[extropy-chat] proto-extropian religions

Eliezer Yudkowsky sentience at pobox.com
Tue Oct 12 20:36:57 UTC 2004


Bryan Moss wrote:
> 
> An interesting question might be: In lieu of actual knowledge, is it 
> better to have unjustified true belief or justified untrue belief (given 
> a pop philosophical notion of justification [and knowledge])?  I.e., who 
> would you prefer, someone who believes evolution to be true (a true 
> belief) but knows none of the arguments and none of the evidence, just 
> the raw fact "evolution is true," or someone who believes creationism is 
> true (an untrue belief) but has read the Bible from cover to cover and 
> has a bookshelf full of books on creation science, intelligent design, etc?

I would account this as negative knowledge or anti-knowledge, making it 
more difficult to absorb the real details when it comes time to learn them. 
  I prefer the blank slate, not because blind belief in evolution counts 
for much, but because it will be easier for the one to learn real science 
later.

In Bayesian terms, confessed ignorance is better than a strong confidence 
in an incorrect prediction.  It *is* possible to do worse than a hypothesis 
of maximum entropy, though only humans and contemporary AI programs perform 
so poorly that you can increase performance by injecting noise into the 
algorithms.

> I submit that as long as the proles believe that evolution is true, the 
> details don't matter.

Proles?

I suppose that as long as you believe that, it isn't worth my time to try 
and explain my own lofty position to you.

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



More information about the extropy-chat mailing list