Re: How to make a slave (was: Building a friendly AI)

From: M T (
Date: Mon Dec 03 2007 - 10:57:31 MST

>> Do you understand what your computer does by
>> anthropomorphizing it?
>Sometimes a little bit, the amount of understanding you will obtain by
>analyzing the trillions and trillions of lines of computer code that
>makes Mr. Jupiter Brain tick is zero.
>John K Clark


I was tempted to write before, when you first made this statement about Dr. Jupiter Brain but refrained.

But, "Zero"?

Sensationalist statements like that give me chills.
You know what's wrong with the "zero", so I won't go into it.

But while I'm at it, I have to say that the scenario that you put forward is weak and seems to be created to suit your argument.
Specifically, by the time there is a Jupiter Brain, it's safe to assume that there will be ample humans with substantially improved hardware. Or minor AGIs with an easy enough architecture for unimproved humans to understand, which would help humans in turn to understand the architecture of the Jupiter Brain. And there are many more plausible scenarios that would facilitate understanding.

But you decide to jump way ahead in the future to make an argument seem plausible, as if a Jupiter Brain will explode tomorrow out of Google's research.

If you had put a full-stop after "Sometimes a little bit", your argument would have been stronger, in my humble opinion.

Kind regards,


Support the World Aids Awareness campaign this month with Yahoo! For Good

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:01 MDT