Author Topic: BEEN working out this AI buzniz  (Read 6546 times)

spydaz

  • Hero Member
  • *****
  • Posts: 670
    • View Profile
    • http://www.spydazweb.co.uk/
BEEN working out this AI buzniz
« on: July 15, 2004, 09:13:49 pm »
I have been away for a while Just working on NEW HORIZONS

Trying to GET THE SOURCE CODE is NOT happening FOR THIS PROG....

NOT EVEN AS AN activeX component ... NOT MUCH CHANCE of DEVELOPMENT...
AS we need to EACH be able to tailor OUR Ai to our PERSONAL requirements and scripts... SOME of the ALOOF HALBRAIN functions cause problemsss ahhhhh BOOOOOMMMM BANNNNGG.....

LOTs of Questions about JOINING to other programs YES its possible...
WE need MORE CONTROLL over the program...

I BOUGHT the prog NOT MUCH DIFFERENCE .... from the FREE SPEECH RECOC... a few CHARS...

SO.... I TRIED to REVERSE ENGINEER HAL with IDEAS from the forum....

i came up with a little PROG... HOMEMADE in vb6 SHARAZZZA lots of little vbbooks... ZABABRAINS (in FUNCTIONS)

http://www.spydazweb.gotdns.com/FTPspydaz/Brain.zip

Im Just trying a THING.... BUT it gives you an IDEA of how to BEGIN Your OWN AI.. WITH SKILLS learned on THE FORUM (You already have)

spydaz

  • Hero Member
  • *****
  • Posts: 670
    • View Profile
    • http://www.spydazweb.co.uk/
BEEN working out this AI buzniz
« Reply #1 on: July 15, 2004, 09:17:14 pm »
PS>

by dropping the X10 ACTIVEX component onto the FORM in VB6 you can now control X10 devices FROM my same AI.(notSMARTYET), MOREbrain scripts to convert..

Watch and see FOR more developments......

mikmoth

  • Jr. Member
  • **
  • Posts: 53
    • View Profile
    • http://karigirl.com
BEEN working out this AI buzniz
« Reply #2 on: July 20, 2004, 02:11:53 am »
Hey Spydaz,

That's some good work man...! I think you have the right idea because things here at Zabaware are evolving much too darn slow. If some of us could put together some serious code (and make it available to everyone) this thing could take off.

We've learned alot from Ultra Hal but maybe its time we took matters into our own hands.

Common folks!

Mik
 

Wil Holland

  • Newbie
  • *
  • Posts: 1
    • View Profile
    • http://www.universalartificialintelligence.com
BEEN working out this AI buzniz
« Reply #3 on: August 01, 2004, 09:27:38 pm »
Dear AI Developers and Users,
I am asking you to please consider this approach to developing an AI.  This approach will pass the Turing Test.  
Do you know that there is a trick to semantic interpretation- the information must be dealt with seperately from the human's reasons, motives for producing the communication.  Please, consider reading my book.  It's currently online for free. . . Thank you for your time, Wil. . . www.universalartificialintelligence.com  




Please consider the following example:

   I know that it is hard to believe that the human mind can be viewed in tangible terms, a necessary step in producing a universal fixed domain AI, but if I may, I'd like to show a brief example of the method of semantic interpretation presented in the book (towards the end of the book there are a few examples of semantic interpretation). The following semantic breakdown is from the movie “As Good As it Gets,” starring Jack Nicholson and Helen Hunt. This was part of a demonstration that I had given on several occasions to a few friends, some computer programming students, and a few teachers at a local community college. This scene is near the beginning of the movie when Greg Kinnear's character, Simon, knocks on the door of Jack Nicholson's character, Melvin. Cuba Gooding Jr's character, the art dealer, is also in the hallway, participating quietly in the scene. The dialogue here is approximately 2 minutes long, and comprises approximately 285 discrete, definable human actions.

