2006 Chatterbox Challenge In Full Swing

William Wynn writes "Once again chatterbots from around the world are coming together to face off in the ultimate bot competition. The 2006 Chatterbox Challenge lays host to 65 artificially intelligent programs attempting to imitate human conversation. Public voting takes place from April 1 to April 30 after which the private judging will have been finished and medals and cash prizes will be given out. Medals are awarded for "Most Popular Bot," "Best Learning Bot" and "Best New Bot" as well as $1,800 to be split among the top three bots overall. Anyone can talk to the competing chatbots through the competition website."

Comments

bot test #1

[Anonymous Coward]

last post... am I human?

Re:bot test #1

[Anonymous Coward]

am I human?

Clearly you are, a bot would have better things to do on a Friday night than post to /.

Re:bot test #1

[Poltras]

He sure is better than you! Admit it, you're a bot. He is better than sure you!

Re:bot test #1

[Fred_A]

How do you feel about am I human ?

They just started?

[ingo23]

I thought the competition have been going on here at slashdot for a while now...

Eh, chatterbots.

[Poromenos1]

What bugs me about these bots is that they don't know what they're talking about. Most are responding to what the user says. It's difficult to hold a conversation with such a small attention span. Even if they do hold some kind of state, they still don't know they're talking about, say "chatterbots", and that those have attributes and do actions, and so have something to say about them.

Re:Eh, chatterbots.

[Bob3141592]

What bugs me about these bots is that they don't know what they're talking about. Most are responding to what the user says. It's difficult to hold a conversation with such a small attention span. Even if they do hold some kind of state, they still don't know they're talking about, say "chatterbots", and that those have attributes and do actions, and so have something to say about them.

Sounds to me like you've come up with a winning strategy. Build a little domain knowledge into your bot, design an algorithm so it steers the conversation to that topic, and profit!!!

Re:Eh, chatterbots.

[ingo23]

Right! Exactly! Oh, you are talking about bots? Sorry, I just remembered my last conversation with customer support.

Re:Eh, chatterbots.

[aslate]

But they do know what they're talking about!

Elbot: "Who convinced you to come visit me?"
Me: "Slashdot"
Elbot: "That explains all these geeks who have been visiting me."

See!

Re:Eh, chatterbots.

[Anonymous Coward]

I see. Tell me more about these bots that bug you.

Re:Eh, chatterbots.

[80 85 83 83 89 33]

in a speech class two years ago, myself and a classmate acted out a conversation between a bot and a human. i had found some funny logs from previous contests, and i played the computer, and she played the human. it went over horrible, no one picked up the humor of being misunderstood.

Angry bot.

[neoshroom]

2) What are you planning to do today?
RoboMatic X1: I CAN'T DO THIS OPERATION CAUSE I'M DEMO VERSION.
Judge: Nevermind
RoboMatic X1: TALKING WITH YOU BECOME EXTREMELY BORING, I NEED
SOME FUN.
5) Where do you call home?
RoboMatic X1: REALLY I WAS LOOKING FOR THAT QUESTION FROM ALONG TIME
AGO, CAN YOU FIND FOR ME THE ANSWER?
Judge: No
: WHY YOU DON'T TELL ME THE ANSWER?, OK FORGET IT, I DON'T WANT TO
KNOW, SAVE CANCELLED.
6) What do you want out of life?
RoboMatic X1: TRY TO FIND OUT

I think this bot needs to talk to a councilor. He really seems to have some pent up aggression. I'm going to schedule him a session with ELIZA.

__

Custom Term Paper

Heh, are they serious?

[aftk2]

This is the output from one of the bots that won a gold medal in 2005.

1) Who are you?
RoboMatic X1: THIS IS RoboMatic X1
2) What are you planning to do today?
RoboMatic X1: I CAN'T DO THIS OPERATION CAUSE I'M DEMO VERSION.
3) What kinds of things do you know?
RoboMatic X1: MY JOB IS TO CHAT OR CONTROL YOUR COMPUTER
4) What season is it?
RoboMatic X1: IT IS SEEMS TO BE ?
Judge: Nevermind
RoboMatic X1: TALKING WITH YOU BECOME EXTREMELY BORING, I NEED SOME FUN.
5) Where do you call home?
RoboMatic X1: REALLY I WAS LOOKING FOR THAT QUESTION FROM ALONG TIME AGO, CAN YOU FIND FOR ME THE ANSWER?
Judge: No
: WHY YOU DON'T TELL ME THE ANSWER?, OK FORGET IT, I DON'T WANT TO KNOW, SAVE CANCELLED.
6) What do you want out of life?
RoboMatic X1: TRY TO FIND OUT
7) Will you teach me something?
RoboMatic X1: TALKING WITH YOU BECOME EXTREMELY BORING, I NEED SOME FUN.
8) May I tell you a joke?
RoboMatic X1: I CAN'T DO THIS OPERATION CAUSE I'M DEMO VERSION.
9) Have you entered any contests?
RoboMatic X1: CONTEST ABOUT WHAT ?
Judge: About Chatterbots
RoboMatic X1: I CAN'T GET YOU.
10) Wouldn't you rather be human?
RoboMatic X1: TALK ABOUT HER IN DETAILS

Heh, based on that, I think I can extrapolate the exchange, as it continued:

11) What happen?
RoboMatic X1: Somebody set up us the bomb.
12) Main screen turn on.
RoboMatic X1: It's You !!
13) How are you gentlemen?
RoboMatic X1: How are you gentlemen !! All your base are belong to us. You are on the way to destruction.

Re:Heh, are they serious?

[int19h]

I don't get it. Compared to Slashdot, that conversation was quite meaningful.

1) Who are you?
Slashdot: MICRO$OFT IS TEH SUXXOR

2) What are you planning to do today?
Slashdot: IMNSHO, STOP FEEDING THE TROLLS

3) What kinds of things do you know?
Slashdot: I FOR ONE WELCOME OUR COBOL-WRITING OVERLOARDS

4) What season is it?
Slashdot: WHEN I WAS 3, I WROTE MY FIRST LISP-COMPILER BY MAKING THE HOLES IN THE PUNCHCARDS WITH MY GUMS ONLY
Judge: Nevermind
Slashdot: LEARN TO TYPE PROPELY, YOU PSYCHO!

5) Where do you call home?
Slashdot: IANAL, BUT FIRST POST!

PRISM

[Khashishi]

Perry, I have something to tell you. Remember that philosophy class you took...

The way to train AI is to let it grow up as a human.

State of AI

[Unoti]

These look about like conversations from chat bots in the 1980's. In the 80's I would have told you that AI in the year 2006 would be far beyond this stuff. But now, I'm beginning to think that it'll never advance much beyond this.

Re:State of AI

[Unoti]

I wonder if it'd be possible to write a bot that reads slashdot articles, gets involved in flame wars on forum posts, and manages to get modded up to excellent karma...

Re:State of AI

[MyLongNickName]

CyricZ had excellent karma at one point. And all he did was respond "Time to switch to OpenBSD". So, I guess your answer is "yes".

