Comments on: The Current State of Affairs in Pursuit of a Commonsense AI http://texai.org/blog/2008/01/11/the-current-state-of-affairs-in-pursuit-of-a-commonsense-ai/ An open source project to create artificial intelligence Thu, 20 Nov 2008 22:04:42 +0000 http://wordpress.org/?v=2.1.2 By: Matt Revelle http://texai.org/blog/2008/01/11/the-current-state-of-affairs-in-pursuit-of-a-commonsense-ai/#comment-2007 Matt Revelle Fri, 11 Apr 2008 15:52:41 +0000 http://texai.org/blog/2008/01/11/the-current-state-of-affairs-in-pursuit-of-a-commonsense-ai/#comment-2007 Do you believe that artificial intelligence will be formed without modeling some equivalent of the neocortex? Recognition and memorization of patterns and pattern hierarchies seems like a requirement for true intelligence. Do you believe that artificial intelligence will be formed without modeling some equivalent of the neocortex?

Recognition and memorization of patterns and pattern hierarchies seems like a requirement for true intelligence.

]]> By: Steve Reed http://texai.org/blog/2008/01/11/the-current-state-of-affairs-in-pursuit-of-a-commonsense-ai/#comment-2033 Steve Reed Mon, 14 Apr 2008 14:19:06 +0000 http://texai.org/blog/2008/01/11/the-current-state-of-affairs-in-pursuit-of-a-commonsense-ai/#comment-2033 Yes I believe that the Texai project will create a cognitively plausible artificial intelligence. Furthermore, a multitude of Texai instances acting in concert will behave in a plausible manner with respect to a large human organization. However, I am not directly modeling human cognitive structures (e.g. the neocortex) to create Texai capabilities. Rather, I treat cognitive plausibility as a set of capabilities and constraints into which the Texai implementation must fit. For example, I am using incremental processing for both utterance comprehension and for generation. This algorithm is derived from the work of Walter Kintsch, a cognitive psychologist, but it is a symbolic approach, not connectionist. -Steve Yes I believe that the Texai project will create a cognitively plausible artificial intelligence. Furthermore, a multitude of Texai instances acting in concert will behave in a plausible manner with respect to a large human organization.

However, I am not directly modeling human cognitive structures (e.g. the neocortex) to create Texai capabilities. Rather, I treat cognitive plausibility as a set of capabilities and constraints into which the Texai implementation must fit. For example, I am using incremental processing for both utterance comprehension and for generation. This algorithm is derived from the work of Walter Kintsch, a cognitive psychologist, but it is a symbolic approach, not connectionist.
-Steve

]]>