On Thu, 13 Feb 2003, Alan Grimes wrote: > The [SIAI] idea of goals is not so abhorant but still has the > fundamental flaw of not being applicable to general intelligence in any > meaningful way. The entire approach that it suggests is that it focuses > only on one tiny subsystem of human cognitive function and neglects the > vast aparatus which orchistrates our behavior. I have been thinking > about writing a post to discuss and promote an engineering approach to > AGI... =\
In disagreement with this point in particular and with your overall dissatisfaction with the SIAI Friendly AI doctrine, I immediately saw the wisdom and applicability of a deliberative causal goal system as the principle for ensuring that Cyc becomes a safe AGI - should we first achieve it. Furthermore, following a tip from Ben with regard to the NIST/RCS hierarchical control structure developed by James Albus, I immediately saw the framework upon which to lay Friendly AI principles. Cycorp has an engineering approach to creating AI. We have made common the term Ontological Engineering for example. Although Brain Science may converge with AI and each field will contribute to understanding the other, Cyc is not at all engineered as a model of human cognitive processes, rather we proceed from symbolic logic to develop an ambitious vocabulary to deductively reason about the full range of human concepts, and hopefully from there to use the vocabulary to manifest cognitive behavior. My views are my own and do not necessarily reflect the views and policy of Cycorp. -Steve -- =========================================================== Stephen L. Reed phone: 512.342.4036 Cycorp, Suite 100 fax: 512.342.4040 3721 Executive Center Drive email: [EMAIL PROTECTED] Austin, TX 78731 web: http://www.cyc.com download OpenCyc at http://www.opencyc.org =========================================================== ------- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED]