Re: AI timeframes

From: J. Andrew Rogers (andrew@ceruleansystems.com)
Date: Fri Apr 09 2004 - 21:51:42 MDT


On Apr 9, 2004, at 4:24 AM, Ben Goertzel wrote:
> About secrecy... One thing you miss by being highly secretive is the
> ability to easily draw brilliant new minds into your project. AGI is
> not a problem that lends itself to a LARGE team, but it is a problem
> that lends itself to a team consisting of a number of highly
> intelligent
> and insightful people, all working together, and with different
> "angles"
> and insights.

Yeah, I meant "secretive" in the sense that any business with trade
secrets is, not in the sense of "a lone nut in a bunker". I
communicated that poorly.

A relatively small, closely-held, non-public organization is probably a
better description of what I had in mind.

> About private vs. governmental -- this really isn't a key point. The
> government, at times, is capable of funding research in a way that
> gives
> control to the scientists and doesn't weight them down with
> bureaucratic
> nonsense.

I would argue that the government is only ever an effective catalyst
when a specific line of research requires inordinate amounts capital
AND needs to be accomplished in a short time frame. The Manhattan
project is an example of this, as was early space flight. In almost
all other cases, private industry sprints ahead and one can argue to
what extent it would have happened with or without the government
trying to involve itself.

If the government decided tomorrow that they absolutely, positively had
to have fully functional AGI in the next five years, you would see
remarkable things happen in a relatively short period of time. If you
extend that to ten years, you end up with an organization that gets
strangled by bureaucratic kudzu before it even gets started. It is the
nature of the thing.

The thing is, I don't see the US government (or most others for that
matter) being sufficiently motivated in a results-oriented way to
execute a Manhattan Project type AGI endeavor. A big part of the
problem is that it is hard to define an indisputable concrete need for
the result of such a project that the government can put its weight
behind -- the use of such a thing for the government is nebulous at
best and nefarious at worst -- except for perhaps inside the larger
Department of Defense (which I doubt most people want). Manhattan
Projects don't materialize out of thin air, they are the result of the
government feeling intense external pressure.

j. andrew rogers



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:46 MDT