Ai Dreams Forum
Member's Experiments & Projects => General Project Discussion => Topic started by: ivanv on September 01, 2011, 07:05:53 pm
-
I'm doing some work on behaviour of AI. Particularly on recognition of user intension. I thought that somehow mood about some intension has to be recorded. So a solution came to my mind: to align all doings according to "wanting to know and predict input as much as possible". And that could be the opposite side of "fearing of unknown input". Something like reinforcement learning.
But what if, for predicting input, machine uses some autoritative measures like "do that or else :knuppel2: ". I'm wandering is there some workaround?
-
some experiments with data formatting and visual representation (no, it's not off topic, it's about data!):
Table Data Classification And Visualization (http://www.youtube.com/watch?v=EMzFzNt56LY#)
limeta CASE (http://www.youtube.com/watch?v=t5lQoCeCwec#)
:i20:
-
oh, yes, just for completeness, although it was already published under "ethic" topic
Personality called: "Sky" (http://www.youtube.com/watch?v=BCq-qSBvrJY#)
:i20:
-
Hello everyone!
I'm in dilemma. Do you consider last two posts as spam material? should I delete them?
-
Hey, no not spam by any means, just busy at the moment so haven't commented... perfectly fine by me O0
-
I'm still wondering what you used to make the videos?
-
I'm still wondering what you used to make the videos?
first two videos are made in xara designer.
third one is made in swish max (for creating flash content)
-
The first one sort of reminds me of pivot tables.
The second, I don't really understand what you are trying to say, though it does remind me of how I represent data sometimes. Are you looking for, trying to or have developed a set based database system?
Here's an old image (but still mostly valid) of mind map editing in my neural network designer:
(https://aidreams.co.uk/forum/proxy.php?request=http%3A%2F%2Fjanbogaerts.name%2Fwp-content%2Fuploads%2F2010%2F08%2Fimage1.png&hash=a5d596de223c9d0ed76907d49ab63aa8f87f9076)
The blue one is a cluster of type 'object'. It contains a text neuron with the value 'me' and a regular neuron that's simply labelled 'me'. the object links to a neuron labeled 'home' and the link has a meaning of 'at', so you can read it as 'me at home' :o
-
Are you looking for, trying to or have developed a set based database system?
just for a while now, i'm experimenting with different formats for storing knowledge base, which i find very important for artificial intelligence. i test these formats on real-world commercial demands. once i have a complete system for knowledge base, i'll move to next step - implementing behaviour mechanism. i have no interest in commercial application development (at least for now), but i would be happy to participate with the idea.
the first video shows a sort of inheritance mechanism known from OOP
the second video shows how knowledge base could be defined through sets and N:1 relations known from relational databases. some background knowledge about databases in required to spot progress over current technology (SQL). but as for ai, the principle is this: each set has elements. each element can be connected to elements of other set through relations.
<below, in attachment is example>
well, first two videos show separately element in sets (first video) and definition of that elements (second video). my next station is to merge those two concepts into one single diagram. also, some dynamics for data are ought to be added. wish me luck :-\
-
as i was fastly enter the trusty member hall, i'm trying to prove myself a hard time. i was experimenting with that graph representation of knowledge. look at what i managed to construct with paradigm i called binary relational modeling (brm). brm has its own dedicated site under construction where that paradigm is being described in detail.
-
as i was fastly enter the trusty member hall, i'm trying to prove myself a hard time. i was experimenting with that graph representation of knowledge. look at what i managed to construct with paradigm i called binary relational modeling (brm). brm has its own dedicated site under construction where that paradigm is being described in detail.
Here is an article about data modelling and conceptual schemas that may help you, in case you haven't already come across it.
http://en.wikipedia.org/wiki/Conceptual_schema (http://en.wikipedia.org/wiki/Conceptual_schema)
-
Here is an article about data modelling and conceptual schemas that may help you, in case you haven't already come across it.
http://en.wikipedia.org/wiki/Conceptual_schema (http://en.wikipedia.org/wiki/Conceptual_schema)
boy, you scared me, particulary with this one: http://en.wikipedia.org/wiki/Concept_mapping. (http://en.wikipedia.org/wiki/Concept_mapping.) but i'm dealing with sets having elements and binary relations, not with concepts and n-ary relations. after folowing some links, i concluded that my brm isn't subset of any graphing methods i found in wiki about conceptual schemas, or at least it differs in groundings.
these are example relationships i found on wiki:
Each PERSON may be the vendor in one or more ORDERS.
- done with brm (essence)
Each ORDER must be from one and only one PERSON.