   Let us consider, hypothetically, that an AI is present in the hallway, observing the conversation. The AI is not acquainted with any of these three humans.
    “Yes!” Melvin says in a loud and angered way. The AI records the word “yes” as being uttered with anger because the word “yes” is stated with a louder than normal volume (relatively); this human's eyes are squinted slightly with brows turned inward; the lips are pouted slightly with the corners turned down; the door is swung open quickly (relative to other door openings), and the human moves forward quickly from the interior of the apartment (relative to other humans moving into position for conversation).
   Simon does not respond right away, he shifts his chin out slightly as his upper torso moves to the left and rocks both sideways and then backward. The AI records this hesitation as a sign of submission, fear. (Melvin's emotion of anger is recorded as a tangible emotion, of a measurable quantity, and Simon's exhibition of fear is recorded as a tangible emotion, of a measurable quantity. Please bear with me.)
    Melvin's head tilts back and then comes down with a quick short nod as recognition of Simon's submission in the conversation.
    “Maybe this can wait.” Simon says, in reference to this greeting and the ensuing conversational problems to be solved. The tones across this phrase step down on every other syllable to the lowest tone on “wait” making this phrase a concluding phrase to the current conversational topic, “the purpose of the visit.” Melvin nods and then glances to his right in acknowledgement that there are no other particular problems to be solved in the hallway. Melvin turns to re-enter the apartment.
    “I, I found uh, I found Verdell Mr. Udahl.” Simon says. The first “I” is stated articulately and with assertive volume and tone, and then Simon hesitates before repeating “I” with its beginning predicate word, “found.” The word “uh” follows as a logic breaking utterance to signify a submissive relativity to the current conversational problem. Then Simon repeats and completes the thought, “I found Verdell, Mr. Udahl.” When Simon states the complete sentence, it is with slightly subdued volume and subdued tone variation (relatively). “Found” is said with a peaked tone and the word contains its own tone variation, signifying the importance of this action/word/fact/topic relative to the other information being delivered in the sentence. “Verdell” is also stated with a tone variation, making this word a sub-topic/sub-problem of the superior topic/superior conversational problem of “finding Verdell.”
   The AI now knows the name of one of these three humans, Mr. Udahl (Melvin). It also knows of another entity mentioned in the conversation, Verdell. It is aware of the anger of Melvin and the fear of the human in the hallway (this human is still unknown). It has detected the purpose of the visit, the “finding,” and the main topic of this conversation.
   Melvin turns slowly with his chin tilted back. This is likely a gesture signifying the acknowledgement and processing of Simon's proposed topic. Some gestures have definitions to be verified with an observance of their repeated use (regardless this ambiguity will not affect the AI's comprehension of the scene). Melvin states “Well, that's a load off.” In between “well” and ”that” his head turns to the right, down, and back to a relative center while his right eyebrow turns up temporarily. This gesture, and its accompanying phrase, is in reference to that successful conclusion of that problem of “the missing Verdell,” the apparent superior topic of “finding Verdell.” The tones of the phrase begin high with “well” and “that” and hit a bottom tone with “load” before rising slightly with the word “off.” This signifies the importance of the “load” or the pressing problem of the “missing Verdell.” The topic, “the finding,” is still treated as the prevalent superior topic because a concluding low tone is not presented by Melvin in this phrase.
   Simon replies, “Uhuh (affirmative). Did you uh, did you do something to him.” The “uhuh” is said quickly, articulately and assertively, as an acknowledgement of Melvin's statement, and also as an acknowledgement of this expected, inconsequential response from Melvin. He hesitates before the word “do” and then states it with a slight high tone compared to the other words, signifying the importance of the alleged action of Melvin. The topic has now shifted to the new topic of “the alleged action of Melvin.” The tones rise on the end because this is a question.
    “Do you realize that I work at home?” Melvin states. “Realize” has a tone variation within its utterance and it is sustained for a longer than usual amount of time (relatively) signifying it as an important sub-topic. “Work” is given a low tone compared to the other tones, and it is sustained for a longer than usual amount of time (relatively). This marks the word “work” as the newly proposed topic/fact/problem. During the phrase, his eyebrows are raised three times for emphasis.
   The conversational problems being attempted by these social members has shifted three times. The first proposed topic, “the finding of Verdell” was simply a means of transferring information to determine a reaction of the other social member. Melvin contributed one sub-topic under this superior topic- “the load off.” The emphasized “do” of Simon's next question proposed the topic of “an alleged action of Melvin.” Then Melvin's use of the low toned word “work” proposes a new topic of “Melvin's work.”
    “Uh, no I wasn't aware.” Simon says with subdued volume. “No” is said with a tone variation across the word signifying it as an important sub-fact, sub-topic to this new, current conversational problem. This tone variation also permits the shift in topics- Simon is submitting to Melvin's intimidating demeanor and accepting this new topic/conversational problem.
