Re: 10x?! (was guaranteeing Friendlyness)

From: Ben Goertzel (
Date: Mon Dec 05 2005 - 10:49:37 MST

> So if somebody let you have one of these you'd be happy?
> <>
> 96 processor deskside workstation, peak performance of 230 gigaflops,
> running off a single electric plug.
> Cost around 100,000 USD.
> But big savings on data centre costs, power, space, rackmounts, air
> conditioning, management, etc.
> BillK

This would be an excellent start, indeed.

Care to trade one to Novamente LLC in exchange for some shares in the
company? Send me a private email and we'll talk ;-)

My view is that the hardware exists for achieving AGI, and at least
one viable design exists as well (Novamente).

What is missing is the availability of funds to pay a small dedicated
team to work on AGI for the required number of years (or, alternately,
an appropriately skilled and motivated team that is sufficiently
wealthy or free of responsibilities to have no need for funding).

Society does not choose to allocate funds to this sort of work, and
therefore AGI may potentially be a long time coming, even though the
needed hardware and designs exists.

This is unfortunate in my view, but far from unprecedented in human
history. Many substantial advances have taken longer than necessary
to eventuate due to lack of appropriate vision on the part of
society's resource-holders.

On the other hand, all it would take would be one wealthy individual
or corporation with the vision to significantly fund Novamente (or
some analogous advanced AGI project, if any exist), to change the
picture substantially. Or else the advent of say 5-6 individuals with
no income requirements and a mastery of both software engineering and
AI theory, with the motivation to spend years of their lives working
uncompensated toward AGI and the Singularity.

You never know...

It is a source of great regret to me that, when I did get a fair bit
of money for AI development (in the late 90's), the AI design I had
then was half-baked and overcomplex. (Not that it was bad -- with
sufficient refinement it would have worked -- but it was so complex
that the refinement process was too time-consuming and
labor-intensive.) Now that I have a much better AGI design in hand
and a partial implementation, I find the economy is in a much more
pessimistic state and funds are much harder to come by. We are making
steady progress toward the end goal, but with the current level of
work that is being focused on the project, the progress is
frustratingly slow.

ben g

On 12/5/05, BillK <> wrote:
> On 12/5/05, Ben Goertzel wrote:
> > I'm not in the business of trying to extrapolate improvements in hardware.
> >
> > My point was that my projections about when a
> > Singularity-launch-capable Novamente could be created (given adequate
> > funding for the project) did not assume more than a 10-fold increase
> > in computing power over the next 10-15 years.
> >
> > In fact they don't strictly assume ANY increase in computing power,
> > but the 10-fold increase would be nice in order to bring the cost of
> > fulfilling the computing requirements down.
> >
> > I am almost certain that a human-level self-modifying AGI could be
> > created with a few thousand machines right now, and I think it might
> > be possible to get by with a few hundred if one paid enough attention
> > to optimization. Based on these (admittedly rough) estimates, a
> > tenfold increase in computer power would bring human-level AGI into
> > the domain of relatively easy affordability.
> >

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:54 MDT