wu :: forums
        (http://www.ocf.berkeley.edu/~wwu/cgi-bin/yabb/YaBB.cgi)
      

        riddles >> general problem-solving / chatting / whatever >> the web as a killer of novel applications
        
(Message started by: amichail on Aug 6th, 2006, 3:02pm)
Title: the web as a killer of novel applications
Post by amichail on Aug 6th, 2006, 3:02pm
From my experience thus far trying to get a startup going, I am now convinced that the web is a killer of novel applications.

The reason is that the average user is willing to look at a web service only briefly. If it doesn't make sense in less than a minute, he/she will leave.

But do novel applications really make sense that quickly with minimal effort invested in understanding them?

I suspect that web apps (and to some extent, also GUI apps) have led to impatient and unsophisticated computer users.

There are several potential solutions to this problem:

  • design GUI and web apps in a way that encourages exploration of more complicated features that require some time to understand
  • require everyone to learn programming; since computation is the most powerful form of self-expression ever created, one can make a case that it would be unethical for a society to make programming optional
  • have some sort of way of certifying novel/complex sites that are worth your time
  • have something like required art appreciation courses but for novel computer applications
Title: Re: the web as a killer of novel applications
Post by towr on Aug 6th, 2006, 3:44pm
From my experience, I have to disagree. The web has tons and tons of novel applications that millions of users discover, use and promote.
True, there are tons more applications that are left out, but almost certainly some similar app somewhere is getting attention. There is simply a lot of competition.
But on the up side also more of a chance that you get noticed. Regular software devellopers used to fail in silence en masse, but on the net people can stumble across them much more easily than in the 'real world'.

Certainly, a snazzy GUI will help, simplicity in design will help, etc. But those are qualifications for normal software just as well. An application ought to be usability based, not feature based (as we programmers all to often tend to do). Many people still can't program their VCR; yet potentially they can do more than ever. Lots of features which people can't use or have no use for in the first place.

I suspect, the problem may not lie with the user. If they wanted to result those extended features would provide, they'd take a small effort. But if documentation is poor, and the use is needlessly complicated, it just ends somewhere. The result has to be worth the effort.
And I can hardly believe you're actually suggesting changing the users. Instead of making bigger doors, chop people's heads off so they'll fit through. Certainly there'd be a lot of benefit if people could all understand a bit of programming. It'd also be great if programs were made as userfriendly as possible, though. Some cognitive ergonomics deserves consideration.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 6th, 2006, 4:09pm

on 08/06/06 at 15:44:06, towr wrote:
And I can hardly believe you're actually suggesting changing the users. Instead of making bigger doors, chop people's heads off so they'll fit through. Certainly there'd be a lot of benefit if people could all understand a bit of programming. It'd also be great if programs were made as userfriendly as possible, though. Some cognitive ergonomics deserves consideration.

Changing users is not so unreasonable.

An analogy would be a society where the average person is illiterate.  Changing people in that  case would be desirable. So why not here? Why not raise the standards of computer literacy to new heights so that we can build more novel applications?
Title: Re: the web as a killer of novel applications
Post by Icarus on Aug 6th, 2006, 4:15pm
I understand your frustration, but you have the wrong culprit. The same problems plagued every inovation. The problem is, most adults are not willing to invest the effort to learn to use something new until they are convinced that their situation will be improved by it. Children are far more likely to explore new tools simply for their own sake, but we who have passed that stage find that there are far too many things out there clamoring for our time and attention, so we want to know it works before we get too far involved.

Designing GUI and web apps to encourage exploration is good. But if you try to force it in any way, you will drive people away instead. Which exactly why your enforced programming and app appreciation ideas are definitely the wrong way to go. As for certifying sites, we already have something that I promise already works better than any sort of official certification does: endorsements. People read reviews and other endorsements from sources that they know are reliable, and are interested in the same things they themselves are. If you tried to replace that with official certifications, it wouldn't work. The official certs are from sources that have different interests than the user does, so he is going to stick with sources he knows.


What you really need comes down to that most hated of words (in my mind, anyway) "Marketing". If you don't sell it, people are not going to buy it. You have to make your product known to people who can give you good endorsements. You have to bring your app to people's attention, and convince them that it is worth exploring. And you have to convince them on the basis of what they want, not what you think is really cool and inovative. It doesn't matter how fantastic a mousetrap you build, someone who doesn't have a mouse problem isn't going to buy it.

