Re: Singularity Institute - update

From: Barkley Vowk (
Date: Wed Apr 30 2003 - 23:34:34 MDT

On Thu, 1 May 2003, Eliezer S. Yudkowsky wrote:

> Barkley Vowk wrote:
> >
> > Don't get me wrong, I've read much of the stuff you've
> > written, its clear to me that you are most likely not off your meds.
> Great. Go read "Levels of Organization in General Intelligence", and then
> if you're interested in criticizing any of the specific theory found
> there, feel free to do so.

I'm not criticizing your work, I'm criticizing your attitude to the
academic world. Your attitude screams crackpot. And given how hard it is
to get any serious attention in the AI field (or whichever field you
decide is the correct audience for your particular take on AI), you should
try very hard to avoid that. (and it just irritates me.)

For any research project to be science, and not pseudo-science you require
3 things:

1) a credible research group, filled or associated with properly
credential-ed researchers. (why? because without this, nobody will waste
the time to review your work without good reason, see below).

2) a credible body of work, that has been suitably reviewed (by a group
like the one described above).

3) a credible and logic-ly correct methodology for producing the above
body of work.

Why bother with all that? Because real science gets real funding and
recognition (eventual, but required for large scale funding however),
pseudo-science hangs around the Internet, waiting for someone to foolish
read it.

> Academic fields are sometimes wastelands. Nature has no a priori tendency
> to produce truths that are easy to explain to a lay audience.

I'm sure many free energy crackpots consider physics a wasteland because
their ground breaking ideas are not even mentioned. And you'd be amazed
how difficult their "truths" are to explain to anyone who does not have
their collection of logical errors to reason with.

> "Credentials" are not the proper object of a university system, learning
> is.

Credentials are a token with which one proves he/she has been taught the
skills to learn, and the background knowledge to at least start an
understanding in their field. Dismissing credentials is a common
sign of a crackpot.

> The "journals of an entire wing of academia" may easily be the wrong
> audience for any given piece of work. And it is not very hard to be more
> Bayesian than most people.

In many cases "researchers" who are self-described geniuses (or perhaps
just more X than most people), dismiss journals because their "peers" that
do the reviewing toss their papers (or use them for lining bird
cages/cleaning up coffee spills). Usually this is followed by the author
stalking legitimate researchers to get their ideas heard. This step is
generally followed by a restraining order or meds being administered.
Sometimes the above is marked with statements like "Academics in this
field are filled with anti-knowledge!", or "Those reviewers aren't smart
enough to understand my paper!", or even "There is a conspiracy against
me, my ideas are just too good, and are being suppressed".

My 2 cents.

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:42 MDT