Re: Everything I know is wrong

From: Eliezer S. Yudkowsky (
Date: Wed Jan 04 2006 - 09:14:38 MST

Michael Vassar wrote:
> Those assumptions may be wrong.
> I don't even know how, as a human, to assign a probability to the
> possibility that they are wrong.
> But I do know that if they are wrong then "everything I know is wrong".

If psi is *not* part of the world, as I believe, then it may not be
possible to make psi part of the world without everything else being
wrong. Lies need not be consistent with truth, and a single
contradiction proves everything. Reality, on the other hand, is always
consistent with reality; there are many maps but only one territory; and
apples did not stop falling when Einstein's *theory* overthrew Newton's
*theory*. So if psi was real, then there would be some way to integrate
psi such that not everything else we knew would be wrong - apples would
not stop falling. It is only because psi is not real that you think
that psi would invalidate everything, and it is because psi would
invalidate everything that you think it is not real. This is, in cold
sober fact, circular logic, or more kindly a consistent worldview, and
there is the possibility of both arcs of the loop being invalidated

> From my perspective, from decision theory for humans purposes, the entire
> huge space of universe states describable as "everything I know is
> wrong" simply renormalizes out of my decision function.
> It's not worth our talking about it.

"Everything you know is wrong" has been true through quite a lot of
human history, at least in the sense of being able to encompass things
just as shocking as magic - such as the total absence of magic. In
fact, most people still haven't come to terms with the total absence of
magic, which tells you something about how shocking it is.

Eliezer S. Yudkowsky                
Research Fellow, Singularity Institute for Artificial Intelligence

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:55 MDT