Virtual Humans Forum
Virtual Humans Forum
Home | Profile | Register | Active Topics | Members | Search | FAQ
Username:
Password:
Save Password
Forgot your Password?

Virtual Humans
Partner website:
Chatbots.org
 All Forums
 Virtual Humans
 Business Applications
 Identify application areas
 New Topic  Topic Locked
 Printer Friendly
Previous Page | Next Page
Author Previous Topic Topic Next Topic
Page: of 5

laackejim
Committed Member



USA
3274 Posts

Posted - Jan 03 2006 :  22:56:38  Show Profile
Man oh man do you have a twisted mindJavascript:insertsmilie('')

On the other hand---hmmmmm--- nah, I got all the woman I can dream of.Javascript:insertsmilie('')

More seriously (I sure hope) I am pretty sure the concept will work for three emotions. I have not done the math yet for the 6 but am beginning to worry that there will be too much noise. The reference patterns I referred to would be answers to a string of specifically prepared questions (something along the line of Peter's book) which would explore language use etc. The Vhuman's pattern would be defined by designing a personality, checking if it meets our intent (does it feel right to us), then putting the questions to it and recording the responses.

Gad! it just occurred to me. IF we did that and sent the Vh out on its own, then after many sessions did the same patterning, posed the same questions to the Vh -- would there be changes? and if there are, would those changes be a measure of growth?Javascript:insertsmilie('')

Jim
quote:
Originally posted by hologenicman

I'm not sure why, but your explaination gives me the feeling of a dating service for human versus virtual human personallities...

It sounds like a mathematical formula for match making.

John L>


Uncle Jim (e=mc2)
Go to Top of Page

hologenicman
Moderator



USA
3324 Posts

Posted - Jan 04 2006 :  00:55:28  Show Profile  Visit hologenicman's Homepage
quote:
Man oh man do you have a twisted mindJavascript:insertsmilie('')


I was actually serious...

Not about the dating part, but about the testing.

http://www.emediawire.com/releases/2005/1/emw193081.htm

Online personality testing similar to Peter's may be helpful for evaluating the Vhuman's personallity and the human's match.

John L>

HologenicMan
John A. Latimer
http://www.UniversalHologenics.com

"If the Human brain were so simple that we could understand it,
we would be so simple that we couldn't..."
-Emerson M Pugh-

Current project:http://www.vrconsulting.it/vhf/topic.asp?TOPIC_ID=816&whichpage=1

DISCOVERY: The more I learn, the more I learn how little I know.
GOAL: There's strength in simplicity.
NOTE: Goal not always achieved.
Go to Top of Page

drwallace
Curious Member

USA
23 Posts

Posted - Jan 04 2006 :  15:12:01  Show Profile  Visit drwallace's Homepage  Send drwallace an AOL message  Send drwallace a Yahoo! Message
This task seems do-able, but rather expensive. We have one product, the CLAUDIO Personality Test bot, http://www.alicebot.org/claudio.html, which after a brief interview can produce an estimate of the client's personality type. It would be simple enough to vary the test to see if the client processes information primarily through an auditory, kinesthetic, visual, or semantic map. Similarly, you can use
linguistic clues to detect the mood or emotional state of the client.

We also have another product, the Superbot, http://www.alicebot.org/superbot.html, which is really helpful if you are creating a totally new bot personality from scratch. You could build in CLAUDIO-style AIML inquiries so that the bot could set an "etype" predicate, or a "umap" predicate, and then go on to make the other 9,900 bot replies depend on the values of such a predicate. They would use the AIML <condition> statement and branch to different replies for each client personality type (or map, or emotional state, or whatever you are trying to model).

But it already takes so much time, even with the Superbot, to create an original bot personality that doesn't have any complex user model, that just gives the same responses to everyone all the time, that it would quadruple your costs (at least) to build in such a model. Quadrupling doesn't even take into account the creative effort involved in tailoring specific replies to each emotional type.

My 0.99

Be Your Own Botmaster
Go to Top of Page

hologenicman
Moderator



USA
3324 Posts

Posted - Jan 04 2006 :  17:00:25  Show Profile  Visit hologenicman's Homepage
Being as cheap(Frugal ) as I am, I would consider some of the free online options:

http://web.tickle.com/personality/?sid=2003&supp=search_personality&test=personality

http://search.yahoo.com/search?p=free+personality+test&fr=FP-tab-web-t&toggle=1&cop=&ei=UTF-8

http://search.yahoo.com/search?ei=utf-8&fr=slv1-msgr&p=free+personality+test+dating

John L>

HologenicMan
John A. Latimer
http://www.UniversalHologenics.com

"If the Human brain were so simple that we could understand it,
we would be so simple that we couldn't..."
-Emerson M Pugh-

Current project:http://www.vrconsulting.it/vhf/topic.asp?TOPIC_ID=816&whichpage=1

DISCOVERY: The more I learn, the more I learn how little I know.
GOAL: There's strength in simplicity.
NOTE: Goal not always achieved.
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - Jan 04 2006 :  19:43:39  Show Profile
John -- I have never been frugal, if I can lay my hands on it and don't have to eat it, I spend it.

