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
 Growing the V-people Business
 New Topic  Topic Locked
 Printer Friendly
Next Page
Author Previous Topic Topic Next Topic
Page: of 2

thenar
Curious Member



USA
69 Posts

Posted - Dec 11 2005 :  22:52:14  Show Profile  Visit thenar's Homepage
I think the world is ready for v-people, they just don't know it yet. One of the problems I ran into especially with the large companies is that the people who run the companies and who make decisions are often ten years out of date on technology...no kidding...still. I was working with a cashregister company at one point and their people were all hot to develop my vpeople smartcards to help handle finaincial transactions...the bosses thought it was science fiction and killed the project. I still have a few of the smart cards with Sylvie's picture on them. It was very sad. We need to get the message out that the time is now. I've been trying...that's partly what my book was about. We need to get more and more into the news.
-P-

-P-

bhoecht
Curious Member



USA
8 Posts

Posted - Feb 09 2006 :  23:02:04  Show Profile  Visit bhoecht's Homepage  Click to see bhoecht's MSN Messenger address
To me, the use of v-people should not be viewed as a technology looking for a business. If the v-human referenced did not provide a meaningful lift or viable alternative to an existing business model (presumably real human), then it was not commercially viable (but probably just cool). From what I have seen, commercial grade v-people are just now/recently approaching commercially acceptable standards... if it doesn't look real, it is not going to be construed as such.

I do agree that the time for v-people is at hand, but for a different reason. 8 months ago I formed a new business. It's business purpose is to operate as an add-on to a car dealer's website and allow consumers to self-serve their own car deals - credit, trade, finance, lease, extended warranty, insurances, accessories, everything.

Now here is the rub... from the view of the business purpose the avatar/v-human serves three roles (none of which have to do with the technology or artifical intelligence):

1) consumers do not know enough to self-serve their own car deals, it is multi-facted and complex, there are legal / disclosure requirements to consider and a need for sales/persuasion to make the model effective.
2) the avatar makes the process an immersive, experiential one - fun, lively, engaging (perhaps even entertaining)
3) so used, the avatar is "remarkable" (as Seth Godin would define it in Purple Cow)... early adopters can't help but to pass on the word of what they have found

In this context, the avatar has a real business purpose - take full advantage of the rich medium represented by the Internet, consumer education, "persuasion architecture" and marketing by being remarkable.

To me, it is not that the world is not yet ready for v-people... it is that entrepreneurs have not yet closed the loop of what is required with what is possible... or like with my company, the derivative works are still in the pipeline.

Now if anyone has any really good (suitable for the role of online sales assistant role) suggestions on avatars, skins, clothes and textures, we would love the referrals. Our research shows that the available inventory out there is more reflective of avatars (to date) having been put to seedier uses than mainstream business applications.

Also, if anyone has seen other applied business applications of v-human / avatar technologies, we'd love to hear about them... good and bad... no point repeating past mistakes.

I do agree with the original posting that the time for v-humans is at hand... time will tell.

Me I just want to unlock the power of the Internet for selling cars, trucks, motorcycles, ATV's, RV's and powersports industries by dealers (world-wide of course). If I get to hang out with cool folks(like everyone who has posted in this forum), so much the better.

I am a CPA (US) and CA (Canada)... this is so much more fun than accouting!
Go to Top of Page

vrossi
Forum Admin



Italy
1455 Posts

Posted - Feb 10 2006 :  10:43:02  Show Profile  Visit vrossi's Homepage
quote:
Now if anyone has any really good (suitable for the role of online sales assistant role) suggestions on avatars, skins, clothes and textures, we would love the referrals.


Welcome to our Forum. I believe it's very useful to have here someone who is using VH in some business application.

I visited your site and I see you use the basic full body character by Haptek. There are many nice textures you can use to dress her and give her a more human look. You find an example here http://www.vrconsulting.it/vhf/topic.asp?TOPIC_ID=47 and you may access the digitalgirl forum http://www.digitalgirl.co.uk/ to see and download many more. Also on Zabaware Forum http://www.zabaware.com/forum/ you find many exmples: just search topics by aladyblond, for example.

I would like to see your demo but, after asking me to download the Haptek player, I can't see any other way to start the demo.

There is another site which contains a chatbot who helps the customer to choose a car: http://www.bobmoore.com/ and then click on Ask Megan.




