RE: Why SIAI is a nonprofit

From: Ben Goertzel (ben@goertzel.org)
Date: Sat Oct 23 2004 - 09:53:19 MDT


> >>80% of all new businesses fail. Even with an
> >>excellent business plan, I think Eliezer is right, the
> >>chances of success wouldn't rise above 50%. Also
> >>correct is that the business would become your life.
> >>You'd have to do nothing else for at least 5-10 years.
> >
> > Well, I am doing AGI research plus running 2 small narrow-AI-based
> > businesses. It's hard but not impossible.
>
> Duly noted, but I must point out that you have not yet *succeeded* in
> making lots of money off the narrow businesses and simultaneously
> developing a successful AI. So you haven't quite proved it's a workable
> idea, at least not yet, even leaving aside all safety issues. The two
> small businesses might be a good beginning, but
> Intelligenesis/Webmind got
> much farther than that before it failed. Even if some AI progress was
> salvageable into Novamente, that probably was not the outcome you
> originally had in mind when you founded Intelligenesis. Meaning
> no offense
> by it, just pointing out that there's farther to go before the for-profit
> strategy is validated as one that can succeed all the way to the end.

Eli, I pretty much agree with all your comments -- except with the comment
that Webmind Inc. got further than my current businesses before it failed.
Webmind Inc. had more money and staff, based on venture funding, but it was
never profitable; whereas my current businesses are small but profitable, so
I'm not sure they're really less far along than Webmind Inc. by business
standards.

I'm not arguing that what I'm doing now with Novamente is a great strategy
in general, and I'm not necessarily recommending it to others. I think it's
a reasonably decent strategy for me and Novamente right now, but that's
based on a host of particular contextual factors.

While I have not yet succeeded in getting rich nor in completing a
superhuman AGI, I *am* succeeding in paying some clever folks to work on the
Novamente AGI system via my AI-based startup businesses. That was my point,
which is valid whether or not I ever get rich or ever succeed in creating a
superhuman AGI (though I hope to do both!).

I do think that others could follow this same sort of strategy. I believe
for instance that James Rogers is doing so.

-- ben g



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