise that it needs to make that choice?
Hey, I never said "idiot". I said "ignorant". The good thing about the latter is that it can be fixed.

EDIT: I just happened to be reading a text for Epistemology and thought this passage was relevant:
A beginner is unlikely to make a significant scientific discovery, because the ideas he may come up with are likely repetitions of other ideas that ideas that have already been tested, or they might be in collision with facts or theories he or she doesn't know about.
-Carl Hempel, Philosophy of Natural Science, ch. 2.
Last edited on
Also, what you described wasn't "intelligence"...


Stop right there! I already know it isn't intelligence. I never said it was true intelligence. It's artificial intelligence, which will never be as cool and awesome as real intelligence which humans have.

Edit: Oh and please, don't get confused by when I said that it will actually "think", that statement doesn't mean it will be true intelligence. I didn't mean that it would have conscience or anything. Oh wow, you guy's are really dependent on specificity.

Ah, an "expert" in religion and philosophy, that doesn't know the difference between the two.


You wasted a post and you apparently did not read all of my own post, which you probably think is a ludicrous one. Quickly I must say that a religion contains believed philosophies, ethics and morals which are designated by their "God/Gods". I am stating this fact only because this has nothing to do with me being zealous, it deals with you, Pax. Religion contains everything of knowledge; it contains everything I said that I was expert at in a concise way. And if you read my posts completely, you would not have come up with yet another pointless attack.

EDIT: I just happened to be reading a text for Epistemology and thought this passage was relevant...


Yes, that is well understood by me. I am not done with research, but I also need to do thinking of my own. For example, I didn't even know that clever-bot existed, until a couple months after I began my research. I know how clever-bot works, I might make my AI get it's word knowledge a little bit like how clever-bot does it, but I'm also thinking of new ways to make my AI capable of processing so many things.
Last edited on
Your stated aim was to create a true "artificial intelligence" that could "think", unlike and superior to anything created by any other (including people who actually know what they're doing). Artificial in the context of the AI meaning not-natural, not inferior to the so-called intelligence of humanity.

You claim to be expert at everything, are you an omniscient entity? Forgive me, I wasn't aware we were in the presence of the deities on this forum.

So you claim that religion and philosophy are the same, and you are expert at both? Why then list them separately? Trying to look clever?
Viblicent, why dont you just start working on your AI...?

Instead of trolling the whole cpp.com community with your 'expertise'

Viblicent wrote:
HA! The map editor that I'm making will be just as good as the Unreal Editor

http://www.udk.com/features
http://img299.imageshack.us/img299/7546/screenshot000032.png
http://instantsfun.es/swf/badumtss.swf
Last edited on
@Pax
Your ignorant response (oh ignorant in too many ways to say) indicates that you have a zealous personality, like me. Lets get into some programming again and kill this stupid discussion! Please wait, I will be back soon with another question. During the mean time, leave the topic alone.

@Skillless
I can leave you alone.
Last edited on
@Skillless
I can leave you alone.


Thank you :*)
good luck with your map editor, you wont need it ofcourse, little expert of ours ;)
Last edited on
You're calling me ignorant?

What qualifications do you possess, little boy? What products have you shipped? What contribution have you made so far to the world?
closed account (D80DSL3A)
Viblicent sounds like an intelligent, curious minded person. A bit of immaturity may be expected of a teen.
I like programming because I can experiment around and immediately find out if my work is correct or not.
@Viblicent. I encourage you to keep trying! If your ideas aren't valid then your programs won't work. Simple as that! Your own computer supplies the proof.
Good luck with your very ambitious goals.
Ok, ok, STOP. EVERYONE. What do you gain from this flamewar? Honestly, you sound like children! (Albeit more intelligent)

Viblicent, just remeber that you're 15. You can't be an expert at ANYTHING, as you've only been capable of self-conscious thought for around 7 years. Also note that most of the people you're arguing with are 20-50, some with college degrees, all more knowledgeable about C++ than you are. Also, every single person has told you that this project is too difficult, even for people who have spent their entire life researching this stuff. Don't you think you should listen?

Now, I'm not telling you to give up the project. By all means, continue! All I'm saying is that you should take baby steps (as others have already mentioned). Look at the things you need to achieve those. I would even hesitate to call your ultimate goal part of your agenda, as true AI is decades away (at minimum).

Who knows? Maybe you'll BE one of those scientists!
Thanks for the encouragements. But I really do feel ashamed, I won't deny that most of you probably hate me. You guys don't understand, I've been through this process before. I've been zealous a lot, and I've pulled through it with a lot of what I consider success (years later, of course).

@Pax
What qualifications do I have? Well, I have programmed roughly 10,000 lines of code since the beginning of this year. I might be wrong, that was just an estimate based on how many functions and classes I've written multiplied by how many lines each of them have on average. No I haven't shipped any products yet.
Yes! Peace is settled. Now, before another flamewar begins, let's look at some code!

Or, better yet, make a new topic and let the flame die with this one.
Last edited on
Hey, everyone on here. I had a thought: Why don't we apply a little thought to this?
I won't deny that this project is well beyond the capabilities of Viblicent alone (sorry it is... for now), but is it beyond the capabilities of everyone on here combined? Think about it, most people who commented on a least have the semblance of brain, and some have much more. Instead of saying it's impossible, putting him down, insulting him, and being jerks... why don't we try and help him?

here's how things are: Viblicent wants to make a learning AI, he make be going about it all wrong, but that's his intent. everyone on here has prgramming ability, some more than others. i have given this much thought and believe i have a workable framework for his idea.

before you give the "Viblicent treatment" please listen to my suggestion for it is worth at least listening to.

AI is a simple equation. we ARE AI. In the beggining, we throw random output (crying etc.) and get input. we then classify the output based on the input. we may do it badly but that's what we do.
why don't we apply what we do to a computer program? this might not be as crazy as it sounds. the theory is that we make a program that writes it's own program. as it gets input, it adds new lines to the program. the lines would be logic derived from input. after enough time... the program would start being coherent just as babies start understanding things after a few years... lets say we have a microphone hooked up to it and that it knows high sounds cause "pain". the program is constantly throwing output and getting input through it's mic. you then walk in and yell "Hey!" the progrma would then write "ouput" (which it was throwing when you yelled "hey') = "pain" (pre-defined) = "bad"
all you do then is add more inputs, and get a server.
it's obviously not that simple but that's the frame work. if you guys want to help the poor guy instead of making him sound stupid and arrogant, maybe you should think about this and actually help...
we might be able to do this together...
Pages: 123456