Since the day I had chat with Cleverbot, I started to love Chatbots too. I agree about the programming part, it's not going to be easy. I wish good luck with your Chatbot development. :)

Cleverbot really does not know the current issues, not even the latest blockbluster movie that hits the world. I talked about the movie Avatar and it kept talking about the children TV Show. LOL.

Thank you for your thought about the reasoning and Neural Networks. I believe those two elements are the key to advancements in strong AI, like you have said.

Programming "Common Sense?" That sounds great but yeah, it's pain. There are even humans not having sufficient common sense. :P

Thanks for your comment, Michangelina.

I haven’t chatted with a chatbot in a while, Cleverbot is more interesting than most. It tends to ask more questions than most normal bots, has a sense of humor, and it can make typing errors.

Unfortunately, if you address issues that are going on in the world today. For instance, the catastrophe in Haiti. The bot will have no knowledge of it. The designer would have to constantly upgrade the chatbot’s software in order for it to discuss all sorts of topics and ideas. Yet, can a bot have an opinion? The answer is “not really”.

However, it is definitely a good example of understanding language and comprehension. Reasoning is still a problem, which I think cannot be accomplished without neural networks. Neural networks are the key to advancements in strong AI. Pattern recognition is just computer science running on automatic responses.

I don’t think a bot will be able to pass a Turing Test for a while… At least not until the Singularity happens (Kurzweil). Minsky is trying to do something with his bot that functions like a three year old… something like that… programming “common sense” is a pain. >.<

Thanks for sharing. :)

Thank you for stopping by and for the performance report. I had no reports of recent Loebner Prizes. After reading your comment, I actually tried to find out more about it. I found that Jabberwacky was not really getting better compared to the competitors after 2006. It did not achieve the required standard of Turing Test achieved by those competitors.

I'm not well knowledged about Chatterbots or AI and so, most of the statements I made came from different sources. After you said, I found that variety of judges was actually used for 2008.

Thank you for posting and leaving the informative link.

I have edited the article and credited you. :)

Also, your statement, “The decision are based on comments by several users consisiting of experts, non-experts, children, adults, native and non-native English speakers” does not cite a source. Only in the 18th Loebner Prize were all those category judges used:

http://www.reading.ac.uk/research/Highlights-News/featuresnews/res-featureloebner.aspx

You may wish to read Shah & Warwick’s Analysis of Jabberwacky’s Performance, during a Chatterbox Challenge and a Loebner Prize for Artificial Intelligence: “Constraining Random Dialogue in a Modern Eliza” in Schmidt, C.T.A (Ed). Essays on Computer and Philosophy: From Man Through To Artefacts To Man, AAAI/MIT.

Here’s Jabberwacky’s laconic response to Question 9 “how many is a million?” in the knowledge phase of Chatterbox Challenge 2005: “More than 12″! :-)