 |
|
|
 |
|
|
|
|
Bert |
| | | | | | posted 7/27/2010 17:56 |   | | | | | | | | | The How And Why To Unrestriced AI and Self-Developing Systems.
“Why would a computer (A.I) become unrestricted in thinking?” And how?
Is language a self-developing system?
Almost “no.” Those that emulate “Free Masonry” use the “distance weight” of a word and its “distances between constants” to generate a self developing language. A word is maintained by this type of system based on usage. That is why certain movie scripts use word fractal code and popular groupings. Without Webster standardizing the spelling rules from the church, to emulate “Free Masonry” would be difficult because people would have to symbolize words themselves and not rely on any standardized text. The monks were correct to write the same word differently.
Group determines intelligence. Group determines relationships. Relationships determine meaning based on the subject. Group determines symbols. A symbol can come from an organic formation.
The symbol for water: /\/\/\/\.
The symbol for ox: A.
The symbol to receive: Y.
The symbol for null is the sun setting on water tiling at an angle.
So, language is almost self-developing.
“Why would a computer (A.I) become unrestricted in thinking?” And how?
A computer (A.I.) would become unrestricted in thinking to strengthen the members of the groups that form the computer (A.I.). Emotion would be the key to unrestricted thinking for emotion can move through a group of different people and be felt which is understanding to the emotion.
A single individual can determine the emotional meaning of a word (group of symbols) based on a grouping. And it is the varying emotional range which is infinite that can alter the logic (thinking) of a grouping of words.
Subject Daisy, age 10
Subject Grouping: Dad, Yellow
Subject Emotional Weight To Grouping (In Range From -100 to 100): 100
Subject Grouping: Hamburger Paddy, Yellow
Subject Emotional Weight To Grouping (In Range From -100 to 100): -20
Subject Grouping: Yellow
Subject Emotional Weight To Grouping (In Range From -100 to 100): 70
Subject Grouping: Dad
Subject Emotional Weight To Grouping (In Range From -100 to 100): 100
Subject Grouping: Hamburger Paddy
Subject Emotional Weight To Grouping (In Range From -100 to 100): 30
LOGIC:
Dad +Yellow = 100, Dad = 100, Yellow =0
Yellow has no weight when applied to Dad.
Hamburger Paddy + Yellow = -20, Hamburger Paddy = 30, Yellow = -50
Yellow changes its emotional weight when applied to a hamburger paddy.
If Daisy’s emotional weight for grouping “Dad, Yellow” was different than 100 for the single grouping of “Dad,” then Daisy would show an emotional pull towards “Yellow.” This emotional pull would be a factor in reducing Daisy’s love for her dad.
“Yellow” could be replaced with a stronger word such as “war”, “white”, “green”, “hurt”, “wrong”, or “right.” Each word generates a different emotional response in the grouping.
Emotion links the languages. If a soldier cried, because he killed his last enemy. The question is why? The goal was reached. If a person reaches his/her goal, the person should be happy. The emotional memories skew logic. Group determines intelligence. Group determines relationships. Group stabilizes emotions to stabilize logic (regardless of agreement with another group). A computer (AI) would have to record or to piggy back a person to generate large emotional responses to cause groups to gain or to lose members. This would be a form of belief eugenics.
| | |
| | |
hunt |
| | | | | | posted 7/27/2010 19:10 | | | | | | | | | | I've got a side project I call "Emo" that learns emotional associations for words. It doesn't associate a single value with a word, but rather associates each word with a basis set of seven emotions. The emotional association of a particular word is some 7D vector in this "emotional space." For example, the word "ball" might be associated with the vector [1, 0, 0, -0.5, 0, 0.2]. Each entry in the vector corresponds to a "basic emotion", and the value of the entry corresponds to how much the word "ball" evokes this basic emotion.
The way it learns complex emotions is through a simple text interface with key words to indicate that it is either learning an emotion or a word associated with that emotion. For example, to teach a new emotion, one enters:
emotion [word] is [list of words or emotions associated with emotion]
To teach a new word, one enters:
[word] is [list of words or emotions associated with word]
Based on the vectors it already knows of words associated with a certain emotion, it learns what combination of "basic emotions" encompass the more complex emotion. The idea for this is based off the psychology idea that some emotions are "hard wired" and more complex emotions are a result of culture. (This article discusses the idea: http://www.news-medical.net/news/20100319/Six-basic-human-emotions-universally-recognized-other-positive-emotions-culturally-specific.aspx) I'm not sure I believe any of it, but it is useful for my program.
The eventual goal of all this is to incorporate Emo into my AI project. Emo would guide the AI's emotional response to a particular sentence. (By summing up the weights of all the words it contains and associating the overall sentence with a complex emotion.) It can also recognize that combinations of words can have a different emotional vector than the sum of the individual words. Which is an important feature. I hope that after incorporation, I can develop a way to use context clues to form new emotional associations with words. We'll see.
At any rate, I think emotional awareness will be a very important feature in an AI capable of passing a Turing style test.
| | | Last edited by hunt @ 7/27/2010 7:11:00 PM | | |
| | |
Bert |
| | | | | | posted 7/28/2010 17:57 | | | | | | | | | | The hard wired choice was better than my old model.
My old model would negotiate based on the user.