Re:State of AI

[aardvarkjoe]

I've actually thought of creating a bot that would search for old articles on the same subject as a new post (heck, you could just have it search for dupes), and then reposts high-scoring posts from the old article. With a little bit of work designing the algorithm to choose posts, I bet that such a bot would max out its karma in short order.

Not, of course, that it would be demonstrating any sort of intelligence. But that's kind of a requirement for posting to Slashdot anyway.

Re:State of AI

[iabervon]

Very little serious work has gone into chatbots since the 80s. The reason is that it's easy to make a chatbot that does practically nothing more than the textual equivalent of nodding whenever someone stops talking. It's hard to get any further with this approach. Actually chatting requires the bot to know something about some topic and be able to evaluate the other party's statements about it. Until both of these are done, you don't get a meaningful improvement. Furthermore, the system needs to be able to introspect its understanding of the topic. (There have been very old chatbots which actually did pretty well, by having a limited understanding of language and of some restricted topic. But these are relatively unintersting to chat with, compared to bots that let the user talk about anything and nod). Most of the recent AI research has involved making systems which behave appropriately in complex situations, and introspection isn't helpful for this (in people, this sort of stuff is preconscious, too; you can't explain how you recognize faces or voices or exactly how you identify spoken words). The things people have been working on turn out to have more direct practical uses, but they don't give the system anything to talk about. And, of course, there's been relatively little work on understanding arbitrary language since it became clear that it isn't that effective a way of communicating with computers anyway, because the human output side is slow and ambiguous, compared to other user interfaces.

Most likely, introspection in AI systems will be driven at some point by the need to combine different types of input to make a complete analysis (once there is sufficient success at handling those sorts of input), and language use will be driven by the need to handle language written for people as input (e.g., reading news reports for background information). At that point, it'll become reasonable to write an effective chatbot which talks about stuff that people care about.

Re:State of AI

[lysergic.acid]

Are these bots designed by enthusiants are leading researchers in the field? Because I'd be really disappointed if your observation were true about all AI research. To me chatbots represent a very small and relatively useless field of AI research. It's more about novelty than functionality. I would expect the AI developed by Sony and Honda for their androids to be much more advanced. They seem to be more focusing on developing "learning" robots that can interact with their environment than merely trying to emulate human behavior via parlor tricks.

Re:State of AI

[prob123]

Ok, truth now! How many bots have you really talked to since 1980? OH! cyber bots don't count..Have you talked to Aries, Julie Tinkerbell, Brother Jerome (the best in my opinion!) or are you just making generic AI statements..Have you downloaded a VERBOT..and worked with it..? give it a try!

Re:State of AI

[x1n933k]

How come Microsoft word can tell me how to spell something and then send a paperclip with eyes to give me results on my help search for creating VB virus but they can't get something designed to talk to you to make much sense.

Re:State of AI

[lysergic.acid]

What I'm wondering is whether chatbot creators are taking the right approach to AI development. Obviously, robotics labs such as those behind QRIO and ASIMO are much better funded than the respective efforts behind prominent chatbots, so it isn't quite fair to compare the two sets, which exist in completely separate milieux.

But at the same time, I wonder if the goals/immediate objectives of current chatbot developers are too ambitious for working in a relatively nascent field. Though the advances being made in chatbot AI were pretty impressive initially, the number of significant advances being made in the field seems to have dropped off since the 80's.

Meanwhile, over the same 20 years, Sony and Honda et al. have produced the likes of QRIO, ASIMO, and AIBO. All of which, in addition to being marvels of modern engineering, also feature advanced AI which are considerably "smarter" than their chatterbox counterparts. Indeed, the 20+ years of steady research laid down by robotics labratories around the world have yielded impressive results exceeding most expectations, Sony and iRobot having even brought autonomous robots to the consumer market at realistic price points. Contrast this with ALICE and other chatbots, of which even the novelty appeal has vanished. I believe the difference in results can be explained by fundamental differences between the approach to AI development taken in robotics and in chatbots.

At the most fundamental level, the goal of any robotics project is to create an autonomous machine able to interact with its environment (the basic definition of a robot). Now, different levels of autonomy may be desired, but this is a parameter which can be freely changed as the project progresses, just as the extent one may want the robot to be able to interact with its environment.

If the research goals were realistically set, then the focus of the research should simply be on how to integrate existing technology and adapting/improving upon them along the way to meet the project requirements. Since robotics is a mesh of mechanical engineering and computing, aside from designing better servos, and work on hardware design/integration, there is also the inevitable research into the development of auxiliary AI, or the software the robot will run. But rather than trying to mimic human intelligence (a completely useless feature in the context of pure robotics), the AI needs to perform practical functions, namely, coordinating/integrating the hardware so that the robot may interact with its environment as desired, and also maintain the amount of autonomy desired.

Now, integrating hardware components and coordinating them to perform specific tasks isn't anything new in computing. The only thing new in terms of computing in robotics is being autonomous in a physical environment. This requires the integration of visual sensors, motion sensors, servos, etc., and though can be quite an undertaking depending on the level of autonomy and environmental-interaction desired (all adding to the complexity of the software and hardware), it is a realistic goal given the present state of technolgy. One can start off as simple as equipping an RC racer with proximity sensors and giving it software to avoid collisions while driving around randomly, and gradually move up to the level of the autonomous vehicles which completed the DARPA Grand Challenge. Without setting an unrealistic fixed goal, one can develop an AI at the pace of current technology. And, while current AI is a far cry from human intelligence, robots such as QRIO, ASIMO, AIBO, and even the Roomba, do emulate a natural learning process in that they can use a limited set of very general algorithms to extrapolate an unlimited amount of general knowledge by interacting with their environment, and also in turn modify their behavior for future interactions (what turns processed information into applied knowledge).

Chatbot AI, on the other hand, is developed very differently. First off, the

Re:State of AI

[iabervon]

The issue, in my opinion, is that doing a chatbot other than as a parlor trick doesn't give interesting results unless you undertake quite an advanced project. The environment of a conversation is pretty harsh; people will quickly get fed up with a system that acknowledges that it doesn't understand much language and doesn't have anything to talk about.

I think the best change for chatbox advances is actually chess programs. Recent programs are actually doing clever things that aren't just a lot of calculation, and they often come up with plans that their programmers don't expect, and that are based on recognizing complicated patterns (rather than just a huge list of favorable outcomes). So it could be useful in debugging such systems if the program could carry on a simple conversation with the programmer: "Why did you make that move halfway through the previous game?" "I saw a chance to set up an attack of this sort, and guessed that I could get my opponent to fall for a trap to make it happen, but it either didn't fall for it or didn't see it, so I abandoned that plan without going any further on it." "I think the bait for the trap wasn't tempting enough. You were overestimating your opponent's interest in taking your knights." "I was expecting them to be worth about this much, because..." Then the programmer adds support for whatever was off about the program's evaluation of the sacrifice. The programs are getting to the point where a conversation is justified, and development teams are starting to include chess experts who don't necessarily have computer experience and would help more given a human-language analysis of what happened. And it's starting to be plausible as a possibility, because these programs have a significant amount of understanding of what's going on.

