> you can't address [the existence of God] scientifically
Maybe one day science and philosophy will meld to the point where this might be possible in a peer reviewed in the Nature journal kind-of possible way, God cannot be addressed by the current scientific method.
What's valuable about Paul Graham's line-of-thinking is that he can distinguish between science and religion - something that very few can do (and is likely going to make this HN post explode into a pointless evolution vs. creation argument - which is missing the point of the experiment entirely).
> Maybe one day science and philosophy will meld to the point where this might be possible in a peer reviewed in the Nature journal kind-of possible way ...
No, that's not possible, because of two essential properties of science that distinguish it from philosophy -- empirical evidence, and falsifability. The first (empirical evidence) means science is steered by evidence drawn from everyday reality in practical experiments or observations. The second (falsifiability) means it must be possible to test scientific ideas against reality, and those ideas that fail the test must be discarded.
These properties of science mean that science and philosophy will never merge.
> God cannot be addressed by the current scientific method.
And the "current scientific method" is how science is defined. The only way to get around that definition is to abandon science.