All of this was just as true in the stone age as it is in the information age. The web had nothing to do with it.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 6th, 2006, 4:38pm
This is what I am working on:

http://targetyournews.com/

It appears that most people have no idea what the novelty is or why it is interesting.

Also, very few people actually read the documentation. But perhaps there's enough in the interface to see what is going on?
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 6th, 2006, 4:42pm

on 08/06/06 at 16:15:04, Icarus wrote:
What you really need comes down to that most hated of words (in my mind, anyway) "Marketing". If you don't sell it, people are not going to buy it. You have to make your product known to people who can give you good endorsements. You have to bring your app to people's attention, and convince them that it is worth exploring. And you have to convince them on the basis of what they want, not what you think is really cool and inovative. It doesn't matter how fantastic a mousetrap you build, someone who doesn't have a mouse problem isn't going to buy it.

All of this was just as true in the stone age as it is in the information age. The web had nothing to do with it.


Marketing is difficult when self-promotion is discouraged on the web.  

My service actually addresses this exact issue: it creates an environment where self-promotion is acceptable and potentially effective.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 6th, 2006, 4:54pm

on 08/06/06 at 16:15:04, Icarus wrote:
I understand your frustration, but you have the wrong culprit. The same problems plagued every inovation. The problem is, most adults are not willing to invest the effort to learn to use something new until they are convinced that their situation will be improved by it. Children are far more likely to explore new tools simply for their own sake, but we who have passed that stage find that there are far too many things out there clamoring for our time and attention, so we want to know it works before we get too far involved.

But just how fundamental is computation? Should it not be a required subject in high school?

It seems to me that programming, as the most powerful form of self-expression ever created, is pretty fundamental!
Title: Re: the web as a killer of novel applications
Post by towr on Aug 7th, 2006, 3:41am

on 08/06/06 at 16:09:26, amichail wrote:
Changing users is not so unreasonable.
It is if it's just to solve a problem that lies with your application. Why is your application so important it demands people to change? If it was that revolutionary, people would change themselves for it.


Quote:
An analogy would be a society where the average person is illiterate.  Changing people in that case would be desirable.
True, but not for the purpose of selling, say, classic novels. Literature class has done nothing for my appreciation of the 'classics'; quite the contrary. If the story has no merit, why should I care 'how well written' is is. If an application doesn't serve my needs, why should I, likewise, care 'how well written' it is?


Quote:
Why not raise the standards of computer literacy to new heights so that we can build more novel applications?
It's hardly a lack of novel applications that a problem. Adding more clutter to the field won't help worthwile projects being discovered.
And most people will appreciate computer literacy as much as mathematics and regulare literacy. i.e. not. You can't force people to appreciate things.


on 08/06/06 at 16:54:56, amichail wrote:
But just how fundamental is computation? Should it not be a required subject in high school?
It should be available, but it's not something everyone should be forced to do. Some people have enough problem with mathematics, and computation builds on that.


Quote:
It seems to me that programming, as the most powerful form of self-expression ever created, is pretty fundamental!
I seriously doubt programming is the most powerfull form of self-expression. First of all, math encompasses programming, and can thus express at least as much. Second, people hardly express themselves in programming in the first place. Natural language is a much more powerfull form of self-expression, it's easier, more available and used more. Most programming on the web serves to promote natural language interaction. Chats, forums, blogs, webpages, etc. The underlying application expresses nothing to the users, except perhaps the wish to communicate.


on 08/06/06 at 16:42:52, amichail wrote:
Marketing is difficult when self-promotion is discouraged on the web.
Spamming is discouraged, not self-promotion. You can make a website devoted solely to your application, you can buy adds, you can add appropriate keywords so search engines will direct people there etc.


on 08/06/06 at 16:38:23, amichail wrote:
This is what I am working on:

http://targetyournews.com/

It appears that most people have no idea what the novelty is or why it is interesting.
It looks a lot like something there's already a browser plugin for, which look better and works easier.
Can't remember the name at the moment though.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 7th, 2006, 3:49am

on 08/07/06 at 03:41:36, towr wrote:
Spamming is discouraged, not self-promotion. You can make a website devoted solely to your application, you can buy adds, you can add appropriate keywords so search engines will direct people there etc.

It looks a lot like something there's already a browser plugin for, which look better and works easier.
Can't remember the name at the moment though.

The novelty here is a method to make free advertising work in a way that avoids spam. The key idea is that the submitter of the advertisement has to target it to the right people. Poor targeting is penalized.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 7th, 2006, 3:55am

on 08/07/06 at 03:41:36, towr wrote:
I seriously doubt programming is the most powerfull form of self-expression. First of all, math encompasses programming, and can thus express at least as much. Second, people hardly express themselves in programming in the first place. Natural language is a much more powerfull form of self-expression, it's easier, more available and used more. Most programming on the web serves to promote natural language interaction. Chats, forums, blogs, webpages, etc. The underlying application expresses nothing to the users, except perhaps the wish to communicate.


One can easily see how mathematics can be embedded in software (e.g., symbolic math packages, theorem provers, etc.). How would you embed arbitrary software into mathematics? Perhaps you are thinking of functional  or logic programming?

As for self-expression, see for example:

http://web.genarts.com/karl/

Automation is a big deal. It's not something that comes with mathematics or natural language.  I believe automation is so fundamental that everyone should learn programming.
Title: Re: the web as a killer of novel applications
Post by towr on Aug 7th, 2006, 4:31am

on 08/07/06 at 03:55:30, amichail wrote:
One can easily see how mathematics can be embedded in software (e.g., symbolic math packages, theorem provers, etc.). How would you embed arbitrary software into mathematics?
What does a CPU do, if not math? adding subtracting, transforming. It's pure math.
Except a computer is finite, and math isn't. (A turing machine is purely a mathematical notion being infinite)


Quote:
Perhaps you are thinking of functional  or logic programming?
Not really. Although all types of programming can be expressed in dynamic logic.


Quote:
As for self-expression, see for example:

http://web.genarts.com/karl/
Most programming is not used in such a way though. And even this seems to be used more as a tool to make pretty pictures, rather than the program itself as art.
Not to mention it doesn't seem to express much. What is it supposedly trying to convey?


Quote:
Automation is a big deal. It's not something that comes with mathematics or natural language.  I believe automation is so fundamental that everyone should learn programming.
If there was no language, there would be no society, there would be no extelligence, there would be no machines. There would be nothing to automate. If there was no mathematics, there wouldn't be complex machines, there wouldn't be computers, there wouldn't be programming languages.

Besides, not all people can learn programming. It's unethical to require people to do things they cannot.
I really think you're overinflating the importance of computer science in the grand scheme of things. It's a very important tool, just like chemistry, biology and physics  etc. And certainly, it'd be nice if everyone could and would learn all those things. But there's just too many 'fundamentals'. And most of those have no direct bearing on people's lives. Unlike literacy which most people use every day, most people don't need to make/read code every day, or for that matter solve mathematical problem beyond simple arithmatic, or physics, biology, chemistry beyond what intuition and experience provides.

You know what they should teach in schools? To relativize.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 7th, 2006, 4:52am

on 08/07/06 at 04:31:38, towr wrote:
Besides, not all people can learn programming. It's unethical to require people to do things they cannot.
I really think you're overinflating the importance of computer science in the grand scheme of things. It's a very important tool, just like chemistry, biology and physics  etc. And certainly, it'd be nice if everyone could and would learn all those things. But there's just too many 'fundamentals'. And most of those have no direct bearing on people's lives. Unlike literacy which most people use every day, most people don't need to make/read code every day, or for that matter solve mathematical problem beyond simple arithmatic, or physics, biology, chemistry beyond what intuition and experience provides.

You know what they should teach in schools? To relativize.

I'm actually disappointed with the way computing has been studied. I've mentioned this here and in other forums:

http://weblog.fortnow.com/2006/07/science-and-art-of-computation.html

http://www.scottaaronson.com/blog/2005/10/doctor-of-hacking.html

http://groups.google.com/group/alt.architecture/browse_thread/thread/11478a1aca448281/d6f500ab4b5c3303?lnk=raot#d6f500ab4b5c3303

I certainly don't see computers as just a tool.

Automation is a great way to demonstrate your ideas to other people. In that sense, they are an excellent tool for self-expression.
Title: Re: the web as a killer of novel applications
Post by towr on Aug 7th, 2006, 7:16am
Btw, here's the plugin I was thinking of earlier http://www.stumbleupon.com/
I'm not sure how it compares precisely to your application, but there's some similarity.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 7th, 2006, 7:23am

on 08/07/06 at 07:16:39, towr wrote:
Btw, here's the plugin I was thinking of earlier http://www.stumbleupon.com/
I'm not sure how it compares precisely to your application, but there's some similarity.

StumbleUpon uses a collaborative filtering algorithm to make recommendations. It is also not intended for self-promotion.

Target Your News leaves the task of making recommendations to the submitter of the advertisement.  

The rationale is that if you are going to promote your product/service, then you will have an incentive to spend some time on targeting it to the right people, particularly since the score of your submission will depend on how good your targeting is.  The score is important since it determines your submission's rank among people's recommendations.

This is actually a real-time game since you can modify your targeting in real-time as you react to the votes that you receive on your link. You also get feedback on specifically which targeting rules are helping you and which ones are hurting you.
Title: Re: the web as a killer of novel applications
Post by towr on Aug 7th, 2006, 7:38am

on 08/07/06 at 07:23:34, amichail wrote:
The rationale is that if you are going to promote your product/service, then you will have an incentive to spend some time on targeting it to the right people, particularly since the score of your submission will depend on how good your targeting is.  The score is important since it determines your submission's rank among people's recommendations.
Wouldn't it make sense to automate targetting based on the scores people give, and the other sites they like?
Datamining would give a much better precision than a person trying to guess based solely on who he targetted and their scores. Aside from the first classification. (Perhaps it'd make sense to start with a small target group, and incrementally increase the targetting group as precision increases)

From the perspective of the user (rather than site-promotor), I'd prefer stumbleupon. I'd have more confidence it'd give me sites I like. Besides it has a more fluffy appeal.
Of course I'm not a websurfer, more of a websearcher, so neither actually appeals to me much.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 7th, 2006, 7:47am

on 08/07/06 at 07:38:29, towr wrote:
Wouldn't it make sense to automate targetting based on the scores people give, and the other sites they like?
Datamining would give a much better precision than a person trying to guess based solely on who he targetted and their scores. Aside from the first classification. (Perhaps it'd make sense to start with a small target group, and incrementally increase the targetting group as precision increases).

I'm betting that submitters can be more intelligent than collaborative filtering algorithms, particularly when recommendations need to be made with sparse data (which is often an issue with recommender systems).
Title: Re: the web as a killer of novel applications
Post by Icarus on Aug 7th, 2006, 6:06pm
The mechanics are not my forte, but here is how I see your site:

I look at your home page, and it doesn't really suggest to me what its purpose is. I have put pieces together - mainly from your three instructions at the top - to figure it out. And note that I came in with some understanding already from this discussion. It also doesn't suggest to me why I would want to take part in this. And, it doesn't tell me why it is that this is any different or any better than other sites. All three are things you need to make clear as possible to new viewers when they first visit the page. That first impression has to catch their attention and give them reason to explore further. Generally, this is done by catchphrases - short easy-to-remember sentences that give the a "top-level" description of what this is, and why it's going to be a great thing for your audience.

When I do go exploring to find out more, what do I see given as the purpose of this site? "Self-promotion". This immediately suggests to me that if I take part, I am going to be bothered by all sorts of people telling me about what they want, not what I want. At this stage, I am saying "No thanks", without even bothering to look further, and figure out that your site is actually designed to weed that out. The way you have set things up, the good is hard to find, but the bad is right out in plain sight! It is no wonder you are have trouble raising interest!

What you need to do is tell people right on your home page why this is good for them: "Tired of hunting for content? Tailor your surfing to what is interesting to you! Find others who share your interests and share sites you like with them!"