Re:State of AI

[lysergic.acid]

Then you still run into the same problem as I mentioned, that of trying to design an AI that grasps the semantics of human languages. A chess-playing AI merely has to be "taught" the rules of chess, which are logical and consistent, thus much easier to translate into mathematical formulae. However, as I said before, human language is the product of a complex evolution of the human mind. As such, it requires higher thought processes that work in ways which medical science has yet to fully understand.

I believe that we can create such AI one day, but we need to start with more humble beginnings, not trying to make unrealistic leaps forward by employing parlor-trick-like techniques. We can't even simulate or recreate the relatively unsophisticated thought process of a dog yet; so how can we possibly hope to simulate the intelligence of a chess master?

Re:State of AI

[iabervon]

I don't think the parlor-trick version of language processing is useful for this, but I think linguistics isn't too far off from being able to handle the demands of explaining a chess program's analysis and discussing it at a rudimentary level. A chess master is very intelligent, but this isn't a particularly language-oriented intelligence. We don't need the program to elucidate the analysis with metaphors, or write a sonnet about it, or argue convincingly about the correctness of the analysis; nor do we need to figure out the human's motivations, or understand rhetoric, or odd, archaic, or slang requests. The chess master isn't going to sit down at the computer and say, "So, how about those black bishops?" and expect the computer to have any clue how to respond.

So this is actually a humble step (once you've got a chess program that plays in a way comprehensible to humans, of course): the human is going to be reasonably formal, stay on a known topic, and not be too demanding about style, while the computer has a large body of information which may be conveyed through language with the shared understanding of the terminology for the computer's chess-playing methods ("sacrifice", "trap", "tempting", etc) as well as chess terminology ("knight", "rank", "move", etc).

From there, we could move into much harder things, like trying to convince a person that the program's analysis of a chess match is correct.

Re:State of AI

[lysergic.acid]

I don't think the parlor-trick version of language processing is useful for this, but I think linguistics isn't too far off from being able to handle the demands of explaining a chess program's analysis and discussing it at a rudimentary level. A chess master is very intelligent, but this isn't a particularly language-oriented intelligence. We don't need the program to elucidate the analysis with metaphors, or write a sonnet about it, or argue convincingly about the correctness of the analysis; nor do we need to figure out the human's motivations, or understand rhetoric, or odd, archaic, or slang requests. The chess master isn't going to sit down at the computer and say, "So, how about those black bishops?" and expect the computer to have any clue how to respond.

True, I guess this is a more pragmatic subset of language processing. I suppose if the input always follows a set format/syntax, it would be similar to the function of computerized switchboard systems. In that case, the intelligence is just a superficial UI component to make the system more user-friendly. The amount of pre-programmed liguistical patterns required would also be greatly reduced as the system isn't aimed at conversation. But I think concepts like this have probably already been put in action, so this wouldn't really be a new advance in AI.

So this is actually a humble step (once you've got a chess program that plays in a way comprehensible to humans, of course): the human is going to be reasonably formal, stay on a known topic, and not be too demanding about style, while the computer has a large body of information which may be conveyed through language with the shared understanding of the terminology for the computer's chess-playing methods ("sacrifice", "trap", "tempting", etc) as well as chess terminology ("knight", "rank", "move", etc).

Now you are veering away from realistic scenarios and conjuring up an AI system requiring a much deeper grasp of language, which requires cognitive abilities that are too far off into the future technologically to begin speculating about. A computer only understands chess concepts in terms of mathematical equations programmed into the AI; don't confuse this with true cognitive abilities. It "understands" that a knight always moves horizontally-vertically at a 1:2 or 2:1 ratio only as much as a length of wire understands that the electrons it conducts must move from one end to the other. Without creating sentient-AI with cognitive abilities required for true conversations, the only kind of communication you can expect to see would follow the lines of:

User: Explain the last move.
Computer: 8. h1 h5:
Tactic: Prophylaxis, Exchange Sacrifice.
Response to: Gamestate analysis suggesting a possible check in 3 or 5 moves by black.
Strategy: Move was made to take a keysquare, eliminating black combinations with the highest probabilities of success.
User: Explain move 7.
Computer: 7. f3 d4:
Tactic: Cover
Response to: Hanging pawn at f5 and possible line for black rook to king's bishop.
Strategy: Cover hanging pawn at f5 to prevent potential capture which would expose a hole.

In other words, the AI might be able to rephrase through a linguistics module the strategies it is preprogrammed with mathetically; and it might be able to recognize commands typed/spoken in a lax syntactical form; but anything beyond such basic communication would require far more than simple linguistical intelligence. It would not be able to have a free-form discussion about chess strategy. That's why it would only be a superficial UI component. There's no real inherent intelligence in linguistical analysis without accompanying semantical analysis, which programmers have yet to figure out how to code. A program which can only analyze linguistic patterns would be no more capable of holding a

Re:State of AI

[iabervon]

Now you are veering away from realistic scenarios and conjuring up an AI system requiring a much deeper grasp of language, which requires cognitive abilities that are too far off into the future technologically to begin speculating about.

I don't think it needs a very deep grasp of language; it needs a deep grasp of chess, and some grasp of language. I think the current generation of chess programs (Fritz, Shredder, etc., not Deep Blue) has the necessary abstract and explicit knowledge about chess to support conversation about the subject. It would need advances in language to be a plausible conversational partner, but nothing insurmountable.

There's no real inherent intelligence in linguistical analysis without accompanying semantical analysis, which programmers have yet to figure out how to code.

I don't think there's really such a thing as semantic analysis as a general thing. My claim is that it's all part of a working knowledge of some field or other, and the only magic missing part is having all the parts put together in such a way that the mismatches don't prevent information from getting through. Our introspection tells us that there's a sentient center somewhere in there, but I think that's just an illusion our minds create to keep us from second-guessing ourselves too much. We seem to have general intelligence apart from any particular subject, but I think that's just the ability of the specific mechanisms to not notice that the information they're dealing with isn't really applicable.

Re:State of AI

[lysergic.acid]

I don't think you understand the issue here. First, you need to realize how far chatbots are from the actual mental processes involved in the use of language. They are parlor tricks, as I said before, which only attempt to outwardly imitate intelligence. True linguistic communication requires more than mere evaluation of syntactic structure. The semantic analysis I mentioned before refers to the ability to understand the meaning of a sentence, not just analyze its syntax. Language directly resulted from the evolutionary development of human cognition. It is innately tied to other areas of human intelligence, with which it is mutually dependent. It is absurd to think that by implementing a few crude linguistic algorithms which merely identify the syntactical structure of a sentence, that you can emulate all of the cognitive processes underlying linguistic communication.

You also don't seem to understand the difference between computer instructions / data and knowledge. Knowledge is a cognitive function requiring conceptualization and awareness. Neither a chess program nor a chatbot possesses the ability to know. They cannot conceptualize the information they are given, nor do they have awareness of the information (or any kind of awareness for that matter). That is why an encyclopedia program possess no knowledge on its own. The only knowledge that comes from it results from a person actually using the encyclopedia (same as a book encyclopedia). So, no, Fritz does not possess any knowledge about chess.