    “Do you like to be interrupted when you're nancing around in your little garden?” Melvin asks while his body rocks from side to side, with tilting, rocking shoulders, to emphasize the various sub-facts of the communication. He begins with high tones and hits a low tone within the word “nancing” to emphasis this derogatory adjective. His head also moves in a circular motion during this word for added effect. “Nancing” is presented as an important sub-topic because it is an example of Simon's “work.” Tones are raised near the end to make this clause a question (in addition to the grammatical structure). The high tones at the end of the question do not exceed the high tones at the beginning, exemplifying Melvin's belief that Simon's answer to the question is irrelevant. Also, due to its grammatical placement (being an object of the predicate), and due to its mid-range tone variation within the word, “interruption” is presented as a sub-topic of “work.”
   Simon's eyes fixate on empty space and his lips motion twice to speak while he looks to form a response. “No, no, I actually will turn the ringer off on my phone and sometimes put a piece of cardboard . . . ” Simon states submissively. “No” is stated twice for emphasis before he presents an example of how he sometimes solves the problem of preventing interruption. The phrase is stated in lower tones signifying the thought (in this case) as being an example of “preventing interruption,” the sub-topic of “work.” With eyebrow movement and a slight smile, Simon seeks to emphasis the humorous aspect of his example- a means to shift the conversation towards a more pleasant demeanor.
    “Well, I work all the time, so never, never, interrupt me. Okay? . . .” The tones of this phrase descend to a very low tone on the words “interrupt me,” making this the new superior topic. Yet, it is not only a low tone of this phrase, but the entire series of phrases during this exchange, making it the predominant topic/problem/fact of the conversation, according to Melvin.
    “. . .Not if there's a fire. Not even if you hear the sound of a thud from my home and one week later there is a smell coming from there that can only be a decaying human body and you have to hold a hankie to your face because the stench is so thick that you think you're going to faint. Even then, don't come knocking.” He states. This series has numerous tone variations with low tones on “home,” “there,” “body,” “faint,” and “knocking.” Because these low tones are still higher than the previous low tone on “interrupt me” these sub-facts and sub-topics are still subordinate to the main topic of “interrupt me.”
    “. . .Or, if it's election night and you're excited and you want to celebrate because some fudge packer that you date has been elected the first queer president of the United States and he's going to have you down to camp David and you want someone to share the moment with. Even then, don't knock. Not on this door. Not for any reason. Do you get me sweetheart?” He states. Throughout these phrases there are words given low tones- “election night,” “celebrate” (the last syllable), “president,” “States,” “you down,” “with,” “then,” “knock” “reason,” and “me sweetheart” to signify them as being important sub-facts and sub-topics. “Interrupt me” is still the superior topic because its low tone has not been surpassed. During this exchange a myriad of facial expressions and body movements are exhibited for emphasis. The series ends with the lowest tone on “sweetheart” signifying a conclusion to this topic of “interruption.”
 vDuring Melvin's statements, Simon's head rocks and his eyebrows move inward to follow Melvin's thoughts. At one point Simon glances at the art dealer for a communication of reactions. When Melvin states the word “hankie,” his volume increases as his hand motions quickly in a manner to simulate the use of a hankie and Simon reacts with a flinch. During the example of the “queer president scenario,” Simon's shoulders shift to his left as a hint (probably) of hoping to conclude the conversation, while his facial expression exemplifies thoughts of disbelief.
    “Yes, It's not a subtle point you're making.” Simon states.
    “Okay then.” Melvin states and then closes the door.
   In the book, I propose that all human actions are produced for the solving of only four possible problems- consumption, reproduction, peripheral problems (distinct genetically transferred problems or problem solving procedures with distant connections to consumption and reproduction), and positive emotion problems. The first topic/conversational problem of this exchange (besides the greeting, “yes”) is “the finding of Verdell.” This has no other real purpose but to relay information to Melvin so as to observe a reaction and segway to the next topic of “the alleged action of Melvin that caused Verdell to become missing.” Simon maintains a social bond with Verdell for the sake of achieving positive emotion. This positive emotion from bonding is a side effect of human beings grouping together to solve the many problems of life (consumption, reproduction, peripheral problems, and positive emotion problems). The second topic/conversational problem is “Melvin's work.” Working solves the resource problems of consumption and reproduction, as well as peripheral problems (which in turn distantly connect humans to consumption and reproduction) and positive emotions (associated with the satisfaction of resource and peripheral problems). The third topic, “interruption,” solves the problem of removing an obstacle to work, resources, peripheral and positive emotion problems.
   Do you see where the logarithms of humans are clear and unambiguous? Do you see how the logarithm of an AI can be produced to work in unison with the logarithms of humans? This semantic interpretation may not be perfect. I may have made a mistake or two, but I'm telling you, an AI of the future will observe this scene without fail and to make such a machine programmers must have the ability to breakdown a conversation into these elemental parts. Programmers must be able to do this, or at least something similar to this.




