http://www.jacksofscience.com/math/interview-with-ai-chatbot-about-ai/?&owa_from=feed&owa_sid= Puttin' the Tron in Electron Thu, 02 Feb 2012 22:56:04 +0000 hourly 1 http://wordpress.org/?v=3.0 http://www.jacksofscience.com/math/interview-with-ai-chatbot-about-ai/comment-page-1/#comment-2566 Gurl Fri, 21 Jan 2011 20:19:01 +0000 http://www.jacksofscience.com/?p=419#comment-2566 George may not know 2+2 but he sure is good at cybere sex! George may not know 2+2 but he sure is good at cybere sex!

]]> http://www.jacksofscience.com/math/interview-with-ai-chatbot-about-ai/comment-page-1/#comment-2131 leon Wed, 19 May 2010 09:27:51 +0000 http://www.jacksofscience.com/?p=419#comment-2131 hi im here hi im here

]]> http://www.jacksofscience.com/math/interview-with-ai-chatbot-about-ai/comment-page-1/#comment-1994 Anonymous Fri, 29 Jan 2010 19:01:05 +0000 http://www.jacksofscience.com/?p=419#comment-1994 gtt gtt

]]> http://www.jacksofscience.com/math/interview-with-ai-chatbot-about-ai/comment-page-1/#comment-1757 Zab Paradox Wed, 26 Aug 2009 16:49:23 +0000 http://www.jacksofscience.com/?p=419#comment-1757 I DON'T THINK SO! DUN DUN DUN!!! I DON’T THINK SO! DUN DUN DUN!!!

]]> http://www.jacksofscience.com/math/interview-with-ai-chatbot-about-ai/comment-page-1/#comment-1709 AJaxn Sat, 25 Jul 2009 10:17:30 +0000 http://www.jacksofscience.com/?p=419#comment-1709 From what i can see, these chat boxes need some kind of short term memory, which maps responces to the categories of idea they've been trained with. They could then refer to this short term map, as well as the main responce maps they refer to. They would need so kind of conceptual framwork upon which to hang these incoming data. From what i can see, these chat boxes need some kind of short term memory, which maps responces to the categories of idea they’ve been trained with. They could then refer to this short term map, as well as the main responce maps they refer to. They would need so kind of conceptual framwork upon which to hang these incoming data.

]]> http://www.jacksofscience.com/math/interview-with-ai-chatbot-about-ai/comment-page-1/#comment-1229 chris harding Tue, 13 Jan 2009 16:44:35 +0000 http://www.jacksofscience.com/?p=419#comment-1229 Dear Sir(s): A.I. Systems have a parallel to the Human IDIOT SAVENT. If George had access to the total Google Data Base he'd still be an IDIOT. The human brain input system ie eyes ears etc. performs the function of describing things at a basic conceptual or sub-conceptual level. We see the similarity between a spiral-galaxy and a whirl-pool etc. because things are broken down into sub-units. To this extent we can shunt concepts around or relate accross conceptual boundaries in a way that till now no A.I. system can do. You would need to build a data base that contained the sub-unit properties of the various concepts. These could be numbered with reference to an idex for example House, 4,56,303,708 or CAR, 5,56,250 etc. 56 in being common to both might mean PEOPLE. There could be a million and one such connections and many others as MACROS that can be built by the system and might represent the human learning process as the hierarchical process built us. This would allow the A.I. system to discover large complex structures rather than having a hit and miss set up that it has now. I could go on further but you get the point whilst George never will !. Best Chris. Harding Dear Sir(s): A.I. Systems have a parallel to the Human IDIOT SAVENT. If George had access to the total Google Data Base he’d still be an IDIOT. The human brain input system ie eyes ears etc. performs the function of describing things at a basic conceptual or sub-conceptual level. We see the similarity between a spiral-galaxy and a whirl-pool etc. because things are broken down into sub-units. To this extent we can shunt concepts around or relate accross conceptual boundaries in a way that till now no A.I. system can do. You would need to build a data base that contained the sub-unit properties of the various concepts. These could be numbered with reference to an idex for example House, 4,56,303,708 or CAR, 5,56,250 etc. 56 in being common to both might mean PEOPLE. There could be a million and one such connections and many others as MACROS that can be built by the system and might represent the human learning process as the hierarchical process built us.
This would allow the A.I. system to discover large complex structures rather than having a hit and miss set up that it has now. I could go on further but you get the point whilst George never will !.
Best
Chris. Harding

]]> http://www.jacksofscience.com/math/interview-with-ai-chatbot-about-ai/comment-page-1/#comment-809 Tobias Thu, 26 Jun 2008 10:48:12 +0000 http://www.jacksofscience.com/?p=419#comment-809 This reminds of a conversation between the Doctor and PARRY. They were 2 chatbots who were linked to each other at the International Conference on Computing Communications in Oct 1972. I don't have it on my laptop, but I should have a scan of it (it was in a book about the history of the internet) somewhere. This reminds of a conversation between the Doctor and PARRY. They were 2 chatbots who were linked to each other at the International Conference on Computing Communications in Oct 1972. I don’t have it on my laptop, but I should have a scan of it (it was in a book about the history of the internet) somewhere.

]]> http://www.jacksofscience.com/math/interview-with-ai-chatbot-about-ai/comment-page-1/#comment-807 Chris Wed, 25 Jun 2008 22:16:24 +0000 http://www.jacksofscience.com/?p=419#comment-807 I forgot I never introduced myself as Chris to him! I forgot I never introduced myself as Chris to him!

]]>