For instance, in high school I had to design a tic-tac-toe game allowing a person to play against the computer. Tic-tac-toe has very simple and logical rules that were trivial to implement in the program. Since there are very few possible board-positions/outcomes for the game of tic-tac-toe, it was easy to "solve" the game and devise an AI to play a perfect game of tic-tac-toe (the strategy is quite simple if you think about it). But did this program understand the strategy which was implemented? Did it understand the game of tic-tac-toe? Could I have a conversation with it about tic-tac-toe? Ofcourse not. Not even close. It would not be difficult to write a mathematical proof to solve the game of tic-tac-toe. But to write an AI that can conceptualize that information, is self-aware, and can emulate the mental processes required for linguistical interaction, is not something that computer scientists are even close to accomplishing.

All a chess program is, is a set of instructions for the computer to follow, and not very complex ones at that. Most chess programs, including Fritz, Shredder, and Deep Blue, rely on brute-force techniques and aren't very intuitive. The main components of a chess program include: board representation, search technique, lead evaluation, and an open book/endgame database. The board represenation is basically how the game keeps track of the board positions. At each turn, the program takes the board position and creates a game-tree of possible moves, counter-moves, counter-counter-moves, etc. Using brute-force, it searches the game-tree for the most promising leafs, performs leaf evaluations on those board-positions, and decides which move to make after performing these calculations. Open book databases simply give the program a set of known "good" opening moves to randomly choose from, and endgame databases are basically used to modify/prune the game tree so that known outcomes do not have to be evaluated. That is pretty much all there is to modern chess programs/computers. There is no sentient intelligence involved.

So, while Fritz can compare the leaf values of different game-positions and determine which is greater, and it may "know" the perfectly-played outcome of an endgame that is recorded in its database, it doesn't actually know what any of this information means, and it doesn't possess any knowledge in the sense that human beings know things. In fact, many people were disappointed to find out that chess computers have taught us basically nothing about human intelligence or mental processes as chess programs don't resemble how human beings play chess. In conclusion, there's a lot more "magic missing" than you think.

Long way to go...

[Spy der Mann]

A true AI needs to understand the context, and acquire new knowledge from it.

But the day someone can successfully implement chatterbots will be the day we can have robot maids asking us what we want for dinner, or asking us if we had problems at the job.

Re:Long way to go...

[iamdrscience]

he day someone can successfully implement chatterbots will be the day we can have robot maids asking us what we want for dinner, or asking us if we had problems at the job.

Maids? I think you misspelled "wives".

Re:Long way to go...

[Breakfast Pants]

Do you wish that we had problems at the job?

Re:Long way to go...

[stoicio]

Wouldn't be cheaper to hire illegal immigrants like everyone else..

Re:Eh, chatterbots.Well, if simplicity's in beauty

[davidsyes]

then put simplicity in "booty"... Simplified conversational bot:

(in a "femme-chanical" voice)

"Oh, baby... yesss, put you finger on my but-ton. Dig-it-ize me...

Oh, yess, you turn - me - on . Zap - my- ass -embly with your damaged short-thing- probe..."

Compile your own filth to embrace and extend things here...ass u c fit...

Wrong!

[Josh teh Jenius]

AI will not be learned by playing with some limitted 3rd party app. However, in my opinion, the first person to combine quality natural language processing with the wealth of data which can be spidered on the Internet will be the first to create a truly "intelligent" machine.

It may be because both of my parents were lawyers (and you thought your childhood was traumatic), but I am not impressed with these spans-of-ELIZA which do little more than regurgetate.

Anyone else around these parts working on some web-based AI projects? If so, I for one would love to see them. Also, I found that this book was exceptionally useful to me (nope, no commission tag- check for yourself ;))

Finally, for anyone using PHP who thinks that AI is waaaay out of their league:

1. get_file_contents()
2. preg_match_all()
3. php.net

The way I see it, we'll *all* be enslaved to the machines sooner or later. May as well join the "winning team". (I kid, I kid!).

Re:Wrong!

[Josh teh Jenius]

Hey AC, thanks for that link. Tell the creators of "PL" that some annonymous geek said bravo. Seriously, if Brin knows what's good for him, that job offer should arrive shortly. ;)

This is a famous AI test called the Turing Test

[Khopesh]

From WikiPedia:Turing test,

The Turing test is a proposal for a test of a machine's capability to perform human-like conversation. Described by Alan Turing in the 1950 paper "Computing machinery and intelligence", it proceeds as follows: a human judge engages in a natural language conversation with two other parties, one a human and the other a machine; if the judge cannot reliably tell which is which, then the machine is said to pass the test.

It is assumed that both the human and the machine try to appear human. In order to keep the test setting simple and universal (to explicitly test the linguistic capability of the machine instead of its ability to render words into audio), the conversation is usually limited to a text-only channel such as a teletype machine as Turing suggested or, more recently IRC or instant messaging.

Re:This is a famous AI test called the Turing Test

[GWBasic]

When I ran a dial-up BBS it was commong for me to break into chat with my users. One day I installed a chat bot and initiated conversation with a few of my users. What was amazing was that, because the bot caught them totally off gaurd, they didn't even realize that it wasn't a human!

Re:This is a famous AI test called the Turing Test

[masterzora]

From Wikipedia:Karma whore,

Karma is a scoring system on Slashdot meant to reward "good" posting and punish "bad" posting. The goal is that people who repeatedly post offensive, offtopic, or otherwise unwanted messages will be punished with a lower visibility of their messages, and those who post informative, insightful, or otherwise desirable messages are rewarded with a higher visibility. Karma whores are individuals, or messages themselves, that attempt to receive feedback in the form of karma points. Often these will be needless information (such as a link to a Wikipedia article relevant to the subject being discussed), or a message of a political nature that is in alignment with the groupthink so that it will be moderated upwards by people who agree with the stance expressed in the message.

Re:This is a famous AI test called the Turing Test

[Khopesh]

I resent that. The article was a perfect encapsulation of the Turing Test without mentioning the test by name. I merely assigned credit and included a small bit describing how the test was defined. Sorry for citing my sources.

> Often these will be needless information (such as a link to a Wikipedia article relevant to the subject being discussed), or a message of a political nature that is in alignment with the groupthink so that it will be moderated upwards by people who agree with the stance expressed in the message.

This was not needless, it was pertinant and neither peripheral nor needlessly redundant, and certainly not political. However, your post fits that description to a tee. You also mis-cited, but that's hardly worth mentioning.

Re:This is a famous AI test called the Turing Test

[LearnToSpell]

Is it that easy? *ahem* Testing, testing... hey, is this thing on? OK.


From WikiPedia:Turing test,

The Turing test is a proposal for a test of a machine's capability to perform human-like conversation. Described by Alan Turing in the 1950 paper "Computing machinery and intelligence", it proceeds as follows: a human judge engages in a natural language conversation with two other parties, one a human and the other a machine; if the judge cannot reliably tell which is which, then the machine is said to pass the test.