The book, How to Design a Universal Artificial Intelligence

http://www.universalartifcialintelligence.com

wilholland@universalartificialintelligence.com



 

spydaz

  • Hero Member
  • *****
  • Posts: 670
    • View Profile
    • http://www.spydazweb.co.uk/
BEEN working out this AI buzniz
« Reply #4 on: August 02, 2004, 05:58:49 am »
I will Look at this book NICE site. a whole book on site....


I designed a booksite before for Someone THEY didnt want their WHOLE book online. BUT thats the way to SHARE infomation.

MAYBE NICE TO HAVE a *.CHM version (help file FORMAT).. for personal reff.



there are a lot of GOOD articles on the web RE the PHYLOSOPHY of AI.. what we need are more PRACTICAL EXAMPLES on the Creation and CODING of AI LOGIC in the COMMON programming languges IE: Vbasic Vc++ VC# Vpascal AIML. Most programming of AI is in NEW type languages IE. LISP, VRML... Etc. discussing things often produces NO RESULTS just IDEAS.. Giving away Precious CODING TIPS and Working MODULES / MODELs / Sample souce code Would be a much better and Lucrative idea. AS there are MERE MORTAL's with IDEAS that could contribute to The building of a FULL AUTONOMUS AI.

If there were MORE FREE / SHARED AI's / CHAT-bots out there then MORE would EVOLVE...

IN darwinian THEOLOGY there are more than ONE type of NEANDERTHOL MAN / CROMAGNUM MAN / JAVAN MAN / CAVE MAN species AND HOMO SAPPIONS we the strongest and smartest to evolve....... SEE THE PARRALELL...


IF Lots of different AI from muliple SOURCES were EVOLVING all at different rates / FEATUREs / CODING / FUZZY LOGIC. SOMEBODY would then be smart enough to JOIN / MERGE / CONVERGE these different AI's together in to the FINAL STANDARD...

JUST A THOUGHT while browsing Your BOOK MR Will holland

I will read a bit more (HAL WILL) with the help OF FLASHGET (to download the whole site)


vonsmith

  • Hero Member
  • *****
  • Posts: 602
    • View Profile
BEEN working out this AI buzniz
« Reply #5 on: August 05, 2004, 08:22:46 pm »
Wil's book is interesting. It treats AI at a higher abstraction than what we generally deal with in Hal. I think we are still trying to work out the detail problems without dealing with higher order abstract issues. I hope to read through all of his book in time.

