RE: How hard a Singularity?

From: Stephen Reed (
Date: Wed Jun 26 2002 - 19:49:13 MDT

On Wed, 26 Jun 2002, Ben Goertzel wrote:

> Government is not going to solve this problem. And I say this as someone
> with fairly democratic-socialist tendencies, not as a typical extropian
> libertarian. If government tries to manage the Singularity, it's just going
> to drive real Singularity development work underground or overseas. The
> problem has to be solved by maturity and responsibility on the part of the
> people doing the development, I feel. This is scary, but to me, it's less
> scary than thinking about the government handling something of such
> importance.... While it's true that gov't is generally good at halting
> action from occurring, by bogging it down in endless bureaucracy, gov't's
> have also been responsible for a hell of a lot of fanatically unwise
> actions -- governmental involvement is far from a prescription for wisdom!!

I suppose I am in the tiny minority who favor heavy government involvement.
Rather than try to persuade most of you why this is should be the case,
let me simply describe how I think events will unfold.

Currently I manage Cycorp's participation in a Darpa project and I
trust this process - and believe it will work for ensuring Friendly AI.

I believe that in time Darpa will sponsor research to address this
problem. As evidence mounts that the Singularity is possible - such
evidence being things such as software systems capable of deliberative
goal seeking, learning, problem solving and self-improvement, then Darpa
will act. The recent Darpa Broad Agency Announcement for a 2003 project in
Cognitive Information Processing Technology is a small step in this
direction. But in the out years 2004-2005 this project could lay the
foundation for a more ambitious follow-on.

I speculate that a future Darpa BAA will borrow heavily from CFAI.

It now occurs to me that Eliezer (or any other researcher on this
list), with goal of becoming a Darpa program manager or deputy program
manager, could begin a career path that would develop the required skills
and get the attention of Darpa management. I gave a brief presentation to
the Darpa EELD program manager yesterday and these persons enjoy wielding
the leverage they have over funded project participants. You do not have to
possess a PhD, but publications and graduate work is desirable - you need
only to have a well thought out vision, and be able to communicate that
vision to senior Darpa management - in competition with other program
managers. There is a good deal of turnover with Darpa PMs so Darpa is
allays looking.

Another idea is to get some Darpa staff on this list. (Or are you already

The European Union also sponsors AI research and it might happen that a
small EU project is the first to address what we call Friendly AI. It
would be to the benefit of those of us here who wish to accelerate the
Singularity to encourage competition between the respective government
sponsors of such research - knowing that Darpa has the highest funding


Stephen L. Reed                  phone:  512.342.4036
Cycorp, Suite 100                  fax:  512.342.4040
3721 Executive Center Drive      email:
Austin, TX 78731                   web:
         download OpenCyc at

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:39 MDT