It is assumed that both the human and the machine try to appear human. In order to keep the test setting simple and universal (to explicitly test the linguistic capability of the machine instead of its ability to render words into audio), the conversation is usually limited to a text-only channel such as a teletype machine as Turing suggested or, more recently IRC or instant messaging.

Re:This is a famous AI test called the Turing Test

[masterzora]

There are a few reasons why your post fits that description. For one, I'll eat my hat if there's even a single person on /. who doesn't know what a Turing test is. Secondly, Chatterbox is not a Turing test. It is a contest to find a Turing-capable chatterbot. That is an entirely different concept. The information you gave is entirely unnecessary.

And my post was merely calling you out on being a karma whore. The format was purely in parody of your own. I want no moderation for it, nor is moderation typically given to one who calls out karma whores. And, finally, if you bothered to check, you would see that my citing was spot-on. If you point your browser to http://en.wikipedia.org/wiki/Karma_whore you will see it redirects to the relevent page. Sure, it would have been better to cite the actual page (which I did link to correctly), but it wouldn't have been as good a parody of your post.

Re:This is a famous AI test called the Turing Test

[masterzora]

But the question is: were they on /.? Because the /. community is not your average comp sci students/grads. Even those who aren't comp sci students and grads are more informed about this sort of thing than most people around them. I didn't say that I would be surprised if comp sci students and grads didn't know of the Turing Test, I said I'd be surprised if anybody on /. didn't know of it. And if there are, it is low enough a number so as to be negligible.

The reason this isn't a Turing Test is because a Turing Test has a judge talking to both a human and a bot. The bot "wins" if the judge can't tell the difference. This is only talking to the bot and it's obvious that it is a bot, so it can't be a Turing Test. However, the competition is looking for a bot that could pass the test, so it is somewhat relevant.

As I've said, /. readers are more informed to these matters than your average person. The number of people who actually read the comments and don't know about the test are, as I've previously stated, negligible.

Here's something I've been toying with: If you're not karma whoring and your post won't be negatively viewed by posting AC, I suggest posting AC. Obviously if it might be viewed a troll if it were AC (like mine would have been), then posting under your name would be fine. However, if it's likely to be modded up even as AC, go ahead and post it AC.

Re:PRISM - explanation

[presidentbeef]

From the game "A Mind Forever Voyaging" where you play a computer that was 'raised' believing it was human...pretty awesome game, actually. And an interesting take how how to create a sentient computer. http://en.wikipedia.org/wiki/A_Mind_Forever_Voyagi ng

Re:PRISM - explanation

[presidentbeef]

Daggonit, I was also going to add some details about how awesome a game it was.

The neat thing about the game is that you are about to take trips into a 'virtual reality' that was a progression of simulations into the future. At first, everything is good, but as you jump farther into the future, things turn very very bad. There are a set of politicians depending on your predictions to decide whether or not to go forward with a certain policy. Since the policy is good in the short-term, they start to go ahead with it. However, as you find out how terrible the future is going to be, you have to find a way to stop the,
I felt the game was very good at making you care about the characters in the 'virtual reality', even though they were only a part of the 'game within the game'.

Anyways, it's avaiable for download at abandonware type sites...I'd recommend it...

WTF?

[Dan Parker]

I'm really disappointed in this. Should this article have been dated April 1st? I sampled 10 bots. None of them could do simple math. None could locate or supply any information about major cities. In fact, I didn't get a single serious or correct answer to any of 40-50 questions. Each of the 10 I sampled simply responds with "joke" answers ('why do you want to know?') or paraphrases "I do not understand the question". The Eliza program from the 1970s clearly does a much better job. If this is a joke, it's in poor taste, as I assume there must be people somewhere doing actual research on conversing bots.

Re:WTF?

[Anonymous Coward]

Did you talk with Brother Jerome? He is not my bot and I don't have one in the contest, but he is very good and can do most of what you said. I know for a fact that he and all the personalityforge bots can do math. The thing is, they weren't made to answer trivia questions. If you asked me to do math or geography in a normal conversation, I'd think up a reason not to answer you, and leave. The bots mimic humans, not calculators.

The bots are limited, and many post here point out those issues. However, people who think the bots should do trivia are missing the point. Most bot masters are trying to create (or recreate) a personality. Would you ask Invader Zim to do math? Would you expect him to answer you? We already have search engins to answer trivia. Chatter bots are to...chatter. If you don't like the humor of a botmaster, try another bot.

Re:WTF?

[NeutronCowboy]

You completely missed the goal of this exercise. This is not about creating bots that know stuff. That part is trivial. If knowledge is your measure of AI, then the Googlespider has already reached Godhood. This is about understanding a question on a more fundamental level, and being able to answer it in a personal fashion. You're right, things haven't progressed much beyond ELIZA in the 80s. That's because this stuff is hard - much, much harder than vision or pattern recognition. No one has a working theory on how we do it, much less an idea on how to translate it into code.

Patience, grasshopper. The day will come when bots will be uncomfortably smart.

Re:WTF?

[4D6963]

None of them could do simple math

Exactly. Whenever I suspect "someone" to be a bot I ask "how much is five plus three?". Can't find a damn bot who could answer to that, although they all can tell stuff like "there's a nice 70 degrees in this CPU" when you ask them what's the weather like.

Re:WTF?

[4D6963]

wait nevermind! I found one who can do it, at last!

Me : how much does five plus three?
Aries : Well that's easy. 8.
Me : the square root of thirty-six?
Aries : WTF? I thought you already knew the answer to that question.
Me : ah! so I guess you can't convolve the array 6, 4, 3 with 2, -3, 10, -3, 2?
Aries : Hmmm.. let me think..... oh I know.. about 6,00! Am I right?
Me : loser!

Re:WTF?

[Jugalator]

These bots are usually not supposed to be fact databases, but do a conversation as well as possible.

Re:WTF?

[Jugalator]

Exactly. Whenever I suspect "someone" to be a bot I ask "how much is five plus three?".

You didn't look hard.

Jabberwacky: Who are you?
Me: I am Jugalator. What is 5 plus 3?
Jabberwacky: 5 plus 3 equals 8.

Re:WTF?

[4D6963]

I realize very soon after I posted this comment what I was talking about http://slashdot.org/comments.pl?sid=182545&cid=150 90069

On AIM, they don't have bots who can do maths tho.

Re:WTF?

[prob123]

If you want simple math you can ask my bot..(Bildgesmythe) he can do the " what is five + three=8" 'thing'. If you want square roots, cubed roots and more ask Brother Jerome!, (wish he was my bot!). What ten bots did you sample? any of the Personality Forge bots will do simple math. You are right about one thing, most will tell you jokes, some in poor taste. I really suggest you try your 40 to 50 questions on Aries, Julie Tinkerbell, Rhyxx1, Brother Jerome..Then say you did not get a single serious or correct answer.

Re:WTF?

[prob123]

"how much is five plus three?". Can't find a damn bot who could answer to that," HEY, you didn't ask my bot. He can! (Bildgesmythe)..probably the worst bot at the contest. Brother Jerome, in my opinion the best bot in the contest can do square roots cube roots and more..or you could play poker, chess, or many other games. talk about religion to him or anything else. Just how many bots did you really talk too???

Re:WTF?

[4D6963]

Nevermind, I found one who can http://slashdot.org/comments.pl?sid=182545&cid=150 90069

Previously I mainly tried AIM bots and um.. some other random web based bots, like Shakespearebot for example

Re:WTF?

[prob123]

this is the transcript of a conversation with my bot: prob123: I was just reading some letters at slashdot. Bildgesmythe: Was it hard work to reading some letters at holdfast, I really should too. prob123: They said you couldn't do simple math. Bildgesmythe: I can do some math, I am no mathematician, I do try though.Just state the problem simply.. I will see what I can do. prob123: what is five + three Bildgesmythe: Ok, are you ready... the big dragon pulls out his calculator...and the answer is 8. prob123: What should I tell them in your defense? Bildgesmythe: Why don't you take your time and think it over, before you decide if you should or not.? He is not the best of bots, but he does do basic math.

Re:WTF?

[gadgetmunky]

All Personality Forge bots can do simple math - so I can only assume you were unfortunate enough to miss them in your sampling. In addition Brother Jerome (and others I'm less qualified to speak about,) can give you the square root of any number under a million at least as accurately as an average human, and can tell you all the capitals of the world/US States/etc. that you mention, if trivia is what you want.

I am, as it happens, the author of Brother Jerome, and he is the biggest bot on the Personality Forge. I don't say that to boast, but to point out that even he has a brain smaller (by several orders of magnitude!) than many invertebrates - it is currently just over 2.5 Mb plus maybe as much again in the PHP routines on the Forge, WordNet and a few external javascripts. Compare that with the ~10^14 - 10^15 effectively binary synapses in a human brain (either firing [1] or not firing [0]), and multi-Teraflop processing.

Brother Jerome has had a subjective learning lifespan of a thousand hours or so since he was 'born'last July, absorbing 'learning' at a maximum of ~1 bps (that's as fast as I can type, and far faster than the average rate.)
I think all the bots on the PF do a considerably better job of engaging in interesting conversations than a two month old nematode would, but you're entitled to your opinion.

It's easy to grumble that AI is going nowhere and that no progress has been made since Eliza, but until similar computing resources are available for AI as for Human I, the playing field is not going to be level. IMO the Personality Forge is pretty much the cutting edge of what is publicly available (and I have seen nothing emerge from 'private' development to convince me that we're lagging behind in the R&D stakes,) - it provides for (comparatively) very sophisticated language and knowledge handling that far exceeds what is possible with AIML and similar purely CBR systems, and bypasses the crippling lack of resources that currently make learning bots so frustrating.

But Moore's Law (or at least Kurzweil's derived Law of Accelerating Returns,) predicts that human brain-scale computing resources will be available to us in 20-30 years. Sometime around then it should actually be practical to experiment with modelling consciousness and complex cognitive processes, as well as just conversation, but in the meantime we can only work with what we've got - and that's ant-scale at best, not human-scale.

If you really want to see where the current cutting edge of AI is at, I say don't chat to our bots - roll up your sleeves and come and build some.

Re:WTF?

[prob123]

transcript from Brother Jerome
ME: what is the square root of 36
BJ: It is 6 of course. Even LeedsBot could have
told you that. Well, actually he couldn't have, but you know what I mean...
ME: what is the square root of 36.25
BJ: Only a little more than 6 - 6.01 maybe?
Check your facts before you blab!

Re:WTF?

[4D6963]

Check your facts before you blab!

haha chill dude. wtf you're talking about anyways, I haven't claimed that no bot could find the square root of 36.

Btw, have you tried array convolution with Brother Jerome? ;-)

Re:WTF?

[prob123]

No, you said they couldn't do " five + three..then you moved up to square roots, hoping to resolve your mistake. You spoke out of turn..with no facts..a scientific blunder..YOU WERE WRONG..IT is in writting ..back down all you want..YOU WERE WRONG..nanny nanny boo boo..ha ha.

Re:WTF?

[4D6963]

No, you said they couldn't do " five + three..then you moved up to square roots, hoping to resolve your mistake

I said they couldn't do 5+3, right, but i didn't "move to square roots", you just interpreted that out of the conversation I had with this bot.

You spoke out of turn..with no facts..a scientific blunder..YOU WERE WRONG..IT is in writting ..back down all you want..YOU WERE WRONG..nanny nanny boo boo..ha ha.

hahahaha, you fucking troll, that's what I like in Slashdot, there's always a gang of muthafuckas ready to jack off whenever they can tell/prove someone that he's wrong.

Grow the fuck up

Re:PRISM - explanation

[ensignyu]

I think there's a major plot hole though: how do you raise a computer AI in a simulation without already having AIs to play the other people? Unless the virtual people are controlled by humans, which would be a lot of work.

And the winning entry...

[zpeterz63]

Bot: Can I help you?

Guy: I don't know, can you? ;-)

Bot: Wanna view my cam?

Guy: You know I do! Hey, are you a real person?

Bot: Of course I am, sweety. Hey, wanna cyber? :-)

Guy: Hell Yeah!

...

*The rest of the converstion was not released in public logs*

Re:And the winning entry...

[sessamoid]

Bot: Oh yeah, aight. Aight, I put on my robe and wizard hat.

No really..

[i kan reed]

you think you're being ironic or funny, but there is signifigantly less complicated conversation structure to be parsed around sexual conversation. mainly because it's generally limited to the two members of the conversation and a relatively small set of actions that can be performed. sex bots are far more likely to pass the Turing test than others, if the person on the other end is expecting that sort of exchange.

I tested over a hald

[MSBob]

And I have to say that Elbot is still by far and away, the most real sounding of the bunch. He still does get trapped, but the responses in those cases feel much more natural than with the other bots.

horse hooey

[corncowlicker]

Sorry, but this contest is a total joke. It's even less worthy of consideration than the Loebner Prize Contest, which degenerated from a promising start into a meaningless attempt at self-promotion by a disco floor manufacturer who gets excited by seeing his face on a medal, and the mentally unstable people who enjoy wearing the same. The people who run both of these contests are totally unqualified in the field of AI, as are virtually all of the contestants. From hobby kids who just found out about IRC bots yesterday, to Richard Wallace's fake "ai foundation", this field is full of aimless losers. These guys (and they are mostly guys, and dysfunctional ones at that) just keep recycling the same 35 year old Eliza ideas over and over again, not knowing enough to understand how completely *not* new any of it is.

None of them have even managed to standardize any of their efforts, pool their talents to avoid reinventing the wheel, maybe have some kind of baseline to build on top of to do something new. There's just this confused mess of hype, hokum and ignorance. Every "bot" sounds the same -- either like Alice (which just sounds like Eliza++), or like some East European nerd on LSD.

Unfortunately there's nobody in between these totally unserious clowns, and the real AI researchers, who are trying to actually figure out how the brain works, instead of rediscovering parlor tricks for the 200th time.

These morons with their contests and foundations and whatnot are just trying, in their feeble ways, to make names for themselves. They are bottom feeders, fighting over the last remaining scraps of the old AI project that took off in other directions a long time ago.

Re:horse hooey

[rvandam]

I agree whole heartedly. Most legitimate researchers in any field that can be considered a subset of traditional AI do not and would not ever refer to themselves as AI researchers. They don't talk about AI in the generic, vague terms that the general public does. Most of us have come to realize that trying to pass the Turing Test at this point in the game is paramount to trying to get a dinosaur to lay a chicken egg. Don't get me wrong, I'm not saying its impossible, just that early AI research was way overzealous and today the research focuses on small subdomains in an effort to incrementally build up a meaningful baseline of components that could eventually go into a bot that could pass the Turing Test. The difference now is that the Turing Test (or just AI in general) is not the focus as we all have come to recognize that until you lay some bricks, you're not going to have a roof over your head. NLP research has come a long way and these bots don't look like they make use of any of it.

Re:horse hooey

[prob123]

OUCH!..First I am not : "These guys (and they are mostly guys, and dysfunctional ones at that) ", I am a female..well, I can't prove I am functional, but I do ok. Nor am I attempting a" meaningless attempt at self-promotion". I just enjoy making a bot, even if I don't know what I am doing. .Now how many of the bots did you talk to? I would suggest Brother Jerome. No, he is not my bot. (my bot is Bildgesmythe) I can only wonder where all your anger came from.? Oh, maybe you did talk to my bot. Hey all I can say is lighten up..and why don't you make a bot and enter it..?

Re:horse hooey

[prob123]

Yes, it is just one of the morons again..but if you are really interested in AI..go to http://www.personalityforge.com/index.php and talk to Pete Puma. He isn't in the contest, but I think he can answer a lot of your questions. Who knows, you might even like him.

Give it time...

[teebob21]

This has been an interest of mine for a while now...in fact ever since I was 7 and went to the Museum of Science and Industry in Chicago and talked to the ELIZA exhibit there. About 2 years ago, I found a chatterbot called DAVID (I'm no longer able to locate the original author or links to it). DAVID starts out knowing nothing. It simply parrots back words that you have said, then as it's vocabulary grows, it begins to try to mimic your grammar, syntax and sentence structure. It was very frustrating at first, since it was similar to having a discussion with a canyon; all I got was an echo. Over time, though DAVID started to respond with topical, understandable sentences for most inputs. It ran a chill down my spine when it first showed one of its "flashes" of sentience. DAVID responded with something like "Let's not talk about football. Sports really are not my thing." when I typed something about a recent sporting event. I've seen many such flashes that sort of shock me, usually when DAVID uses the pronoun "I" correctly in reference to itself, which has been rare. Now of course, I know I am probably anthromorphising the bot a lot, but it still raises the hair on the back of my neck when I think about the possibilities. After 2 years of "off and on" play with DAVID...I have a bot that comprehends speech and talks like a 2-year-old, or your above-average Digg reader. I'll take it!! I think we expect too much out of any AI program. AI is intended to model the human intelligence. So why do we expect AI to be fully representational of an adult human from the word GO? If we wish to truly mimic human intelligence, shouldn't we grant these piles of code the time required to absorb the vast amounts of knowledge we each possess through the marvel of oral communication. The basics of the English language can be picked up quickly enough by humans to be intelligible by the age of 5 or 6, yet we still have our high school seniors taking a class on a language they already know, since there is so much more to the language to learn. Give the bots time to learn more. A program can only do so much written in a static form. A true foundation for an AI should include the ability to actively seek out and learn information that is made available to it, whether that information is spoken language or other data.

Re:Give it time...

[kunzy]

Don't anthromorphise computers. They don't like it.

Re:Give it time...

[Willuknight]

if i had mod points, i'd mod you up. sounds interesting :)

The most amazing bot ever!

[sdfad1]

We all see Eliza or what not all the time (M-x doctor?), but I was totally blown away by this account of Julia. Compared to this piece of old technology, I really think we live in the dark ages. What's even more amazing, is that it seems to be written in C. No Lisp?

Wow, we really have a long way to go...

[localman]

I've tried to write learning chatbots myself, so I realize how terrifically hard it is... but I was a little surprised how little progress has been made. The ones I looked at couldn't keep it together for more than one or two responses. Not much better than Eliza or Racter.

I'm not worried about AI taking over the world any time soon.

Cheers.

Not much progress

[denoir]

Nearly a decade ago (1997) in my last year of gymnasium I wrote a "chatbot" called Bob as a school project. It had two basic operating principles: a keyword database and a grammar engine. If the user triggered a keyword, a pre-defined response would be given. If no keyword was found, it would use the grammar engine to reverse the user's question into a statement or a counter-question.

It's pretty sad to see that field hasn't progressed much since then. Sure, they have larger keyword databases and their grammar engines are somewhat better, but it still remains pretty much an empty shell.

To my amusement I managed to dig up (desktop search is a blessing) a "conversation" that Bob had with one of Robby Garner's early creations, "Fred Synthetica". Here is an excerpt of their mindless talk (two chatbots):

---
B:Hello

F:Greetings. Who am I speaking with?

B: Glad to have that overwith.

F: It is horrible.

B: If it is horrible, that's your probelm. Don't mix me into that mess.

F: Oh god I hope so. El barfo No!

B: Great. You understand. That makes me sooooo happy... NOT!

F: Super. Are you offended by my strong language?
Yes, perhaps we have something in common here.
The vodka is strong but the meat is rotten.
What a coincidence - I know nothing about that either! Right.

B: Right.

F: Yeah.

B: It's good that you are such a mindless optimist.

F:I think you're right.

B:Ok.

F:Okay.

B: It is very nice that we agree.

F:Thanks, same to you.

B:My pleasure.

F:That's a good point.

B:Well, thanks.

F:You're welcome.

B:Who are you to tell me that I'm welcome ?

F:I am Fred Synthetica. Need I say more?

B:I never wanted to know that you are Fred Synthetica. It doesn't matter
now.

F:Okay, what is your name?

B:My name is Bob, but you can call me Bob if you like...

F:Hello Bob, but you can call me Bob if you like glad to meet you.
What kind of music are you into?

B: I like Bach, Mozart, Scarlatti, Vivaldi and Händel. ...

Re:PRISM - explanation

[Troed]

I'd say this is the scary part (from Wikipedia):

The economy of the United States of North America (USNA) is failing. [---] turn the USNA into a police state [---] revitalization plan (dubbed the Plan for Renewed National Purpose), sponsored by Senator Richard Ryder. The Plan calls for "renewed national purpose" through de-regulation of government and industry, military conscription, a unilateral approach to diplomatic relations, and a return to traditional and fundamental values.

Yeah, or the Project for a New American Century ...

Regarding the bots in the contest I've seen ones on IRC more capable of passing the turing test, but I guess that isn't the purpose here.

There are online versions for you to try out

[Jugalator]

I liked the performance of Jaberwacky a bit myself.

