🚀 go-pugleaf

In article <3076@news.cerf.net> snodgras@cyber.net (John E. Snodgrass) writes:
>      Your suggestion that computer programs model themselves, in this or
>any other way, however, strikes me as farfetched. To create a coherent
>model of any type, I would suggest it is necessary to have a purpose which
>that model serves. This provides a means for prioritizing and organizing
>the data in the model. Thus without even criticizing the obvious point
>that all computer programs have a 'bottom level' which is simply a set
>of conditionals relating input to output, no computer program could have
>a purpose with respect to _itself_, which would (by my definition) exclude
>any possibility of consciousness.

I see no reason to believe that "purpose" is such a fundamental concept.  I
see it as an emergent concept in complex systems.