From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Thu Jan 18 2001 - 19:41:50 MST
Samantha Atkins wrote:
>
> "Eliezer S. Yudkowsky" wrote:
> >
> > Why the shift? Well, mainly:
> >
> > 1) Instead of visualizing the Singularity as a far-off celebration on the
> > day that the whole human race gets together and networks our computing
> > power to build the seed AI that is the final culmination of all our
> > accumulated programming expertise and augmented intelligence, I figured
> > out how to build the darn thing myself on a Beowulf network.
> >
> > The other reasons, in no particular order:
>
> Really? Where is this writeen up? Where are the project development
> plans? Where is the funding and staffing plan? I don't mean to pick on
> you but I am hearing extravangant (imho) claims for one of the grandest
> projects of all time but I am not seeing or hearing about much real
> substance. Where is the beef?
I said I thought I knew how to do it. I didn't say I had written it down,
produced an architectural sketch and incremental developmental pathway,
obtained funding, hired a group of highly intelligent engineers, and
started work. It takes *time* to do all that. I'm certainly a heck of a
lot visibly closer to "first line of code" than I was a couple of years
back - compare CaTAI 2 with CaTAI 1.0; or compare "Time and Linearity" in
CaTAI 2.2 with the earlier sections.
> > 2) I figured out that human cognition actually uses all that parallelized
> > power mainly for performing operations in one neural "step" that are
> > intrinsically serial - in other words, the brain needs a massive amount of
> > power because the transistors run at 200hz. You'd need a massive amount
> > of parallel power to run a spreadsheet if it ran on a 200hz CPU. So my
> > estimate of the real computing power required to get human-equivalent
> > intelligence got stepped down by three or four orders of magnitude.
>
> Well that's an interesting theory. Please point to a write-up and to
> evidence and hopefully some peer-review.
Hey, if you think it's a good idea, write it up and send it in to your
local journal. I'm not making a career out of that argument, any more
than I made a career out of the 10^17 ops/sec @ 2025 estimate for human
equivalence. I just have the responsibility of picking what seems like
the best argument and making my best guess using it, whether peer-reviewed
or not.
> So exactly what is being done to produce the SI beyond the papers and
> such? How do people get involved if interested and possessing needed
> talents or desiring to invest?
Events occur in the following sequence:
1) We receive tax-exempt status from the IRS.
2) The AI project is put into initiable form (i.e., design moved down to
a level low enough that highly intelligent engineers can start work on
it. See previous comment about comparing "Time and Linearity" versus
"CaTAI 1.0".)
3a) We obtain funding (private foundations or individuals) to hire some
number of highly intelligent engineers;
3b) We find highly intelligent engineers who are willing to be hired;
4) Engineers are hired and work starts.
-- -- -- -- --
Eliezer S. Yudkowsky http://singinst.org/
Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:05:02 MST