Friday, December 03, 2004


So, in the last year that I haven't been accomplishing much on Seductotron, I've been watching my niece grow through her first year. It's really interesting watching a human learn how to interact with their environment, and I feel that it's given me some critical insights into what must be done in order for a project such as the one I envision to get off the ground.

The problem of a self-programming program isn't completely new, and I've encountered is as a paradox before. I wanna' say that it was Ashby? Anyhow, that's not really important. I'm impervious to knowledge. The point is, it seems a paradox. You can't write a meta-program that writes a sub-program without imposing the constraints of a program upon the meta-program that writes the sub-program in the first place. So, it seems that we hit a cognitive limit when we try to envision a truly self-educating program. You can divorce the program from the determinative outcome of a defined data set, but you can't divorce it from the meta-determined outcome of a defined procedure.

Anyhow, all that is really just an exposure of my absolute ignorance. Most of my education was a tragically wasted endeavor. Naturally, the following is merely a travesty of it.

The point for me, is that if it seems a paradox then it stands to reason that I've hit a cognitive limit. However, it does not follow that I have therefore hit a functional limit. In fact, it's been my experience that almost all paradoxes are attributable to a flaw in the logical process itself - generally from framing a statement in such a way that not all of the necessary conditions can be reconciled adequately to one another. Anyhow, that's a caveat I can get into in my drooling-retarded fashion, but let's just stick to this particular roadblock and my attempt to sidestep it without much consideration: