Re: screwing the inscrutable

From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Fri Jul 13 2001 - 10:19:56 MDT


Eugene Leitl wrote:
>
> On Fri, 13 Jul 2001, Damien Broderick wrote:
>
> > her mother confirmed recently. No earthquake, no hands Ma. I dunno. If that
> > *is* the world we live in, things are not so simple. The Matrix or
> > something like it could still be Ockham's cut.
>
> Hmm, perhaps there's value in most of civilized world going Big Brother.
> If your theory is true, blanket surveillance is sooner or later certain to
> pick up something strange.

If the theory was true, "something strange" would have already been
recorded by mall surveillance cameras, or sheer coincidence would have
generated some kind of psychic phenomenon during a press conference. Or
is there a reason why the flaws in the Matrix only appear to groups of a
few believers and only when the video cameras are turned off?

Using my psi-is-bunk hypothesis, I predict that transparency will fail to
capture any psychic phenomena, that the advocates will go on believing
anyway, and that what they cite as evidence will shift even further into
the realm of alleged past victories and present reports of unverifiable
subjective experiences.

As communications technology advances, the impressiveness of reported
psychic phenomena decreases in strict lockstep. Where past psychics could
apparently perform huge miracles on demand in front of crowds of
thousands, present-day psychics are apparently limited to small unreliable
phenomena performed for tiny groups of believers when the video cameras
are turned off.

I'll formalize the statement: "If alleged reports of a phenomenon grow
steadily less impressive with proximity to the present day and the advance
of communications and recording technology, then the phenomenon is not
real, the reports are the product of strictly memetic forces." That
behavior is uniquely predicted by the memetic hypothesis.

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



This archive was generated by hypermail 2b30 : Fri Oct 12 2001 - 14:39:48 MDT