- done with brm (essence)
PERSON is a sub-type of PARTY.
- requires complicating brm (not an option despite of expressive power)
- will be implemented in superduper knowledge representation base that will hold brm also
Each Employee may have the supervisor within Employee.
- done with brm (by recursion)
anyway, thanks for watching and for information that i might not been wasting time. :D
-
i just googled "binary relational model" and it happend! i invented hot water!
link 1: http://c2.com/cgi/wiki?SemanticBinaryModel (http://c2.com/cgi/wiki?SemanticBinaryModel)
"Semantic Binary Model" is "A database model proposed by Dr Naphtali David Rishe.". it sounds just like brm and page was written in 2009.
link 2: http://dl.acm.org/citation.cfm?id=1575688 (http://dl.acm.org/citation.cfm?id=1575688)
"The Use of the Binary-Relational Model in Industry: A Practical Approach"". a book from 2009
link 3: http://dl.acm.org/citation.cfm?id=636814 (http://dl.acm.org/citation.cfm?id=636814)
"The Semantic Binary Relationship Model of information". this is my favourite. the book dates from 1984!
... and so on ...
the damn thing is full of that stuff! :uglystupid2: no, it's not funny! O0
anyway, if you think it deserves a look here is my own link:
https://sites.google.com/site/binaryrelationalmodeling/ (https://sites.google.com/site/binaryrelationalmodeling/)
-
here is a little presentation of how true ai would be implemented.
very breaf, 5 pages, no basic knowledge required:
https://sites.google.com/site/ivanvodisek/AI-goal.swf
-
here is a little presentation of how true ai would be implemented.
very breaf, 5 pages, no basic knowledge required:
https://sites.google.com/site/ivanvodisek/AI-goal.swf
You need a little bit more specificity. Its too general. Too broad.
Have you tested the theory? What are the results?
Edit:
My theory is similar to yours but my end goal is quite different. You are working on recognizing people's intentions. While I'm working on an AGI system that would also be able to not only recognize people's intents, but also recognize what they know and predict what actions they are going to make based on what they know.
-
You need a little bit more specificity. Its too general. Too broad.
Have you tested the theory? What are the results?
published presentation is intended to tease beginners in ai.
the theory is borrowed from classic theorem proving method that exists already for a while. one thing i find particularly hot is genetic programming: http://en.wikipedia.org/wiki/Genetic_programming (http://en.wikipedia.org/wiki/Genetic_programming)
My theory is similar to yours but my end goal is quite different. You are working on recognizing people's intentions. While I'm working on an AGI system that would also be able to not only recognize people's intents, but also recognize what they know and predict what actions they are going to make based on what they know.
have you found a way to recognize people's intensions?
-
Cool! Wasn't that the premise in the movie, The Minority Report?
I'm sure the program works on a much smaller scale and is more language
oriented than thought / action oriented but not far removed from the typing
assist programs that are usually found on smart phone when texting. As one
enters / types text, words appear from which to select based on the partial
word you are typing.
Perhaps it could also be geared toward sentence context....
Art, are you sure you're in a right topic?
-
have you found a way to recognize people's intensions?
Not yet. I'm still in the early stages of coding my theory.
I'm at stage 1 and in that is step 1. which is programming the basic layout of the knowledge base representation (which is different from the memory base because its dynamic and ever changing, while the memory base is static.)
(https://aidreams.co.uk/forum/proxy.php?request=http%3A%2F%2Fimg62.imageshack.us%2Fimg62%2F1936%2Fknowledge2.png&hash=3a12f26868054c4c0b77cac0a1aa101766014e85)
That is a basic sample I manually did. Basically the end goal is a cluster of words connected to each other by word links, forming a knowledge base based on language just like we humans have. The second step is the difficult one, that is programming an algorithm based on thought (thinking) which takes dialogues in example from the movie "Inception", finds patterns in them and then map it to its KB.
That will be the core of the AI, "A dynamic ever changing thought processing system"
There will be no grammar, or sentence parser. Nothing will be manually entered into the KB either
The end goal of stage 1 is is to allows the ai not just to map kb, but to think with its KB in real time aswell. In essence forming words, phrases, sentences all by it self, in its own mind.
The next coming technology age is AGI. It will be a trillion dollar industry, who ever figures it out will be the next Microsoft, Apple, etc. I believe it will be solved in 5 years by someone. But I believe I can do it in 24 months. ^-^
Every possible help, suggestions is welcomed.
I would like to design a site and make a topic, but I want to wait till after stage 1.
