Re: - E04 - Leadership! Google, Guido van Rossum, PSF

"Anton Vredegoor" <anton.vredegoor@xxxxxxxxx> writes:
> Ok. That's a bit harder. I suppose we agree that if we have an
> intelligent program that is more intelligent than a human and have this
> program design an even more intelligent program than things start to
> accelerate pretty fast?

There's your fundamental flaw. Programs aren't intelligent, any more
than submarines swim.

> Now the combination of a programmer with a tool (program) that can be
> used to make a better tool. This gives a better human-machine
> combination, which then can be used to further improve the combination.

Key word: "tool". Nothing is said about the intelligence of said
tool. It doesn't have to be intelligent at all. It just has to be able
to do one single task better than the human. The intelligence in this
combination is coming from the human.

> I don't think I have completely proven my point now, but since the
> danger is very real and big, coming close is already reason enough to
> watch this carefully.

We aren't any closer to having a "real" AI than we were in the 60s. On
my list of man-made, world-ending catastrophes, a world-dominating AI
is easily outranked by grey goo, and just barely beats out the nine
billion names of god.

> 'sorry, I don't want to start a flamewar, but I really believe what I
> wrote here'

I think you're taking to many bad scifi movies seriously.

<mike
--
Mike Meyer <mwm@xxxxxxxxx> http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.
.