[ExI] Science or Scientism?

[Stuart LaForge]

John Clark wrote:

> If you don't use my axiom the phantom limb syndrome tell you nothing
> about consciousness, all it tells you is if you cut off somebody's toe
> sometimes they make noises with their mouth that sounds like "I don't
> have a toe but it hunts anyway".

I have twice proven the undecidability of consciousness on this list,
first as a consequence of Russell's Paradox and then later as a corollary
to Rice's Theorem so I am halfway on board here.

The problem is that unless ALL Turing machines are intelligent or NO
Turing machines are intelligent, then intelligence is undecidable in
Turing machines. In other words intelligence is either trivial property or
undecidable as well.

So your axiom would require two undecidable properties to be correlated.
Which they certainly appear to be experientially. We certainly use the
correlation to infer consciousness in our pets for example.

The caution flags go up, however, because the combination of these two
undecidable properties is undecidable also. Your axiom is very likely only
a heuristic that would work on roughly human scales +/- 3 orders of
magnitude or so.

This is because differences in scale limit communication and for large
differences limit even perception. You can't infer the intelligence of
something you can't see for example. And even if the galaxy was
axiomatically intelligent,  would take millions of years for you to notice
any "intelligent behavior". Which is time most humans don't have. Time
passes very differently for brains at different size scales.

Therefore your axiom will be subject to false positives, where you see one
off illusory patterns in random phenomena like faces in the clouds that
you think imply intelligence.

And you will also have false negatives where you underestimate or fail to
notice the intelligence of beings very much larger or smaller than you are
or equivalently beings very much faster or slower than you.

But all in all, I find your axiom a useful heuristic so long as you keep
in mind its limitations and ultimate undecidability.

Stuart LaForge