-
The next coming technology age is AGI. It will be a trillion dollar industry, who ever figures it out will be the next Microsoft, Apple, etc. I believe it will be solved in 5 years by someone. But I believe I can do it in 24 months. ^-^
Welcome to the site. Hope to see your project on a new topic and read more. Once you are at stage 2, let's see if those 24 months doesn't turn into ...a LOT more. ;) My theory sure looked a lot more simple than the actual prototyping which exposes the endless amounts of exceptions in the human language. But I do agree with you that whoever cracks the nut will be a very wealthy man/woman.
We'll see how far away we really are. Like someone said about the fusion power industry. It's the industry that has been 20 years away from the big breakthrough for more than 30 years.
-
hello genesys
i agree, no hard-coded examples. just general rules. O0
Actualy, there is a way to separate dynamic part of data from static one. We can treat dynamic data as blackbox that can be watched by static part of knowledge. Then, according to input from blackbox, static part of knowledge can adjust output to the same or to some other blackbox that is constantly watching for static output. So output to some blackbox is constantly changing according to input from the same or some other blackbox.
now, blackbox can be anything that represents dynamic data, but it must have interfaces for reading / writing its data. if those interfaces are known to static part and if we have a basic knowledge about some blackbox, we can use the static part to control blackbox' states in cannonical burst of input blackbox. So there go side efects, right into blackboxes.
boy, this was a hard one. it took me at least ten years for getting this idea. :'(
greetings from zombie room!
-
Cool! Wasn't that the premise in the movie, The Minority Report?
I'm sure the program works on a much smaller scale and is more language
oriented than thought / action oriented but not far removed from the typing
assist programs that are usually found on smart phone when texting. As one
enters / types text, words appear from which to select based on the partial
word you are typing.
Perhaps it could also be geared toward sentence context....
Art, are you sure you're in a right topic?
i get it... it's a joke, right? very funny, ha, ha...
-
Cool! Wasn't that the premise in the movie, The Minority Report?
I'm sure the program works on a much smaller scale and is more language
oriented than thought / action oriented but not far removed from the typing
assist programs that are usually found on smart phone when texting. As one
enters / types text, words appear from which to select based on the partial
word you are typing.
Perhaps it could also be geared toward sentence context....
Art, are you sure you're in a right topic?
i get it... it's a joke, right? very funny, ha, ha...
:2funny:
-
Art, you're killing me. it took me two days to get the joke, and now i can't stop laughing! mercy!!! :2funny:
-
That's it...I'm reporting you to myself! :knuppel2: ;)
<try not to take yourself too seriously around here>
We won't either! ::)
-
The next coming technology age is AGI. It will be a trillion dollar industry, who ever figures it out will be the next Microsoft, Apple, etc.
a trilion? i would like to make money obsolete. people could just push the button and get food, bikes, cars, goodies and everything!
what? no way, people, find your own girls, i'll bann that button :)
by the way, these days i have made some progress with knowledge base representation. still working...
-
That's it...I'm reporting you to myself! :knuppel2: ;)
<try not to take yourself too seriously around here>
We won't either! ::)
i'm reporting myself for stupidity ^-^
yes, it's something like those cellphone's prediction of what word we want. like we know first letter (current state of some system), but we know last letter also (wanted state of the system). And we know what words exist, meaning how can we can get last letter starting from first letter (what sequences of actions we can apply). but ai could have addition of combining words (i'll report cellphones too).
pretty simple, right?
-
sets & elements linked :)
https://sites.google.com/site/binaryrelationalmodeling/brm.swf
next stop: dynamics
-
The next coming technology age is AGI. It will be a trillion dollar industry, who ever figures it out will be the next Microsoft, Apple, etc.
a trilion? i would like to make money obsolete. people could just push the button and get food, bikes, cars, goodies and everything!
what? no way, people, find your own girls, i'll bann that button :)
Seeing that everything in the next technologic age will be run by smart AI.
Cars, Flying cars, military vehicles, home systems, phone systems, camera systems, businesses, companies, etc...
Its quite plausibly the first person who achieves the goal will end up creating a company that could be worth trillions.
-
i think i solved dynamics too, without throwing away functional dependencies:
https://sites.google.com/site/binaryrelationalmodeling/brm-elems
as with attach:
input set could be console or microphone.
output set could be console or speaker.
think it's not hard to make kb by observing input and inducting
consequences as happenings. it's not hard to reach wanted
state of the universe once we built the kb.
remains: to detect states we want to reach.
does anyone have any idea?
-
How is the knowledge going to be collected and mapped? (ex; learning.)
