What about if you are not interested in developing new ones, just studying other peoples mistakes?

You''re better off studying the failures and mistakes made in projects which were altogether more successful and comprised of more substance and hard work.

Here''s a suggestion: study the mistakes made in SHRDLU, SAM, PAM, MARGIE, CD, POLITICS, SWALE, and others which advanced the state of the art. The links for some of those are in my signature.

_______________________________

"To understand the horse you''ll find that you''re going to be working on yourself. The horse will give you the answers and he will question you to see if you are sure or not."
- Ray Hunt, in Think Harmony With Horses

ALU - SHRDLU - WORDNET - CYC - SWALE - AM - CD - J.M. - K.S. | LANL | CAA - BCHA - AQHA - APHA - R.H. - T.D. | 395 - SPS - GORDIE - SCMA - R.M. - G.R. - V.C. - C.F.

Actually, I don''t care how they work, I just like talking to them because I have no friends... ;-)

so far checked only SRDLU
and that''s super !
I wanted to implement smth similar. the real challenge is making it understand language constructs by ITSELF. This info can give some ideas.

i once got alice to sleep withvme!

quote:

---

Original post by bishop_pass
Chatterbots are stupid. A true step backwards within the field of AI.

---

Academic AI researchers roundly lambast chatterbots and their designers every chance they get..... it''s true that chatterbot design is more a matter of psychological and sociological cleverness than an application of AI research. Yet I suspect that as AI approaches a serviceable enough level to interact with humans natively, researchers will find that the chatterbot hackers have been tunneling in from the other end, discovering tricks and techniques that will be useful in "polishing" AI interfaces.

---

How appropriate. You fight like a cow.

quote:

---

Academic AI researchers roundly lambast chatterbots and their designers every chance they get..... it''s true that chatterbot design is more a matter of psychological and sociological cleverness than an application of AI research. Yet I suspect that as AI approaches a serviceable enough level to interact with humans natively, researchers will find that the chatterbot hackers have been tunneling in from the other end, discovering tricks and techniques that will be useful in "polishing" AI interfaces.

---

How appropriate. You fight like a cow.

---

What is that ? What are you rambling about ? Your post is as useful as plain 0

quote:

---

Original post by Sneftel

---

Original post by bishop_pass
Chatterbots are stupid. A true step backwards within the field of AI.

---

Academic AI researchers roundly lambast chatterbots and their designers every chance they get..... it's true that chatterbot design is more a matter of psychological and sociological cleverness than an application of AI research. Yet I suspect that as AI approaches a serviceable enough level to interact with humans natively, researchers will find that the chatterbot hackers have been tunneling in from the other end, discovering tricks and techniques that will be useful in "polishing" AI interfaces.[...]

---

AI can pretty much already interact via normal english via text, its been done in several cases. There programs that can do as you command them and then answer questions about why they did things. There are also programs that can read in english text, summarize it, and make inferences based on the data.
Check out Bishop Pass's sig for examples of both and more.

[edited by - Extrarius on May 7, 2003 3:38:06 PM]

quote:

---

Original post by thy MC
What is that ? What are you rambling about ? Your post is as useful as plain 0

---

Is there anything in particular you didn''t understand, or did the whole thing go over your head?

---

How appropriate. You fight like a cow.

quote:

---

Original post by Extrarius
AI can pretty much already interact via normal english via text, its been done in several cases. There programs that can do as you command them and then answer questions about why they did things.

---

You sound like you''re mostly talking about SHRDLU. But that''s not the last mile. It''s true that you can have a conversation with an AI if you know how that AI is programmed and what syntactical constructs it supports. It''s quite a different thing to parse arbitrary natural language, with typos, minor grammar errors, and unfamiliar constructs.

---

How appropriate. You fight like a cow.