AI Zone Admin Forum Add your forum

NEWS: Chatbots.org survey on 3000 US and UK consumers shows it is time for chatbot integration in customer service!read more..

Just For Fun
 
 
  [ # 181 ]

For ‘Take 2’, let’s see a simple text string, say, “hello World” translated into the relevant visemes, and cause the mouth to move accordingly. smile

 

 
  [ # 182 ]

What the A.I. community needs is a
contest for a light weight chatter robot
class, under 300 lines of code and data*:
.
.
.
.
.

http://www.8pla.net/easybot/
.
.
.
.
.
.

* Not including the data log.

 

 
  [ # 183 ]

Reference: http://www.8pla.net/easybot/

Thanks for your very kind participation!
We’ll keep the rest between us. You choose
whether or not to share your session log.

Those who tested it provided helpful feedback.
A simple memory bank was added to this light
weight chatter robot class to handle repeats.

So how does this light weight chatter robot class
with under 300 lines of code and data, compare to
the heavy weight chatter robot class with way over
50,000 lines of code and data?

 

 
  [ # 184 ]

Reference: http://www.8pla.net/easybot/

Here are the results of its first test in basic math today ...

>> How much is 987654321 - 123456789
<< 864197532

>> what is 2 * 2 * 2?
<< 4  (It only calculates two elements for now)

>> How much is 99 / 9?
<< 11

>> What is 2 * 3?
<< 6

>> What is 2*3? (The trouble is no spaces)
<< Blah, blah, blah!

As you can see, it still needs more lessons in math.  Not bad for its first lesson.

 

 
  [ # 185 ]
8PLA • NET - Aug 9, 2011:

So how does this light weight chatter robot class
with under 300 lines of code and data, compare to
the heavy weight chatter robot class with way over
50,000 lines of code and data?

>> are you alive
<< Yackety Yack.
>> what is your favourite food
<< What do you think?
>> what is a robot
<< What comes to your mind when you ask that?

Not too good to be honest.

 

 
  [ # 186 ]

When I conversed (if you can call it that) with the bot, I got the strong impression that I was interacting with Eliza’s kid sister. At only 300 lines of code, you may be doing well with that, but I think that I speak for everyone here when I say that quality is much more important than quantity (or the lack thereof). Tom, I like you a great deal, and you have my respect for many of the things you’ve presented here, but I think it’s time to show us what I’m certain that you’re capable of, and combine the best of what we’ve seen so far, expand upon it, and knock our socks off with a truly remarkable chatbot. I’m sure you’re up to the challenge; but will you accept it? raspberry

 

 
  [ # 187 ]

Reference: http://www.8pla.net/easybot/

Steve Worswick - Aug 9, 2011:

Not too good to be honest.

Honesty is good Steve.  A new bot is sort of like a baby that needs to learn from you.

It’s what makes “Prototype” my very favorite Star Trek episode, though I loved them all:

The prototype suddenly seems to come alive as it sits up and looks around as if disoriented. “Prototype unit 0001 is ready to accept programming,” it states.

Quoted from: http://memory-alpha.org/wiki/Prototype_(episode)

Here is what the baby bot learned from Uncle Steve’s feedback today ...
( Note: This was not logged by Steve.  It is a test run to try to address his feedback. )

>> Which is smaller a cat or a mouse?
<< If I guessed that a cat is smaller, would I be right or wrong?

>> wrong
<< What does that suggest to you?

>> When is New Years Day?
<< I think you know when that is. By that I mean New Years Day.

>> Who is Mitsuku?
<< I think you know who that is. By that I mean Mitsuku.

>> Are you alive?
<< Perhaps I am that in your fantasies.

>> What is a robot?
<< I think you know what that is. By that I mean a robot.

>> What is your favourite food?
<< I think you know what that is. By that I mean any favourite food.

What do you think might be a short list of top AIML for the most comprehensive
response range from the fewest amount of categories?

 

 
  [ # 188 ]
Dave Morton - Aug 10, 2011:

When I conversed (if you can call it that) with the bot, I got the strong impression that I was interacting with Eliza’s kid sister. At only 300 lines of code, you may be doing well with that, but I think that I speak for everyone here when I say that quality is much more important than quantity (or the lack thereof). Tom, I like you a great deal, and you have my respect for many of the things you’ve presented here, but I think it’s time to show us what I’m certain that you’re capable of, and combine the best of what we’ve seen so far, expand upon it, and knock our socks off with a truly remarkable chatbot. I’m sure you’re up to the challenge; but will you accept it? raspberry

Light weight chatter robot real-time learning:

http://www.8pla.net/easybot - Aug 10, 2011:

>> learn Dave is my friend
<< I may learn Dave is my friend

>> explain Dave
<< I explain Dave is my friend

>> learn Steve is smart
<< I may learn Steve is smart

>> explain Steve
<< I explain Steve is smart

>> learn A.I. means Artificial Intelligence
<< I may learn A.I. means Artificial Intelligence

>> explain A.I.
<< I explain A.I. is Artificial Intelligence

Reference: http://www.8pla.net/easybot/

 

 

 
  [ # 189 ]

Are you demonstrating my own learn.aiml to me?

 

 
  [ # 190 ]
Steve Worswick - Aug 10, 2011:

Are you demonstrating my own learn.aiml to me?

Light weight chatter robot real-time forgetting:

http://www.8pla.net/easybot - Aug 10, 2011:

>> remember Steve is asking a question
<< I may learn Steve is asking a question

>> explain Steve
<< I explain Steve is asking a question

>> forget Steve is asking a question
<< I may forget Steve is asking a question

>> explain Steve
<< I explain Steve is but I forgot

Reference: http://www.8pla.net/easybot/

 

 

 
  [ # 191 ]

CHATTER ROBOT IN LISP

For all you chatter robot masters
out there who would like to see a
chatter robot in Lisp made simple:

http://www.ChatterRobot.com/Lisp

 

 
  [ # 192 ]

Light weight chatter robot real-time recount:

http://www.8pla.net/easybot - Aug 17, 2011:

>> Hello.
<< What can I do for you?

>> Blah, blah, blah.
<< I am not sure I totally get it.

>> Yackety, yack!
<< Interesting, I didn’t totally get what it suggested to you,
  but you continued to go on because you felt like discussing this.

Reference: http://www.8pla.net/easybot/

UPDATE: Added recursion to better handle duplicate respones.

 

 
  [ # 193 ]
Steve Worswick - Aug 10, 2011:

Are you demonstrating my own learn.aiml to me?

Obviously, it is not your learn.aiml file which is 579 lines of code by itself with no A.I. engine code.  That is almost twice the size of an entire light weight chatter robot system, data and code combined.

But, thanks for an awesome compliment, Steve, which is very encouraging to all new research into light weight chatter robots written entirely in under 300 lines of code.

 

 

 
  [ # 194 ]
8PLA • NET - Aug 9, 2011:

Reference: http://www.8pla.net/easybot/

So how does this light weight chatter robot class
with under 300 lines of code and data, compare to
the heavy weight chatter robot class with way over
50,000 lines of code and data?

The beta version of Skynet-AI is under 300 lines of code (code+data).
Of course some of those lines are pretty long. wink


Perhaps a better metric would be total code size.

 

 

 

 
  [ # 195 ]

I agree that total byte count is a more accurate indicator, since it’s really easy to put javascript code all on a single line, if one so chose. I’d be willing to bet that with a little bit of planning, the same could be said for PHP.

 

‹ First  < 11 12 13 14 15 >  Last ›
13 of 18
 
  login or register to react