Re: "Morality?" - Composite Reply

Eliezer S. Yudkowsky (
Wed, 08 Oct 1997 02:32:30 -0500

Lee Daniel Crocker wrote:
> > Science is the middle path between the harsh tyranny of unquestionable "truth"
> > and the useless undisprovability of "subjective" reality. Science states that
> > there exists an objective reality, but we do not know what it is. My ethical
> > system makes the same statement with respect to morality.
> A noble goal to be sure, but by what method can you test this morality
> for its correspondence to the objective standard? I know how to
> evaluate statements about reality: they must be meaningful, falsifiable
> by observation, and must have wihstood many observations without such
> a falsification. The standard of judgment is simple: my senses.

"Externalism" (temporary label) is not a science, and will remain unscientific
- like all ethical systems - until an experimental method is invented.
Externalism is not scientific, but it is loosely analogous to science. In
most ethical systems, philosophical thought takes the place of
experimentation. Thus they fail. Philosophy isn't that powerful. In my
version of Externalism, intelligence enhancement acts as an interim substitute
for experimentation. Intelligence enhancement has the potential to be fully
as powerful as experimentation and perhaps more so.

But that isn't the reason. The reason is that I'm too dumb to come up with an
experimental method, and I have to recurse on the problem of becoming smarter.

I realize that "thought without observation is as bad as observation without
thought". I reply that "intelligence enhancement without experimentation is
as bad as experimentation without intelligence enhancement". One can be an
Externalist and a scientist both.

Scientific technology is the key to enhancing intelligence. Intelligence
enhancement is the key to inventing experiments. Which method one should use
at a given time depends on whether experimentation or IA is easier; ideally,
one should use both. In the domain of "ethics", with current technology, IA
is difficult but experimentation is impossible.

--       Eliezer S. Yudkowsky

Disclaimer:  Unless otherwise specified, I'm not telling you
everything I think I know.