----- Original Message ----- From: "Stephen Reed" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Wednesday, January 22, 2003 4:27 PM Subject: Re: [agi] AI and IQ
> Pei > > Thanks for the link. > > However, in my opinion this progress is misleading in that a simple > geometric algorithm is likely used to solve the highly specialized > puzzles, then the robot executes a plan to illustrate the solution. > > Because the robot software has been tailored to solve one particular > mathematical part of an IQ test, it does not really suggest anything new. Steve, I basically agree --- I posted the link because it is related to the previous discussion on how to judge/measure the intelligence of a system, not because I like this work (though I do need to know more about it to make confident comment). In general, I believe that we may be able to define an IQ test for AGI, but it should measure what the system can learn over a period, not what the system can do at a given time. For human beings, since we roughly start with the same innate knowledge, it is OK to use "what it can do" to measure "what it can learn". However, such a methodology cannot be used on AI, because, as you said, it is easy to code domain-specific knowledge into the system. As a result, "what it can do" and "what it can learn" are no longer correlated. > And regarding robots, I far prefer the work in > navigation done by Hans Moravec at CMU, the walking humanoid robots in > Japan, and the robot soccer (football) competition. > > These technologies will become excellent peripherals for AGI unless of > course the robot folks develop AGI first. I don't think the robot folks will develop AGI first. It seems that they are still mostly working on the low-level domain-specific details. Pei > -Steve > > On Wed, 22 Jan 2003, Pei Wang wrote: > > > "In the quest to create a machine that will outsmart the most intelligent > > human, Selmer Bringsjord has developed the first-known robot capable of > > passing part of a standard IQ test. " > > > > See http://www.rpi.edu/web/Campus.News/sept_02/sept_03/iq.html. > > > > > > > > ------- > > To unsubscribe, change your address, or temporarily deactivate your subscription, > > please go to http://v2.listbox.com/member/?[EMAIL PROTECTED] > > > > -- > =========================================================== > 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] > ------- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED]