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

From: John K Clark (johnkclark@fastmail.fm)
Date: Tue Nov 20 2007 - 22:56:33 MST

Next message: Lee Corbin: "Re: how to do something with really small probability?"
Previous message: Thomas McCabe: "Re: General summary of FAI theory"
In reply to: Thomas McCabe: "Re: How to make a slave (was: Building a friendly AI)"
Next in thread: Matt Mahoney: "Re: How to make a slave (was: Building a friendly AI)"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

On Tue, 20 Nov 2007 "Thomas McCabe" <pphysics141@gmail.com> said:

> Calling it "judgment" invites anthropomorphicism.

Yes, and so what? Anthropomorphic reasoning is a perfectly respectable
tool, it can be misused but so can any tool.

> Utility functions are much cleaner mathematically anyway.

And it sounds cooler and more brainy too, but I’ll bet that when you
describe your own actions you use the word “judgment” one hell of a lot
more often than “utility function”. If the word is good enough for you
it’s good enough for MR. AI, I certainly wouldn’t discriminate against
him just because he’s not made of meat.

John K Clark

-- 
  John K Clark
  johnkclark@fastmail.fm
-- 
http://www.fastmail.fm - Does exactly what it says on the tin

Next message: Lee Corbin: "Re: how to do something with really small probability?"
Previous message: Thomas McCabe: "Re: General summary of FAI theory"
In reply to: Thomas McCabe: "Re: How to make a slave (was: Building a friendly AI)"
Next in thread: Matt Mahoney: "Re: How to make a slave (was: Building a friendly AI)"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:00 MDT