On 07-Feb-11, at 12:30 PM, Anand Manikutty wrote:
I have discussed this issue with the people at the Singularity meetup. None of them is able to tell me how exactly AI could engage in "continuous self-improvement", an idea which lies at the heart of the Singularity argument. I believe that "continuous self- improvement" is a purely hypothetical possibility.
At its most basic, a GA requires some variables and a fitness function - http://megaswf.com/serve/102223/
-Taj.
