Re: SIAI: Donate Today and Tomorrow

From: Eliezer Yudkowsky (
Date: Sun Oct 24 2004 - 08:54:17 MDT

Ben Goertzel wrote:
> A start toward this would be for Eliezer to write a systematic treatment of
> his ideas about FAI in the form of a (technical or semi-technical book), to
> be published by a typical academic press. It's easier to get eccentric
> ideas published in book form than in academic journals, because the
> refereeing process is different.

I have a suspicion that this would take a downright amazing amount of time
- on the order of four years or more.

> It may seem odd to recommend paper publication of ideas that are already
> online, but the fact is that paper publication is taken more seriously by
> the powers that be. And I think that the process of putting one's ideas
> into a more "permanent" form often forces one to tighten up one's lines of
> argument.

Isn't that what "Levels of Organization in General Intelligence" was for?
It worked, too, but it took four times as long as I allocated. And as for
"permanent" form, it goes without saying that two years later I was
sneering at the primitive naivete of my own theory.

> Furthermore, there are plenty of European universities that will give a PhD
> based on research only, no coursework. After publishing a treatise on FAI,
> it might well be possible to get a PhD on the basis of that book. I have
> some particular European university connections that might be helpful in
> this regard. I also believe I could help with the publication of the book,
> as I know a few editors at academic presses.

Hm. Sounds like a lot of conjunctive probabilities...

...still, tempting enough that I might take you up on it, but only if the
Singularity Institute completely bogged down otherwise. It's the huge
amount of time required that's the problem. I'm a slow writer, and I seem
to become a slower writer as I become a better writer.

I wouldn't *do* FAI, I'd update LOGI to incorporate Bayes and information
theory, and write the book on the evolutionary psychology of human
"significantly more generally applicable" intelligence. If I wanted
academic respect and a PhD, I'd write on a subject that I already fully
understood, people were already interested in, and that was at least
theoretically possible to explain.

But oh, the time required! Gobs and endless gobs of time! I shudder just
to think of it. And *it might still not work*, and I'd *still* have to do
all the FAI theory and I wouldn't be as young when I did. All my life has
taught me the value of not being distracted. That's why I account it a
last resort.

I wish I wasn't such a slow writer. It would open up more options. But
that's not something I've figured out how to change.

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

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:49 MDT