> I don't think one needs to become as big as Microsoft or IBM to fund AGI
> research very amply, however.  I think AGI is best done by a small,
> tightly focused team, with ongoing feedback from a larger group of
> loosely affiliated scientists.  If I had enough research funding to pay
> for, say 10-15 full-time AGI scientists/engineers to work on *nothing
> but AGI*, then I'd consider this adequate to bring us to AGI.

I agree that 10-15 may be enough, but on the premise that
all of them are working on the *same model*. If you take
into consideration that we're still in dispute about what
the top-level model should be like, the best way may be to
have a number of teams exploring different models.

> Having a
> lot of people and a lot of money brings its own problems, including a
> lot of bureaucratic and managerial overhead.  I don't now have enough
> funding to pay for such an AGI R&D team, and I hope to generate this
> money through my AI business ventures.  But in the context of the US
> business world this requires a relatively modest amount of money, not
> M$-style market domination.  For example, a $1 million annual budget for
> Novamente AGI research would be more than adequate.

I'm currently thinking of a different business model: To
concentrate on AGI research and then license the technology
to software developers such as M$. In this way we can unburden
ourselves of a lot of troubles. I think this is a better biz
model and you may consider it for your own =)

> So anyway, it's not that I don't have an interest in money.  It's that I
> think I have a realistic assessment of how much money is required for a
> maximally productive AGI team.  Beyond a certain size, adding people to
> an AGI project will diminish rather than increase productivity.

My suggestion is whether we can unburden ourselves of some
irrelevant concerns such as implementation, hardware, managing
patents, or marketing. I think there'll probably be a few
things that many of us *here* are not interested in, and other
people are better suited to do the job.

> And, history shows that big companies are not that good at getting
> highly innovative research done.  Even Microsoft, with its culture of
> intelligence and productivity, hasn't really gotten that much out of its
> investment in Microsoft Research.  Useful stuff, but nothing mindblowing
> ... Whereas the world outside M$ has certainly conceived a lot of wild
> and interesting things over the same time-period that M$ Research has
> existed...

They're a company who can throw $millions on a project and
don't think twice about it. I think AGI will not happen in a
vaccuum, and it will depend on a lot of interactions.

> Finally, I see the role of an AGI Consortium as being
> 
> -- the promotion of AGI R&D
> -- the development of tools generally useful for AGI development
> -- the pursuit of AGI-oriented or closely related research not
> specifically tied to one or another AGI software system
> 
> As I said before I don't think an AGI Consortium should be thought of as
> a profit-making enterprise...

I agree with the goals you stated, though I think those
services should be paid. One thing that it may do is to
help manage IP for researchers, sometimes our results are
not big enough for filing a patent but when several results
are combined they may be.  Sharing an IP management team
may benefit all of us especially because software IP is
very complicated these days.

[ BTW thanks for your understanding, I guess I sounded very
paranoid... ]
YKY

-- 
_______________________________________________
Find what you are looking for with the Lycos Yellow Pages
http://r.lycos.com/r/yp_emailfooter/http://yellowpages.lycos.com/default.asp?SRC=lycos10

-------
To unsubscribe, change your address, or temporarily deactivate your subscription, 
please go to http://v2.listbox.com/member/[EMAIL PROTECTED]

Reply via email to