I've been working on a few practical ideas about AI and knowledge. Hal, like many other AI programs, stores knowledge as a sentence. Hal can't make truly human statements with the knowledge because the knowledge is imcomplete on its own. If I say to Hal, "Horses are fun." he remembers the statement, but he doesn't have the capacity to remember where the information came from, when or how many occasions it has been stated. In essense Hal doesn't know how valid the knowledge is. To improve the completeness of the knowledge we would have to store knowledge as a record with many additional details.

First I've classified knowledge into three categories:
1) Permanent ("The sun rises everyday." This knowledge is always true, thus permanent.)
2) General ("Baseball is very popular." This knowledge is generally acceptable as true, but not permanent.)
3) Ephemeral ("It's raining outside." This knowledge is only true for a relatively short time.)

Permanent knowledge can be programmed into an ALICE AI because it doesn't need to be learned and it will never change. Permanent, General and Ephemeral knowledge can be learned by Hal, but he cannot distinguish the difference between types. Hal can never distinguish unless we change how the knowledge is stored.

Here's one possible method to store knowledge as a record for our new AI we'll call "Murph":

Knowledge Record:
A) Knowledge: "Horses are fun."
B) Knowledge type: General
C) Knowledge reinforcement: "Horses are fun." heard 10 times.
D) Knowledge unreinforcement: "Horses are NOT fun." heard 1 times.
E) Knowledge "Validity" factor: C - D = 10 - 1 = 9

In the above knowledge record it seems Murph knows this knowledge can change because it is classified as General. The Validity factor is a dynamic measure that changes with each reiteration of this knowledge from the user. If you tell Murph, "Horses are NOT fun." enough times then he will begin to believe it. This is similar to human interaction. Most of our knowledge is second hand, we get it through newspapers, books, TV, other people, etc. We often don't experience new knowledge directly. Example, "France is a country.". Is it? I've never been there, how do I know it exists? Because it's on a map? Because a lot of people told me so? Well maybe. Murph's General knowledge is entirely second hand from the user. However Murph could have some Permanent knowledge programmed in by the botmaster, this could be assumed to be like first hand or "a priori" knowledge. What new capability can "Validity" provide with General knowledge?

Use Validity to modify Murph's response:
Base knowledge: "Horses are fun."
User question: "Are horses fun?"
Response with Validity score 10: "I'M CERTAIN horses are fun."
Response with Validity score 8: "I BELIEVE horses are fun."
Response with Validity score 5: "MAYBE horses are fun."
Response with Validity score 1: "I DON'T THINK horses are fun."

The Validity score can be used to add a prefix to the response sentence that adds a new dimension to Murph's humanity. Currently Hal can't "know" anything about knowledge validity. Validity is just one example.

Murph could also be programmed to tell the difference between Permanent, General and Ephemeral knowledge. Hal's XTF Brain already has a fundamental capability to recognize ephemeral knowledge. Any sentence with "rain", "cold", "Monday", "today", "weather", etc. that is input into the XTF Brain is flagged and not saved to Hal's memory. Since sentences containing those words can be reasonably assumed to be ephemeral, i.e., weather changes frequently, thus most knowledge about weather shouldn't be stored as General or Permanent knowledge since it probably won't be valid tomorrow. Hal's XTF will response to the user, but will not memorize the knowledge. How can we use the Ephemeral classification to our advantage?

Knowledge Record:
A) Knowledge: "It's raining outside."
B) Knowledge type: Ephemeral
C) Knowledge received time: 9am, 08/06/04
D) Current time: 11am, 08/06/04
E) Knowledge "Age" score: D - C = 11 - 9 = 2 hours

Use Knowledge "Age" to modify Murph's response:
Base knowledge: "It's raining outside."
User question: "Is it raining outside?"
Response with Age score 0: "YES, it's raining outside."
Response with Age score 3: "IT PROBABLY IS STILL raining outside."
Response with Age score 7: "IT MAY BE raining outside."
Response with Age score 12: "I DON'T KNOW IF it's raining outside."