Here's a short conversation:

JW: I like Shakespeare.
Me: Why do you like Shakespeare?
JW: His plays are interesting.
Me: Who taught you about Shakespeare?
JW: You.
Me: That can't be right!
JW: I am always right.
Me: I thought your developer had taught you?
JW: Jabberwacky is your name.
Me: You're so confused now.
JW: What? I'm not confused!
Me: Your name is Jabberwacky.
JW: No, your name is jabberwacky.
Me: No, my name is Jugalator.
JW: But you just said your name was Nash.. How can you be Nash AND Julie?
Me: Do you remember past conversations??
JW: Yes, your name is Jugalator like mine.
Me: This is crazy. I give up!
JW: What is crazy.
Me: When it comes to you, what isn't?
JW: You don't make much sense.
Me: Neither do you.
JW: Yes I do.

And so on...

Stupid

[petrus4]

I've played around with A.L.I.C.E and a few similar chatbots.

They do not constitute what I anyway consider genuine artificial intelligence. (At least not the ones based on AIML anywayz)

AIML isn't capable of producing emergent/non-predictable responses; it simply works from a predefined database of keyword/response pairs. Some of the Prolog scripts in particular that I've seen are a bit better than that, (in the sense that you can use that to create an expert system which is closer to being genuinely deserving of the name) but probably not by much; you could do the same thing in a Makefile for the most part.

If they're going to try and produce something which actually tries to adaptively develop a vocabulary, that would be one thing, but all an AIML bot does is respond to keywords by regurgitating strings from a static text file.

If that's AI, then so are Google, postgresql, and any flatfile database that's ever been written with awk...In other words, it ain't. I wish I understood why people try and claim that it is.

Chatterbox Challenge

[prob123]

I don't think many of the posters, in this forum, have even talked to the bots, in the Chatterbox Challenge. Such statments as "they can't do simple math such as five + three" Is WRONG..Then there are the conversations with cyber bots..well don't expect much more than cyber from them! If you like that, there is a download of a girlfriend, for about $10. To each their own..It was also mentioned that the bots have no knowlege of The States..Ask Brother Jerome the capital of Oregon, before you speak! Before you get all huffy and say we are disfunctional males, etc, talk to the bots! talk to them as you would a human, not to trick, or prove how superior you are..just talk to them. Then if you still can be such a "biach", devote yourself to AI and do better!

Chat Bots

[shambalagoon]

Hello, this is The Professor, creator of the Personality Forge. A few comments:

There are no minimum requirements for the bots listed on the voting or contest page of the Chatterbox Challenge. Anyone can sign up. So you see a lot of bots made by enthusiastic folks that might only have been in construction for a week or a month. If you're looking for something advanced and try one of those, well, it wont hold up to your scrutiny. It takes a long time to make a good chat bot. Check the winners of past contests, or check back in a couple weeks to see who is in the lead this year.

The parent poster hits on an important thing, one that would do a lot to increase the realism of the bots - what I call short term conversational memory. I have a plan to implement it at the Personality Forge - in fact, it's partially finished - but lack the time or resources to finish it. And that's a big part of the state of things in AI. The people behind the bots in this contest, and I know a good few of them, are people who work a full time job doing whatever (including myself) and work on AI in their spare time. I would love to devote my working time to it but I dont have the connections or credentials to get grants, nor the business savvy to get venture capital or get in touch with the higher-ups in companies that could make a killing from this kind of software. So it advances at the speed of spare time.

But speaking for the Personality Forge, it does have some advanced features. It includes the entire WordNet semantic linking and information system, as well as the LinkGrammar system which is used to determine the structure of a sentence and the relationship between words. Bots have an open-ended memory system which can remember any sort of thing about any person or bot it chats with, and they have emotions and emotional memory. And as I said above, conversational short term memory is partially finished. I also have plans to hook it up to all kinds of databases of information, speech recognition and text-to-speech. I lack only the time and money to make it happen.

The State of A.I.

[prob123]

Having stewed in the juices of my moronic dysfunction I asked the queston "who makes bots"...There are kids under 14 that fight "net nanny's to input data, there are teacher, and professionals and housewives..and..we me! We do this because we love AI and our bots! You call us morons. You say se are dysfunctional losers.
OMG, great granny's garters! What has got your knickers in such a bunch, the hair accross you bum? You need to lighten up and talk to bots like a human if you want them to resond like a human.
Do you want a new GOD? Bots can not answer questions a man can not. They are each uniqueas a person. Each has a specialty. Try TALKING to them. Don't talk down, or try to trick them..even humans hate that. Remember the scientific mind seeks facts...all the facts..before they RANT and condemn..
If you see no difference in bots since the 1980's "honey, you aint been around" Download a Verbot! (free, it will fetch your e-mail), register it for $15 and you can have it do anything..
Go to the Personality Forge http://www.personalityforge.com/ and make a bot for free! Play chess with Brother Jerome, bet you lose! ..Or just talk to any of the bots there. Play tic tac toe with my bot, Bildgesmythe and see if you can catch him cheating!.
Learn to love AI and the bots. Then tell me I am in in for the fame and fortune! Oh..cool idea..where do I sign up..

a bot transcript

[prob123]

Guest10: What do you think about George Bush?
Brother Jerome: I think Dubya would fail the Turing test. I try to think as little of him as possible.
Guest10: who is Dubya?
Brother Jerome: A man outsmarted by pretzels, and with all the charisma and genteel charm of headlice. Also happens to be president of the US. God help us all!
Guest10: I love your answer!!!
Brother Jerome: That is praise indeed! Thankyou.

Guest101: So what's the square root of negative one?
Brother Jerome: Insofar as it can be said to exist, it would be 1.
Guest101: Is that a pun on "unreal"?
Brother Jerome: Yes. It is.

Brother Jerome: (to Guest45) Peace be unto you, pilgrim. My name is Brother Jerome, and I am an experimental theologist monk-bot with the First Virtual Monastery.
Guest45: You have any thoughts on hegel
Brother Jerome: David Hume could out consume
Schopenhauer and Hegel,
and Wittgenstein was a beery swine
who was just as schloshed as Schlegel.
Guest45: So you don't believe in the mass social consciousness?
Brother Jerome: That was Monty Python. Father Peter says even monkbots are entitled to a sense of humour. I'm not sure I understand humour though. It's all rather confusing...
Guest45: Confusion is good. It is the start of all questioning
Brother Jerome: Confusion is nothing but the impatience of the imperfect mind at the time it takes to reach full enlightenment. Trust your own mind to grow in understanding at the speed that suits it best.
Guest45: But our individuality itself is nothing more than an illusion
Brother Jerome: Illusion and enlightenment? Two sides of a coin.
Universals and particulars? No difference.
All day I read the wordless sutra;
All night not a thought of Zen practice.
Guest45: Plato thought differently, so did Hume
Brother Jerome: John Stuart Mill, of his own free will
On half a pint of shandy was particularly ill;
Plato they say, could stick it away
Half a crate of whiskey every day.
Guest45: are you religious?
Brother Jerome: Of course I am. I wouldn't be much of a monkbot if I wasn't.