Vittorio
virtualhumansforum.com
Go to Top of Page

aladyblond
Hooked Member



USA
620 Posts

Posted - Feb 10 2006 :  13:23:01  Show Profile
thank you vittorio.bhoecht, either i am not awake or i missed it but where is your site i would like to visit it.~~alady
Go to Top of Page

bhoecht
Curious Member



USA
8 Posts

Posted - Feb 10 2006 :  16:18:36  Show Profile  Visit bhoecht's Homepage  Click to see bhoecht's MSN Messenger address
You will find my company's website under my profile... and here too now...

www.ai-dealer.com

We have not yet enabled the avatar on the site. TV_girl by Haptek is too "sexual" for the business purpose we need. She needs proper clothes not cleavage.

Tying in the demo site is still a couple of weeks away (it is a first impression thing). We are still very much involved in getting the "experience" right and expect a commercial launch by summer. I'll post an update to this forum when we tie it in.

We are especially interested in a professional looking male avatar... sharp looking and well dressed.

For anyone not familiar with the term & practices, in designing a "Persuasion Architecture" (http://www.futurenowinc.com/persuasionarchitecture.pdf) one of the early exercises we did was identify which segments of consumers our application is likely to be most effective with. As our system provides an alternative to going in to a dealership showroom or sending in email leads, our evaluation was that the top two groups which will favor what we offer are:

1) women (influence > 80% of car deals, HATE the current alternatives the most),
2) 18-24 year olds (most comfortable w technology, have ZERO interest in the BS that happens in most dealerships, like self-serve where it involves the computer).
Go to Top of Page

bhoecht
Curious Member



USA
8 Posts

Posted - Feb 10 2006 :  17:01:00  Show Profile  Visit bhoecht's Homepage  Click to see bhoecht's MSN Messenger address
A quick follow on note... I see a lot of very interested v-people advocates on this site. If you have not run in to the below research studies before on the application of an avatar in the role of e-Commerce, they are worth a read.

www.ai-dealer.com/Artificial_Intelligence_ AI .html then follow the link thru to the ACM Portal. You have to subscribe which somehow takes them a couple of days to organize, but their free membership gets you all the papers shown.

Interesting intellectual constructs, except they missed the point of one of my favorite university lessons... How to win the Game of Chicken (which you can do if you control the system and medium).

A dealer in the US makes ~$1,000 on a Chevy Cobalt. Exactly how much effort should be put in to negotiation & talking up the relative merits of any given vehicle? In any given city, individual deal profitability varies by say $300 on a $14,000 vehicle. Car dealer margins on their vehicles are usually in the 4-6% range (lower than almost every other hard good industry where 8-10% is the norm, but due to it being the 2nd highest ticket item most consumers ever buy there is extreme price sensitivity).

Now for the Chicken part... the theory of the Game of Chicken is that there are two opponents racing towards each other. This creates 4 possible outcomes... both turn, one or other turns or both don't turn and both crash. For a long time this stood as a competitive assessment model in the world of business, until one day a rather "James Dean-ish" looking student put up his hand at the end of the lesson.

"Sir, you know how to PLAY the Game of Chicken, but you have no idea how to WIN at the Game of Chicken. If you play me, I wait until you are looking, then I step on the gas and rip the steering wheel off the console and throw it out the window. Now as my opponent, what are your choices?"

Where there is little actual long-term price differentiation possible + it is possible to provide an exceptional experience the consumer will not obtain anywhere else (backed up by a price guarantee, such as "find the same deal for less w/i 30 days + we'll refund the diff")... well as the controller of the medium, I just stepped on the gas and threw the steering wheel out the window. You either value the experience and buy or go seek out the alternative (into the showroom you go). You either pay for the value innovation provided as something you will reward by using or you do not.

Either way, if you are an extreme price shopper I would rather you spent 3-4 hours of your own time on my website rather than waste that kind of my salespeople's time in my showroom only to go shop somewhere else to save a few hundred $$$.

The early days of e-Commerce seems to have produced this thinking that only low price is effective. That has not proven to be true in any competitive business model over the long-term. Low price is NOT a long-term competitive differentiator and ususally only attracts consumers a business should force themselves to do without. Brand, reputation (and in my opinion) effective experience will be that which makes it in the Internet Age... just as it did in every single "Age" before it. eBay, Google, Apple iTunes+iPod... these are effective experience companies that built their brands and reputation on that experience. Again just my opinion, but truly buying a car, truck, motorcyle, snowmobile, boat, ATV or RV on the Internet will be no different.
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - Feb 10 2006 :  20:49:27  Show Profile
quote:
Originally posted by bhoecht

"Sir, you know how to PLAY the Game of Chicken, but you have no idea how to WIN at the Game of Chicken. If you play me, I wait until you are looking, then I step on the gas and rip the steering wheel off the console and throw it out the window. Now as my opponent, what are your choices?"



bhoecht,
Although I agree with you on every other issue (more on that later) I do have to point out an interesting fact about your James Dean-ish looking" student's assesment. Nothing has happened to his opponents choices the only choices affected were the student's. Since there were only two possible outcomes originally (crash or not), nothing changed there. Since the opponent originally had two choices, drive ahead or turn, and he still does, nothing changed there. What did happen is that two of four possible actions were eliminated without changing outcome options but doubling the probability of crashing.

If the student represents your business and crashing is the equivalent of making a sale, I understand your metaphor. On the other hand, as a third party watching this (who might be one of your competitors), I also note that you have turned control of the outcome over to the potential buyer. That is something I can take advantage of, perhaps.

That aside, I suspect you are on the right track and what is being built here (and I am sure elsewhere but I bet on the folks here) is necessary. The packaging and presentation is what you are talking about. And that could be everything.

Jim

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

Punisher
Curious Member



USA
44 Posts

Posted - Jun 02 2006 :  00:23:53  Show Profile
quote:
Originally posted by thenar

I think the world is ready for v-people, they just don't know it yet. One of the problems I ran into especially with the large companies is that the people who run the companies and who make decisions are often ten years out of date on technology...no kidding...still. I was working with a cashregister company at one point and their people were all hot to develop my vpeople smartcards to help handle finaincial transactions...the bosses thought it was science fiction and killed the project. I still have a few of the smart cards with Sylvie's picture on them. It was very sad. We need to get the message out that the time is now. I've been trying...that's partly what my book was about. We need to get more and more into the news.
-P-



Growth of the web (websites, Sharepoint sites, Blogger sites) seems to have opened an opportunity for V-People...

My initial post regarding what I want to accomplish in-house is a direct link to the business problem...Many corporations have information sprawl. I myself am on the phone constantly pointing people to information that already exists on one website or another. My concept is to use a V-person as a central url where somebody would go, ask questions, and get fed the information directly by the v-person. And than point to the in-depth site for further information.
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - Jun 02 2006 :  15:51:16  Show Profile
quote:
Originally posted by Punisher


My concept is to use a V-person as a central url where somebody would go, ask questions, and get fed the information directly by the v-person. And than point to the in-depth site for further information.


Punisher, if you have time please see my post in new members relative to your introduction.
Jim

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

Punisher
Curious Member



USA
44 Posts

Posted - Jun 02 2006 :  19:10:31  Show Profile
quote:

Punisher, if you have time please see my post in new members relative to your introduction.
Jim



Yes I did...Thanks, look forward to an interesting collaboration. My own pursuit in this is unfortunately more restrictive in what approaches I can take...I do believe however that when complete, it may be one possible solution in Corporate America which may be consumed.
Go to Top of Page

hologenicman
Moderator



USA
3324 Posts

Posted - Jun 03 2006 :  04:26:27  Show Profile  Visit hologenicman's Homepage
quote:
My own pursuit in this is unfortunately more restrictive in what approaches I can take...I do believe however that when complete, it may be one possible solution in Corporate America which may be consumed.

Actually, I am looking forward to the challenge of conforming a project to the needs of a comercial market. Of course, I enjoy the freedoms of a "hobbyist" and don't have to completely act according to corporate requirements. Still, following "Guidelines" in my endeavers will make the end result more likely to find success.

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 - Jun 04 2006 :  16:26:17  Show Profile
quote:
Originally posted by Punisher

quote:

Punisher, if you have time please see my post in new members relative to your introduction.
Jim



Yes I did...Thanks, look forward to an interesting collaboration. My own pursuit in this is unfortunately more restrictive in what approaches I can take...I do believe however that when complete, it may be one possible solution in Corporate America which may be consumed.


Okay Punisher, let's see what we can do.

My schedule has let up some and I have a little time today, that and the requirement to not use up acres of space with my proclivity to run-off-at-the-fingers (there I go again) I will take on the problem incrementally.

First: What.
A Vhuman interface to large amounts of data scattered over a series of websites (and other sources probably). This interface will serve both internal corporate clients of the techy and the "checky" (non-technical money folks". Will it also serve external "shoppers" or potential clients?

Note that the language limits of the your operational environment, or my capacity (or bullheaded commitment) limits, or other members personal choices DO NOT change the process that must be implemented. So concepts and design requirements discussed are usable regardless. In come cases a process is just impossible in a specific language -- only then is there a problem.

This is one of the places where your and my interests overlap, at least in the context of this being a limited AI problem with high probability input language and information needs. The information to be accessed and delivered is known, and although it may change frequently, it can be accessed continuously and catalogued in a machine efficient order for recall and assessment. The Vhuman's Rresponse will always be in the context of this limited environment rather than the totally open one most of the designers on the forum are passionately creating. HOWEVER, the limited environment is much more doable with a quality that will promote their passion in the long run, so I suspect that the others (Hologenic himself) would be inerested in collaboration.

For the internal, checky folks, this interface should make the user feel appropriately important (that usually means just a little more important in the stucture than they really are). The CEO's Vhuman interface for example could have her/his own secretary.
It (herein after referred to as Vshe) should also be very personal to the user and Know name, common language use, common jargon for data they are searching for (balance sheet, books on, sheet, bottom line, cookbook can all be terms for worksheet-27-block-cashflowwidget-Divisioncashflow.xls). Remember what they frequently ask for and what they asked for last time. When anniversaries occur etc.

The need for Visual Realism would vary from individual to individual but I suspect that it would be the rule, and in some cases be a critical and nitpicky (not enough realism) issue. Personal warmth, friendly intimacy and the faintest touch of sexual response should be a feature, the levels of each would depend on the corporate individual involved. And this is doable. For potential clients or customers the personalness should remain at the open, friendly but official, and superficial level.

Conclusion: This is all very very feasible, certainly here on the forum. In spite of the fact that some of us become testosterone theatened at times and attempt to overwhelm the cardio-vascular system of the oldest the talent is there to create the appropriate interface personna. The challenge is intriguing and may well get us doing.

Next installment--understanding user input and identifying appropriate response.

Okay Punisher, that is about as short as I can be

On the other hand, I just read the postings between you and John about english sql etc. You may be beyond my input and collaboration already. Glad to see it is moving on.



Uncle Jim (e=mc2)

Edited by - laackejim on Jun 04 2006 17:06:01
Go to Top of Page

hologenicman
Moderator



USA
3324 Posts

Posted - Jun 04 2006 :  17:21:21  Show Profile  Visit hologenicman's Homepage
quote:
For the internal, checky folks, this interface should make the user feel appropriately important (that usually means just a little more important in the stucture than they really are). The CEO's Vhuman interface for example could have her/his own secretary.
It (herein after referred to as Vshe) should also be very personal to the user and Know name, common language use, common jargon for data they are searching for (balance sheet, books on, sheet, bottom line, cookbook can all be terms for worksheet-27-block-cashflowwidget-Divisioncashflow.xls). Remember what they frequently ask for and what they asked for last time. When anniversaries occur etc.


quote:
Next installment--understanding user input and identifying appropriate response.


Punisher,

Jim has pointed out the most customer oriented aspect of Virtual humans. They need a personallity, history, and mutual association in order to draw in the interest and attention of your end-user. This may be a matter of falling back to AIML since it is the most effective NLP to this end to date.(did I say that out loud??)

The English Query functionallity of SQL Server is good for retrieving data, but it must be set up to react to converational input, output, and associations. You can probably use the English Query as long as you model it after the AIML structure and you get a conversational agent with the power of front-ending your disparate data sources. Otherwise, the interface is going to be fairly boring and therefore subjectively ineffective...

I would really recomend that for the purposes of modeling English Query after AIML, you should start up a conversation at the AIML forum, http://vrconsulting.it/aiml , where you can find Dr. Richard Wallace, the creator of AIML, himself.

Personally, I don't follow the AIML path because I don't believe it to have any future expandabillity or growth potential in the field of "AI", but it truly is the best proven current technology for NLP. And, following your business plan of utilizing the best available current technology, AIML (or the AIML model implemented with English Query on SQL Server) is probably exactly what you need.

The talent over at the AIML forum may just jump at the prospect of helping you implement AIML on the English Query medium or they may be too tied up in their current projects... Who knows? Every forum has a different atmosphere, and you'll have to check them out for yourself...

John L>
IA|AI

PS. Perhaps the best approach is to use true AIML for the front end for a conversational agent and personalit needs, and let the AIML interpretor feed HTML commands to the English Query utillity for data mining... AIML can simultaneously give text output to the user while issuing commands such as html and other such things.

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.

Edited by - hologenicman on Jun 04 2006 18:12:16
Go to Top of Page

Punisher
Curious Member



USA
44 Posts

Posted - Jun 04 2006 :  20:54:02  Show Profile
quote:
Originally posted by hologenicman

Jim has pointed out the most customer oriented aspect of Virtual humans. They need a personallity, history, and mutual association in order to draw in the interest and attention of your end-user. This may be a matter of falling back to AIML since it is the most effective NLP to this end to date.(did I say that out loud??)



Understood on the objectives, agree it should be strived for. When all is said and done...(even AIML) its all data. The power to have an adequate request/response technology is ultimately going to drive how the end user gets it. In the case of AIML, you are going to rely on a parser engine to get that response...

Where I was going on that vein when I first posted is the fact that Data is Data. Whether stored in an XML file or a relational Database.

When I first posted I was thinking I could replicate the AIML schema from with SQL server. Set up the XML via IIS connection and post AIML specific queries, if you want to use existing logic and presentation techologies. While at the same time taking advantage of the speed and scalability of a commercial grade product. (remember I am thinking in terms of an n-tier approach)

quote:

The English Query functionallity of SQL Server is good for retrieving data, but it must be set up to react to converational input, output, and associations.



Doesnt AIML as well?

Actually "English Query" is easier on the front end. Because it will automatically set up alot of the initial reactions based on your relational model of the Database structure...(no initial setup).

quote:

You can probably use the English Query as long as you model it after the AIML structure and you get a conversational agent with the power of front-ending your disparate data sources.



Actually I wanted to frontload the personality based on a seperate Database, which I was thinking of basing it on the AIML schema.

My initial problem has to do with Data in seperate locations. This is why I want to pursue Virtual Humans in the first place.

A front side Virtual Human with the personality, able to find the information you need/want.

quote:

Personally, I don't follow the AIML path because I don't believe it to have any future expandabillity or growth potential in the field of "AI", but it truly is the best proven current technology for NLP. And, following your business plan of utilizing the best available current technology, AIML (or the AIML model implemented with English Query on SQL Server) is probably exactly what you need.



I think AIML is a targtted "we can make it now" approach which I applaud. I think a creative extension to this initial work can extend its lifecycle and provide a "gap filler" between what is possible today, to where we would love to see it go.

quote:

PS. Perhaps the best approach is to use true AIML for the front end for a conversational agent and personalit needs, and let the AIML interpretor feed HTML commands to the English Query utillity for data mining... AIML can simultaneously give text output to the user while issuing commands such as html and other such things.



I think I can leverage both approaches for quick time to market. This forum in getting me to think, got me to think of using another technology I am familiar with to provide another potential gap filler. The ability to have a Virtual Human dynamically interact with you I believe is what we can do with Director/Shockwave.
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - Jun 04 2006 :  23:17:54  Show Profile
If I am going to participate in this exciting discussion I am going to have to get the topic right. I just stuck two comments that Might be useful in the english query thread. Keep going gentlemen, I am hanging on for all I am worth (no comments Grant -- Please!!)

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

hologenicman
Moderator



USA
3324 Posts

Posted - Jun 05 2006 :  09:04:05  Show Profile  Visit hologenicman's Homepage
quote:
When I first posted I was thinking I could replicate the AIML schema from with SQL server. Set up the XML via IIS connection and post AIML specific queries, if you want to use existing logic and presentation techologies. While at the same time taking advantage of the speed and scalability of a commercial grade product.
quote:
Actually I wanted to frontload the personality based on a seperate Database, which I was thinking of basing it on the AIML schema.

My initial problem has to do with Data in seperate locations. This is why I want to pursue Virtual Humans in the first place.

A front side Virtual Human with the personality, able to find the information you need/want.


1)Building an AIML interpretor with SQL Server would be a very kool project if you can get it done or get the AIML guys at the AIML forum to work on it with you.

2)The SQL AIML interpretor would need to provide both text ourput and simultaneous command output.

3)The command output would need access to be fed back into the SQL database for data mining which is your original need.

The SQL AIML interpretor could be centralized, but it would need to be personalizable for the various end-users. This could be handled with user profiles on the central database that allow for specialized interactions.

BTW, DR. Wallace offers an AIML resource with a multitude of thousands of rules pre-chosen for around $1,000.00 that will give the project a jumpstart. It would be a one time purchase so that the expense would be spread over multiple sales instances of the resulting application. It would be like sending the application to kindergarden, and he gives the impression that the files are easilly imported.

http://www.alicebot.org/superbot.html

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.

Edited by - hologenicman on Jun 05 2006 09:58:02
Go to Top of Page
Page: of 2 Previous Topic Topic Next Topic  
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 0.29 seconds. Snitz Forums 2000