Eliezer on Science versus Religion

From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Mon Mar 05 2001 - 00:54:30 MST

Damien Broderick wrote:
> There might be
> certain kinds of truths, even the most important, which are inaccessible to
> scientific approaches, let alone to falsification tests.

If so, then they are inaccessible to the people making the claim "There
are untestable truths". As someone pointed out, none of the Books show
prophets talking about untestability. Elijah confronting the priests of
Baal would have understood perfectly the put-up-or-shut-up attitude of
modern-day science. Now this may be because Elijah is a historical
character with superpowers who thus didn't lack for self-confidence, or it
may be because he's a fictional character who was invented before the
defeatism that permeates modern religion. Either way, the concept of
untestability and scientific inaccessibility is clearly a modern
innovation created by a bunch of theologians who, not to put too fine a
point on it, insisted that the Earth was flat and six thousand years old
and got their butts kicked (by Truth, not Science) and have been running
scared ever since.

So now we have theologians who not only make excuses for past failures,
but make up further excuses in advance. Present them with any scenario in
which their beliefs are tested, and they'll immediately start making up
excuses in advance for why they'll fail. The thought that they might
succeed doesn't even occur to them - defeatism, pure and simple. I'm sure
the excuses they're making up in advance conform exactly to the ideals of
religious faith, but whatever part of their brain is deciding which
results they'll probably need to make excuses *for* is so perfectly
atheistic that not even Daniel Dannett could disapprove.

Modern theology has suffered a total failure of confidence. The Books
demonstrate very clearly that religions are supposed to be testable.
Humans innately expect people with religious power to predict the future
and heal the sick. Modern theologians can't heal diddly, which was fine
back when nobody else could either, but now that technology has come along
they're scared sick that somebody's going to notice the difference between
making big claims and delivering on them. But that is simply not
science's problem. It gets solved either when another prophet comes along
who can actually heal the sick, or when religion finally folds in and
admits it was a lie all along. Either way, theology's failure to live up
to its own expectations has nothing to do with science.

Damien Broderick wrote:
> There might be
> certain kinds of truths, even the most important, which are inaccessible to
> scientific approaches, let alone to falsification tests.

I am not playing this game. Every time I say "Well, if we have immortal
souls, then they'll show up in a nondestructive brain scan", the other
person, whose actual Outcome Predictors were coded by rabid atheists, says
"But what if it *looks* like everything is totally understandable, and the
upload doesn't notice any difference, and nobody else can tell the
difference, but the soul is still missing?" and wants to know what I'll do
about it.

Draw a line in the sand, theologians! Pick a place where you stop
retreating! Put up or shut up! Say that the upload will die on the
spot. Say that he'll turn into a remorseless killer. Say that he'll lose
all his artistic creativity. Say that we'll find the neurons are doing
something incomprehensible. Say that no amount of experimentation will
ever explain it enough to correctly predict a single neuron. I can plan
for the possibility of a religious universe if you'll tell me how it looks
different from a totally materialist universe, but as long as you go on
using rabidly atheistic Outcome Predictors, then I'm sorry, but there's
nothing I can do for you.

-- -- -- -- --
Eliezer S. Yudkowsky http://singinst.org/
Research Fellow, Singularity Institute for Artificial Intelligence

This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:59:39 MDT