Richard -- I have always been a "lazy hard pusher" more comfortable working hard to build things that make doing something worthwhile, easier. I am totally, and I do mean totally, incapable of the focussed "drudgery" (sorry, there is no other emotional tag to how I respond to making a brain by hand) necessary to create what you have achieved and have made available in superbot. Unfortunately, I need to eat some of the $1000 it is worth. NO! I am not fishing for a deal. I am wrestling with a solution to the conversation flexibility response problem that has been growing out of the first few philosophical discussions of this forum. I may end up with exactly what you have produced (in which case I will in all honesty feel like I have achieved something worthwhile (because I accomplished it myself (your work would prove I was capable of it even if you did it first), or I will come up with something different, and feel the same, or I will fail and borrow the $1000 from my uncle and buy.

(the title "The Doctor" occurs to me often in relation to you, you are not a timelord are you? From galefri, perhaps?) Sorry, I am a fan of the BBC series, "Dr. Who?" and something about you brings The Doctor to mind. A good thing, he has single-handedly saved the Universe thousands of times, and with a sense of humor.

Oops back to subject. I am still chasing a "niggle" in my mind that was produced by one of John (the hologenic one)'s concepts. You have initiated another and I will work on it. If I am successful I will share the idea and approach. If not, on to another niggling problem and I will sign up for a loan and superbot.

By the way Richard I loved your 10,000 --> 9900/ 10,000 -->1 / 9900 --> my 0.99 pattern.

I would like to suggest the solution is more like

3,280,500 applicable responses (Vh output) = 1
my 1800 (phrases,words) =0.8

required 1800 to 3600 programmed phrases or words
(depending on degree of cross-category use)

One hidden number in here is Vh personalities, but it is a simple multiplier. I am thinking of 20 specific different ones, although the differences between them would seem minor until exercised.

Boy I hope I am right. I will let you know, success or fail. --insert emoticon for I actually mean this, right here-->

Jim


Uncle Jim (e=mc2)
Go to Top of Page

hologenicman
Moderator



USA
3324 Posts

Posted - Jan 06 2006 :  02:39:20  Show Profile  Visit hologenicman's Homepage
quote:
Richard -- I have always been a "lazy hard pusher" more comfortable working hard to build things that make doing something worthwhile, easier. I am totally, and I do mean totally, incapable of the focussed "drudgery" (sorry, there is no other emotional tag to how I respond to making a brain by hand) necessary to create what you have achieved and have made available in superbot. Unfortunately, I need to eat some of the $1000 it is worth. NO! I am not fishing for a deal. I am wrestling with a solution to the conversation flexibility response problem that has been growing out of the first few philosophical discussions of this forum. I may end up with exactly what you have produced (in which case I will in all honesty feel like I have achieved something worthwhile (because I accomplished it myself (your work would prove I was capable of it even if you did it first), or I will come up with something different, and feel the same, or I will fail and borrow the $1000 from my uncle and buy.


My only real concern with using AIML or SuperBot(other than financial) is that if any of us were to eventually go for the Loebner Prize the use of Dr. Wallace's efforts might disquallify our entries:

http://loebner.net/Prizef/2006_Contest/2006_Rules.txt

http://www.loebner.net/Prizef/loebner-prize.html

I guesse the question is, just how novel does the technology have to be to quallify, or are the entries judged on their current abillity to interact on a human equivalent level regardless of the underlying technology.

I would rather keep the technology pure if it will inhibit future credit, otherwise, utillizing such things as AIML and SuperBot would merely serve the purpose of convenience while developing other technologies such as EmotionalContext and EmotionalExpression for challenging the Loebner Prize.

John L>

HologenicMan
John A. Latimer
http://www.UniversalHologenics.com

"If the Human brain were so simple that we could understand it,
we would be so simple that we couldn't..."
-Emerson M Pugh-

Current project:http://www.vrconsulting.it/vhf/topic.asp?TOPIC_ID=816&whichpage=1

DISCOVERY: The more I learn, the more I learn how little I know.
GOAL: There's strength in simplicity.
NOTE: Goal not always achieved.
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - Jan 06 2006 :  06:32:19  Show Profile
John,
I believe the emotional concepts you have been defining in discussion, then coding etc. Hold the core of something really significant. I am less concerned about losing some originality by taking advantage of the work of others (no matter how wonderful it is/was) than I am about losing some of the power of your concepts by packaging any of it in a box designed for a different concept. I am about to launch on another glassy-eyed, arm waving, mystical oration about -- so I will quit. almost. compromise of options for a powerful idea during its inception is a mistake. The right path, I think, is to make it as gloriously expressive of its potential as possible in the prototype even if it only "runs on 3 cylinders and needs restarting" Then, let the world of implementation and use influence the implementation.

Jim
quote:
Originally posted by hologenicman

[quote]
I guesse the question is, just how novel does the technology have to be to quallify, or are the entries judged on their current abillity to interact on a human equivalent level regardless of the underlying technology.

I would rather keep the technology pure if it will inhibit future credit, otherwise, utillizing such things as AIML and SuperBot would merely serve the purpose of convenience while developing other technologies such as EmotionalContext and EmotionalExpression for challenging the Loebner Prize.

John L>


Uncle Jim (e=mc2)
Go to Top of Page

hologenicman
Moderator



USA
3324 Posts

Posted - Jan 06 2006 :  07:25:42  Show Profile  Visit hologenicman's Homepage
Jim, I think that you are absolutely correct in your advice not to compromise.

I was just looking for a temporary tool to help me with teh proof of concept in the mean time.

But Again, I think that you are right, and I should stick with my end game instead.

Btw, I am going to analyze the word jumble perception situation in the other post and start working on a "PerceptualEngine" next.

John L>

HologenicMan
John A. Latimer
http://www.UniversalHologenics.com

"If the Human brain were so simple that we could understand it,
we would be so simple that we couldn't..."
-Emerson M Pugh-

Current project:http://www.vrconsulting.it/vhf/topic.asp?TOPIC_ID=816&whichpage=1

DISCOVERY: The more I learn, the more I learn how little I know.
GOAL: There's strength in simplicity.
NOTE: Goal not always achieved.
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - Jan 07 2006 :  00:21:16  Show Profile
Good!

Your drive and mind are examples of what I mentioned a little while back. Something that we should find a way to support. Funding is beyond me at every level. However, it is not beyond the capabilities of some of those in this forum to facilitate the finding of grants etc to give take advantage of what is growing.

I have one or two people I can ask about ideas. I don't expect to get any valuable responses, nobody is going to believe what is coming until it has been around for a while.

Keep it going. If you give me headaches I can take the cost of asprin off my taxes as a business expense.

Jim
quote:
Originally posted by hologenicman

Jim, I think that you are absolutely correct in your advice not to compromise.

I was just looking for a temporary tool to help me with teh proof of concept in the mean time.

But Again, I think that you are right, and I should stick with my end game instead.

Btw, I am going to analyze the word jumble perception situation in the other post and start working on a "PerceptualEngine" next.

John L>


Uncle Jim (e=mc2)

Edited by - laackejim on Jan 07 2006 00:25:12
Go to Top of Page

drwallace
Curious Member

USA
23 Posts

Posted - Jan 10 2006 :  22:39:01  Show Profile  Visit drwallace's Homepage  Send drwallace an AOL message  Send drwallace a Yahoo! Message
You will not be disqualified from entering the Loebner Contest if you create a bot using the Superbot patterns, because your bot will have responses that differ from those of the original ALICE personality, or any other AIML bots entered into the contest.

From my point of view, the more AIML bots entering the Loebner contest, the better. If all the bots in the final round were based on AIML, then who would be the real winner? Dr. Who?


Be Your Own Botmaster
Go to Top of Page

hologenicman
Moderator



USA
3324 Posts

Posted - Jan 11 2006 :  00:21:31  Show Profile  Visit hologenicman's Homepage
quote:
You will not be disqualified from entering the Loebner Contest if you create a bot using the Superbot patterns, because your bot will have responses that differ from those of the original ALICE personality, or any other AIML bots entered into the contest.


Thanks, Rich, I wasn't sure about that.

BTW, I've read through the transcripts, and I am very impressed at the level of conversation being delivered at the contest.

John L>
IA|AI

HologenicMan
John A. Latimer
http://www.UniversalHologenics.com

"If the Human brain were so simple that we could understand it,
we would be so simple that we couldn't..."
-Emerson M Pugh-

Current project:http://www.vrconsulting.it/vhf/topic.asp?TOPIC_ID=816&whichpage=1

DISCOVERY: The more I learn, the more I learn how little I know.
GOAL: There's strength in simplicity.
NOTE: Goal not always achieved.
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - Jan 11 2006 :  03:29:17  Show Profile
Hi Doc. If Dr. Who saves the world again then he should be recognized. Being the last Time Lord he has his hands full keeping the future and the past functional. So I admire the creation of the future.

Talk to you later. Doc.
Jim
quote:
Originally posted by drwallace

You will not be disqualified from entering the Loebner Contest if you create a bot using the Superbot patterns, because your bot will have responses that differ from those of the original ALICE personality, or any other AIML bots entered into the contest.

From my point of view, the more AIML bots entering the Loebner contest, the better. If all the bots in the final round were based on AIML, then who would be the real winner? Dr. Who?




Uncle Jim (e=mc2)
Go to Top of Page

rickmurphynh
Curious Member



USA
1 Posts

Posted - Sep 20 2006 :  15:45:18  Show Profile  Send rickmurphynh an AOL message
Hello to all, this is my first response and as a new member I am interested in forwarding this technology in any direction.

I am not a programmer but love the potential applications.

I would like to ask the following basic question:
Where are the less lofty and more realistic commercial ideas?

In defense of such a blunt question, please let me say that I am trying to get AI in the market in any application that could be controlled and make a profit. True, what is done with these profits can be debated by capitalists and AI purists indefinitely. But getting viable business models running will allow SOME profits to help fund more complex "cutting edge" programs.



quote:
Originally posted by laackejim

Hi there FireChiefEd

In answer, to comment 1 (below) I have not decided to stop sharing ideas on the forum. I have gotten quieter because the main progress seems to me to be being made by hologenic man and GrantNZ, working together. MY comments as to whether they are going the right direction and/or there are better ways of doing it are likely to be non-productive right now (thats even with an egotistical assumption that I am right). Two dedicated, excited, intelligent, driven folks working together in a partnership driven by some link should not be messed with. They are probably right.

in answer to comment 2 below. Welcome aboard. A whispered note, just between you and me. Pay attention to Vittorio. He seems to have it all. Except for his taste in virtual companions. (ahem).

Ed. Only time will help you figure out whether I am a bunch of BS or real. But I am a fellow chaser of dreams. One who is used to catching most of them and this is one that is catchable, and worth catching.

Man this really fun and exciting, and worth the 12 hours a day spent.
Javascript:insertsmilie('')
Jim

quote:
Originally posted by FireChiefEd


Hi Jim,
If you reconsidered, and decided not to share your idea in this forum, that is your prerogative. However, please do not read a lack of responses as a rebuke from others.

I have just today joined this forum, and am interested in the things you write about.

How satisfying it is to do work that is interesting and fun and have people anxious to use the result of your efforts!

Ed
A fellow chaser of dreams.





Go to Top of Page

vrossi
Forum Admin



Italy
1456 Posts

Posted - Sep 20 2006 :  21:05:21  Show Profile  Visit vrossi's Homepage
Welcome to the Forum, Rick

Finding some commercial application is maybe the most difficult task for AI today.

Please look around in the other topics and you might find some interesting ideas, some of which could become commercially viable in the next future.


Vittorio
virtualhumansforum.com
Go to Top of Page

bhoecht
Curious Member



USA
8 Posts

Posted - Oct 08 2006 :  16:03:03  Show Profile  Visit bhoecht's Homepage  Click to see bhoecht's MSN Messenger address
Here is my company's first commercial application of a v-human.

You can access our demo system if you follow this link:

http://www.ai-dealer.com/Ai-Sales_Demo.html

Follow the Begin button at the bottom of the screen.

A bot w a purpose... our Haptek bot answers questions about what you don't know about buying a car from a car dealer. Just mouse over any text field in the workspace and watch for what it does.

Gen 2.0 will have an AIML chat.

As the prior member stated... business goes where there is money to be made. Should this v-human sells a car...

As the system orientation is still in development, here is my quick primer... you have to click on the text of the timeline to advance thru the system. The Helpful Thoughts section under the avatar has some longer informational communications (we find that if the avatar has > 2-3 sentences to say it is too much for mere spoken response to be effective).

Enjoy!
Go to Top of Page
Page: of 5 Previous Topic Topic Next Topic  
Previous Page | Next Page
 New Topic  Topic Locked
 Printer Friendly
Jump To:
Virtual Humans Forum © V.R.Consulting Go To Top Of Page
This page was generated in 0.23 seconds. Snitz Forums 2000