[NetBehaviour] New developments - On Being/"exist.pl"

[Rob Myers]

On Fri, Jul 25, 2008 at 10:19 AM, Pall Thayer <pallthay at gmail.com> wrote:
>
>>> Is it worth adding a simple neural net so the program can draw its own
>>> conclusions? :-)
>
> But see, that's where things begin to get really hairy. Would such
> conclusions really be the program's conclusions? Wouldn't it be more
> accurate to say that they were my conclusions?

We do not create the structure of our own brains, we receive their
design via evolution (or from God, but either way we don't make them
ourselves). But we eventually take credit for using them.

The structure of a neural net isn't determined by the program itself
either. *Legally* the program's conclusions would be yours, I think.
But *philosophically* is there a reason other than the simplicity of
the program that means credit for its discoveries should go to the
author instead?

AI programs are texts, they are scores. They are more like the writing
games of the Oulipo or the Surrealists or the Beats than a simpler
static text. If they produce "strange loops"
(http://en.wikipedia.org/wiki/Strange_loop) then this could be at
least an analogue to or metaphor for self-awareness.

- Rob.