On the otherhand - and I cannot emphasize this enough - DROP THE BIT ABOUT THIS BEING TO ENCOURAGE SELF-PROMOTION!! The purpose itself is alright - but the image it conjures when people read it is death itself for your application. If you have set up the site right, people will use it in appropriate ways to promote their interests, and others will find it helpful in connecting them. So your reasons will come about. But when you explain it like that, people are not going to be interested, even if they agree intellectually with the purpose. The reason you see this "taboo" on self-promotion is because too many people have abused the web exactly for that. So when web travelers, wearly from this abuse, see the line, they see more abuse, not the cure you are attempting.

What you really need (REALLY NEED) is talk to someone who knows marketing. You are an idea person - this has been clear about you from your first post here. You are good at coming up with ideas, at developing ideas. But you are lousy at selling them. So you need to work with someone who is good at selling ideas, even if they've never had an original one themselves in their life.
Title: Re: the web as a killer of novel applications
Post by rmsgrey on Aug 8th, 2006, 9:38am
If automation offers such a powerful means of expression, why are we arguing the point in English? If the point is that automation lets you show people ideas, why not use it to show us how useful it is?
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 8th, 2006, 2:24pm

on 08/08/06 at 09:38:05, rmsgrey wrote:
If automation offers such a powerful means of expression, why are we arguing the point in English? If the point is that automation lets you show people ideas, why not use it to show us how useful it is?

Automation subsumes natural languages.

I could for example build a bot that would argue my views of the world on my behalf in English. Automating myself in this way may spread my views more effectively, thus providing a better method for self-expression.

In any case, you can view any research prototype system as a demonstration of some idea, and hence as a demonstration of automation as self-expression.
Title: Re: the web as a killer of novel applications
Post by towr on Aug 8th, 2006, 2:38pm

on 08/08/06 at 14:24:22, amichail wrote:
Automation subsumes natural languages.
I very much doubt it.
Semantics doesn't really fall in the realm of computation.


Quote:
I could for example build a bot that would argue my views of the world on my behalf in English.
Could you really now? One that can respond to questions posed in natural language?
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 8th, 2006, 2:44pm

on 08/08/06 at 14:38:22, towr wrote:
I very much doubt it.
Semantics doesn't really fall in the realm of computation.

Could you really now? One that can respond to questions posed in natural language?

There are many systems that can respond to questions posed in natural language, particularly for limited domains.  

I'm not an expert in this area, but take a look at the research.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 8th, 2006, 2:54pm
BTW, automation need not refer to only computers, but also to people via computers.

Take a look at this fascinating talk on human computation by the creator of the ESP Game:

http://video.google.com/videoplay?docid=-8246463980976635143
Title: Re: the web as a killer of novel applications
Post by towr on Aug 8th, 2006, 2:54pm

on 08/08/06 at 14:44:43, amichail wrote:
There are many systems that can respond to questions posed in natural language, particularly for limited domains.  

I'm not an expert in this area, but take a look at the research.
I have, and the current results are very, very poor. You can have better conversations with a 5 year old.
Certainly, you may get a response to your question, even a relevant one. But not anyway near a natural one.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 8th, 2006, 2:56pm

on 08/08/06 at 14:54:33, towr wrote:
I have, and it is very, very poor. You can have better conversations with a 5 year old.

Sure, but the potential is still there particularly if you make use of large data sets (e.g., everything that the person has said in public online, general knowledge obtained through human computation games, etc).
Title: Re: the web as a killer of novel applications
Post by towr on Aug 9th, 2006, 2:46am

on 08/08/06 at 14:56:31, amichail wrote:
Sure, but the potential is still there particularly if you make use of large data sets (e.g., everything that the person has said in public online, general knowledge obtained through human computation games, etc).
Unless you follow a conversation that was held before near perfectly, it would yield a very disjointed dialog. There is simply too little understanding in current systems. They don't understand context or references. Generally you can't even count on them remembering anything you said before. Just repeating sentences said before won't solve that. The single sentence they repeat may be a natural one, but often its context will fail.
Aside from that there's the problem a computer can't a priori know every word a person might use. So it has to be able to learn, and deduce the meaning of words from their context. Not to mention it has to cope with people's typos and grammatical errors; and outside specific domains things like  metaphors, sarcasm etc.
The field is rife with problems that make natural language processing and generation intractable. At least for the current approach. (I suppose technically you could call people automatons, so a similar automatic system ought to be able to cope. But I doubt we could still understand a system complex enough to deal with it. The computational principles behind it would be as insightfull as the physical description of the brain on a subatomic level. i.e. not.)
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 9th, 2006, 3:10am

on 08/09/06 at 02:46:30, towr wrote:
Unless you follow a conversation that was held before near perfectly, it would yield a very disjointed dialog. There is simply too little understanding in current systems. They don't understand context or references. Generally you can't even count on them remembering anything you said before. Just repeating sentences said before won't solve that. The single sentence they repeat may be a natural one, but often its context will fail.
Aside from that there's the problem a computer can't a priori know every word a person might use. So it has to be able to learn, and deduce the meaning of words from their context. Not to mention it has to cope with people's typos and grammatical errors; and outside specific domains things like  metaphors, sarcasm etc.
The field is rife with problems that make natural language processing and generation intractable. At least for the current approach. (I suppose technically you could call people automatons, so a similar automatic system ought to be able to cope. But I doubt we could still understand a system complex enough to deal with it. The computational principles behind it would be as insightfull as the physical description of the brain on a subatomic level. i.e. not.)

But what happens if we use human computation?

For example, to automate myself, I could build a system with everythng I have said publicly online in its database.

During a conversation, a game would be played by human players in real-time.  They would compete to pick the most appropriate reply to the conversation from the database. A scoring system could be used to reward those who picked good replies, etc.
Title: Re: the web as a killer of novel applications
Post by towr on Aug 9th, 2006, 5:13am
[comment]hmm.. finally found what gave the "error 500" problem [edit]but on second thought, perhaps I shouldn't mention it here, in case it's exploitable..[/edit][/comment]


on 08/09/06 at 03:10:54, amichail wrote:
During a conversation, a game would be played by human players in real-time.  They would compete to pick the most appropriate reply to the conversation from the database.
Foregoing for a moment the problems a player would have to choose a response from such a large database, I very much doubt there will in general be a response that fits perfectly in the conversation.
Interpolation may work reasonably up to a point (some response might be close enough to fit sufficiently), but extrapolation is always the biggest problem of computer learning. Even with a closed domain, you'd need to have said everything there is to say about it, and not only that, but in ways appropriate to any sort of questioning of it. (Alternatively you'll need a wide variety of ways to say "I don't know", without sounding like a robot. There isn't even a decent chatbot that can deny knowledge of everything in a natural way)
And, the players would also have to find the appropriate respons to any sort of question applicable to the domain (keyword spotting simply won't cut it; it scarcely even deals with negation appropriately, let alone more complicated sentence structures)
And that's just dealing with question-response, not even a conversation. In general people in a dialog build on earlier parts of the conversation. So to find the appropriate response can depend on the whole previous conversation.
Title: Re: the web as a killer of novel applications
Post by amichail on Aug 9th, 2006, 5:38am

on 08/09/06 at 05:13:58, towr wrote:
Foregoing for a moment the problems a player would have to choose a response from such a large database, I very much doubt there will in general be a response that fits perfectly in the conversation.
Interpolation may work reasonably up to a point (some response might be close enough to fit sufficiently), but extrapolation is always the biggest problem of computer learning. Even with a closed domain, you'd need to have said everything there is to say about it, and not only that, but in ways appropriate to any sort of questioning of it. (Alternatively you'll need a wide variety of ways to say "I don't know", without sounding like a robot. There isn't even a decent chatbot that can deny knowledge of everything in a natural way)
And, the players would also have to find the appropriate respons to any sort of question applicable to the domain (keyword spotting simply won't cut it; it scarcely even deals with negation appropriately, let alone more complicated sentence structures)
And that's just dealing with question-response, not even a conversation. In general people in a dialog build on earlier parts of the conversation. So to find the appropriate response can depend on the whole previous conversation.

It would be known to everyone that the conversation is being constructed on behalf of someone via a human computation game.

To avoid the problems you mention, some guidance can be given as to where the conversation can go. This can take into account the database as well as the areas of expertise of the players.


Powered by YaBB 1 Gold - SP 1.4!
Forum software copyright © 2000-2004 Yet another Bulletin Board