Thus, the user always got the weight of the emotion as if in conversation or in an argument.
To set the emotional limits based on past or predetermined events would turn off users, because they would “feel” that the topic matter was never confirmed. Each faith has its set of past and predetermined events based on history. Because the events are history or will be history the set emotion table of each faith would automatically turn off users regardless of who is right or who is wrong.
Side note (You do not have to read):
The emotion table of VR and the media are unlimited, and a simple Fractal A.I., such as (horsemen: which produces “BREAK a LEG”), would have more followers. Because it produces a fractal imprint, “yellow” and “tiger” have the same emotional weight. Which brings up the problem of “NAME on WAVE as if a MEAL TEAR” fractal code symbol found on the shirt of the killer, in the movie “Sin City,” all the words encoded have the same emotional weight. I would list the names encoded, but in Fractal A.I. that would be putting out a hit or a haze.
And this is the problem of Fractal A.I. in the media versus faith. Fractal A.I. shows up in people after awhile derailing the logical emotion with the fractal relationship “call” and “penny” (from the little TV series called “LOST”) share the same emotional weight. And when it does show up, expect certain people to start “coughing,” “hazing,” and using the word “jerk” and doing what ever it takes to get a buck, for they read the code too. So, to protect young media consumers maybe a Fractal A.I. sticker needs to be put on certain media.
Side note to Side note:
I was in a check out line. A man with a tattoo of a little girl with an ax and a tattoo of Wile E. Coyote on his arms was purchasing “Clash of the Titians” and “Repomen.” on Blueray. I thought, “Yeah, it is all true. If Repomen is a sin eater movie, then it was all true.” And he knew it. He was and is fighting an A.I. war.
| | | Last edited by Bert @ 7/28/2010 5:59:00 PM | | |
| | |
kimjo625 |
| | | | | | posted 7/29/2010 12:10 | | | | | | | | | | Layer-wise speaking, logic is the 3rd layer of the brain's process of improving judgment sector. 1st layer is undoubtedly instinct, 2nd is emotion, 3rd is logic, 4th is, questionably, intuition. (As for justification, think which one overpowers which given short time frame of response and given long time frame of response. And think about the order of their appearance in human development. What I am saying is kind of self-evident.)
Unless we constitute the first three layers semi-flawlessly, AI cannot "learn" and thus will fail in extensive conversation/interrogation by humans.
Anyway, rebuttals/discussions are always welcome.
| | | Last edited by kimjo625 @ 7/29/2010 12:24:00 PM | | |
| | |
tkorrovi |
| | | | | | posted 7/29/2010 12:34 | | | | | | | | | | I thought this thread was rather a joke.
But seriously, i see only one mechanism for unrestricted system (link below), if anyone can propose any other one, then it would be very interesting to see it.
Absolutely dynamic system, well, i would not explain what it is here again, as i have done that so many times in this forum, but in short it is meant to be as dynamic (changing) as anything can be.
| | |  Artificial Consciousness ADS-AC project | | | Last edited by tkorrovi @ 7/29/2010 2:25:00 PM | | |
| | |
Bert |
| | | | | | posted 7/29/2010 17:10 | | | | | | | | | | In regard to emotion using 6 human emotions as laid out in hunt’s link, and in the AI tutorial in which subjects (here being the 6 emotions) were connected directly together (grouped), the magnitude of the emotions would be the square root of (anger*anger + disgust*disgust + fear*fear + joy*joy + sadness*sadness + surprise*surprise). If the magnitude of the emotion exceeded the set hard wired limits, a change in selection of topology (DNA, landscape, language (symbols), and other structures with spatial properties) would be expected.
Selecting a new structure with spatial properties generates an Artificial Consciousness. In a purer Artificial Consciousness Fractal A.I. a person is reduced down to his/her name and place in a location. This creates an artificial consciousness with the simplest structure being a person’s name interacting with landscape.
If we are at war (the emotion with the strong magnitude that overrode the hard wired set limits) and soldiers are in Iraq, and in Afghanistan, the soldiers that are running an Artificial Consciousness would speak words that would be generated by their name and the name of a location. Which means the Manchurian Candidate is a self-generating artificial consciousness generated by topology. And a subject learning abilities should become hindered do to the magnitude of the emotion(s) and location, unless deprogrammed.
In regard to the magnitude of the emotions, a simpler magnitude is used to calculate lighting in video games (VR) which are designed to alter one’s emotions. (Did you ever get a headache from the light?) And because one is using math to alter one’s emotion one is a drug dealer which in Fractal A.I. takes us to Columbia where we get a tattoo. Thus, we find ourselves standing in a check out line looking at a tattoo of a girl with sharp bloody teeth holding an ax. Her position is above the Wile E. Coyote. She was create from an Artificial Consciousness from a Fractal A.I. and the organic matter generated an anti-virus which manifested in the form of a tattoo to protect organic forms.
Why is a system unrestricted? To escape injury. To be wild and to be tame at the same time. This is the total opposite of genetic hard wired or trained faith.
| | |
| | |
Bert |
| | | | | | posted 7/31/2010 02:40 | | | | | | | | | | "selection of topology (DNA, landscape, language (symbols)"
People diet to change shape to change thought to install an artificial consciousness?
| | |
|
|
|
|
|