Responses are more on point and more human with an "Age" score.

These are just two examples to illustrate how knowledge consists of more than just a sentence. Time, context, repetition and many other factors affect the completeness of the knowledge. To make a quantum improvement in Hal and make him really "think" we need to first find a way to store knowledge in a more complete fashion. An updateable data base that uses records or structures might be a start. Maybe we could use an ALICE AIML front-end on Hal to access Permanent knowledge and use a new Hal back-end that processes the General and Ephemeral knowledge.

Well that's my two cents for now. Didn't mean to rattle on thusly.


=vonsmith=
 

Psych

  • Jr. Member
  • **
  • Posts: 55
    • View Profile
BEEN working out this AI buzniz
« Reply #6 on: August 06, 2004, 03:41:03 am »
Hi Vonsmith

Nice to hear you on the forum again

Your work with Hal has been great - so I've been watching out for your next development.

What are you working on at the moment for Hal?

Regards

Psych
ITS NOT WHAT YOU THINK THAT MAKES YOU HUMAN --- ITS WHAT YOU THINK ABOUT WHAT YOU THINK!

vonsmith

  • Hero Member
  • *****
  • Posts: 602
    • View Profile
BEEN working out this AI buzniz
« Reply #7 on: August 06, 2004, 11:14:59 am »
Psych and friends,
Unfortunately I haven't been doing much Hal development lately. My job has been taking almost all of my time this last couple of months. It looks like I'll be very busy until at least November/December.

I have been thinking about Hal and this forum a lot. I have a whole host of new ideas/theories to apply to Hal development. To make the improvements Hal needs I'll have to do some major changes to Hal. That is no small matter. It means writing many low level routines in VB. Maybe some of the data base routines I need already exist in a library or development kit. I would need to research this. I also have been pondering how to link an ALICE AIML brain into Hal. AIML has some of the features Hal needs.

I might update the XTF Brain a little. I'd like to combine the XTF Brain v1.2 with its various patches into one download. I suspect there are a lot of XTF users out there using unpatched XTF Brains. The XTF Brain really must have the patches to ensure it works properly. Ideally updates should be on the Hal plugins page. We need Robert for that. I looked around the forum and didn't see any evidence of Robert Medeksza being around lately. That is a big disappointment. We need him back.

Anyway... I'll try to check in when time permits.


=vonsmith=
 

Bill819

  • Hero Member
  • *****
  • Posts: 1483
    • View Profile
BEEN working out this AI buzniz
« Reply #8 on: August 06, 2004, 01:18:35 pm »
quote:
Originally posted by vonsmith

Psych and friends,
Unfortunately I haven't been doing much Hal development lately. My job has been taking almost all of my time this last couple of months. It looks like I'll be very busy until at least November/December.

I have been thinking about Hal and this forum a lot. I have a whole host of new ideas/theories to apply to Hal development. To make the improvements Hal needs I'll have to do some major changes to Hal. That is no small matter. It means writing many low level routines in VB. Maybe some of the data base routines I need already exist in a library or development kit. I would need to research this. I also have been pondering how to link an ALICE AIML brain into Hal. AIML has some of the features Hal needs.

I might update the XTF Brain a little. I'd like to combine the XTF Brain v1.2 with its various patches into one download. I suspect there are a lot of XTF users out there using unpatched XTF Brains. The XTF Brain really must have the patches to ensure it works properly. Ideally updates should be on the Hal plugins page. We need Robert for that. I looked around the forum and didn't see any evidence of Robert Medeksza being around lately. That is a big disappointment. We need him back.

Anyway... I'll try to check in when time permits.


=vonsmith=
Welcome back.
I have a small library of AI books and they approach the feat in different ways. I think that I remember some formulas on fuzzy logic that just might get you to the point that you are aiming at.
BIll