-
How is the knowledge going to be collected and mapped? (ex; learning.)
i thought that input set changed by input devices could be monitored. rules could be inducted by statistical analyzes. not certain about concrete implementation, but mother-rule could be: happenings that happen together are consequences each to other.
in analyzing console input by testing different combinations, letters that are seen together could be considered as words. words that are often seen together could form some sets (a grammatic predicate can be a flag for grouping other words). the key could be in repeating patterns (every sentence has the same elements, some of elements more or less described). this is behaviour that should be reached by some general rules of induction from any input set (sense of ai system).
universe could be considered as blackbox that inputs states to ai system (sensors) and recieve ai system actions (ai output). now we need to examine what are rules respected by this blackbox. connections between output and input should be concluded. every input set should be analyzed in the same way: to see what sequences are repeated through time and what are connections between these sequences. we have to detect - is some sequence always following some other sequence. then comes grouping of sequences as elements of the same set. then comes analyzing sets as types of sequences to see regularities that are always true, no matter what element is observed. regularities should be guessed more or less randomly. i believe that by inductive learning from input sensors a knowledge base could be formed, describing things that happened in ai sensors. and when kb is formed, when we know what input from senors will bee read after some output, it's easy to calculate what sequence of output device states needs to be applied to achieve some goal.
i'm not certain about concrete rules of induction, but i think the base one is repeating of the same happening after the same action. i don't know, maybe a set of elements could be formed by grouping different happenings that follow the same action. yes, it is another problem, it needs some workout...
-
hi there friends :)
what do you thing of this compromise finder (already published under ethic topic): https://sites.google.com/site/ivanvodisek/
a thought crossed my mind. every thing that happens in the universe has some reason that made it it happen. either it's a wish from a living being, either it's some universe hard-coded rule like gravity amount depending on some parameters. so each action is taken to keep some rules holding.
i was thinking: how about to obey *every* rule recognized in universe (with restriction of making compromises mentioned above)? if it's wish, it should be fulfilled. if it's universe's rule, nothing bad would happen, it's gonna be obeyed anyway. so the problem would be in finding compromises between universe's rules and everyones wishes.
is this sounds as a friendly ai that wouldn't "kill a girl that works in market to get a can of ananas"? i'm scared somehow...
anyone?
-
How long do you think it would take to model the rules in the universe ?
-
i don't know. i guess the system would generalize rules from specific examples of input. so it depends of experienced input. can ever input be totally generalized, or there will always be new unpredictable kinds of input? i bet it's later one.
now a problem is laying on the way: setting up a proper input sensor and rules that generalize that input and store it to kb.
-
When you say the universe are we talking right down to the molecular level too ? I mean laws of Physics and so on, or do you mean an average persons view of the world ?
-
both, if possible
-
'If possible' seems the right thing to say ;D
I think it would be difficult given that they can't even model weather patterns that accurately, so as for the universe....well....
That's why I am more interested in AI in simulated environments because at least that will happen in my life time ;)
Noble cause though, every thing starts somewhere :)
-
rules of the universe. maybe it's a fractal function that goes down and up to infinity, and maybe that function is alterable by living beings. let's see, what should function look like to get a spaceship full of spacegirls land to my houseyard >:D
-
i was thinking: how about to obey *every* rule recognized in universe (with restriction of making compromises mentioned above)?
i've got a good news and a bad news.
the good news is: if we obey all rules (or their compromises) we can't go wrong. at the worst case we'll do nothing (if there's no compromise). so this is safe. :)
but the bad news frightens me: if we put a logic operator "NOT" to intersection of all sane rules, we get an criminal who lives to destroy all good in the world. so one error in programming and... we're doomed :(
but, in its old fashion, the Universe is reaching to us: with little information, a machine can't be seriously harmful, so we have little time at the beginning of its "life" to decide whether we'll going to let it be. risky, as always. is it worth of risk, as at opposite side, we have a world with no suffer, hunger and criminalities?
i'm for it. :-[
-
just another site about true artificial intelligence.
short intro to type theory and automatic rules and algorithms construction.
about 4-5 pages.
i thought someone might be interested:
https://sites.google.com/site/typetheoryrefined/
:)
-
Thanks for posting that link ivanv, it was among the most fascinating 5 pages that I've read in a while. Type theory (and typed feature structures) are slowly emerging as a very promising area of research. For example the English Resource Grammar is the most complete and comprehensive grammar of the english language that is freely available, and it is all based on types. There are also grammars available for other languages based on the same methodology but I don't think any of them are free.
http://www.delph-in.net/erg/ (http://www.delph-in.net/erg/)
-
Thank you, Infurl :)