RE: Adaptive Intelligence Inc - Incubation

From: Ben Goertzel (
Date: Tue Dec 11 2001 - 08:30:44 MST

-----Original Message-----
From: []On Behalf
Of Dmitriy Myshkin
What irked me had nothing to do with SIAI... it was the blatant
declaration of the intent to directly monetize transhuman AI.

Dmitry, the possibility is worth considering that, at the present time, the
fastest route to engineering real AI may be via the commercial world.

Let's look at the other possibilities...

Government research grant funding is in many ways more conservative than
business funding.

The Linux-style model of open-source development is, I'm afraid, far more
appropriate to the building of an OS based on well-known principles than to
the building of something as specialized and advanced as the world's first
digital mind.

The SIAI-style model of not-for-profit-but-secretive development has yet to
be validated. So far Eli is as far as I know the only one working on AI
within this model. Flare development has been offloaded, but Flare is more
Linux-like in that it's a relatively straightforward engineering project.

My intuition is that developing a real AI requires a tightly-knit, full-time
team, and right now the most likely way to pull such a team together for
real AI is within the commercial world. *Unless* someone who is already a
big shot within the (highly conservative) academic AI establishment decides
to seek grant funding for an ambitious general intelligence initiative.
This may happen, but I suspect it will only happen after some of our
maverick projects have achieved greater success.

Using business funding for AI R&D has problems that I know better than
anyone. One is faced with a constant choice between pursuing long-range R&D
and devoting all resources to solving short-term business problems.
Business funding is far from a panacea, unless one is talking about funding
within the R&D group of some very wealthy company (Microsoft, IBM, etc.).

But I don't have much patience for you or anyone else being uppity about
where Eli or Peter or I get money for Real AI development. Unless you plan
to write us a check to cover engineers' salaries and hardware costs, I
suggest you get off your high horse and respect our passion for attempting
to create real AI rather than just talking about it or working on peripheral
projects, and understand our willingness to get real AI done by hook or by

I'm not advocating a 100% "ends justifies the means" mentality. Nor am I a
fanatical right-wing libertarian who believes that everything necessarily
should be done for profit. I'm just advocating a very simple pragmatism and
open-mindedness. Nonprofit does not always mean "good"; the KKK is a
nonprofit organization. And for-profit does not always mean "bad." This
kind of simplistic human thinking, in its 100000 manifestations, is a large
part of the reason why we don't have real AI yet.

-- Ben Goertzel

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:37 MDT