From: Dani Eder (danielravennest@yahoo.com)
Date: Thu Mar 23 2006 - 07:05:10 MST
>
> In summary, I don't think you can say that it is
> *obvious* how the above
> problems would be addressed by someone who went in
> there to teach Daniel
> Bayesian reasoning.
>
As someone who spent most of my 24 years at Boeing
working on the space station project, there is an
obvious approach to me, because I lived it for many
years.
No single person is smart enough to design and build
a space station. I feel it is likely that no single
person will be smart enough to design an AGI.
The way complicated projects get done is to break it
down into smaller pieces and give those pieces to
individuals with specialized knowledge and skills.
So for an AGI, if you want to make use of a savant,
find some piece of the project suited to their
specilized abilities and have them work it.
Note that partitioning a large project into smaller
pieces requires a non-trivial amount of overhead.
On the Space Station project it amounted to 30-40%
of the total personnel to: partition the project
into smaller segments, manage the interfaces between
the segments, optimize and balance the segments,
and test the segments individually and as a group.
The alternative is not being able to do the project
at all.
DRN
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:56 MDT