From: Psy-Kosh (psykosh@earthlink.net)
Date: Tue Jan 07 2003 - 19:45:21 MST
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
> Do you add neurons at random, hoping that they'll arrange themselves
> usefully without costing too much in noise?
>
Well, what I've imagined, given that we get started via a human and
not an AI, would be using the best knowlage we have at the time to do
a gradual general increase (ie, adding "neurons" or equivalent as
best as possible) while keeping somewhere detailed records of my
state at various points in time, and then, once sufficiently
inteligent to truly _comprehend_ the original (or later) brain or at
least some components of it, begin redisigning those then transfer
into "brain 2.0" then gradual growth again untill possising full
comprehention of one of the 2.0 self states and redesigning again...
repeat untill sufficently advanced to have a better strategy
availible.
ie.
1. Upload
2. Store initial upload file.
3. Do the best gradual improvents possible based on best knowlage of
the brain (including the details of the scanfile.
4. When mind sufficiently advanced to truly _comprehend_ initial scan
file, consider and begin redesign.
5. Occupy new version (aka modify self based on new design, etc..)
6. Wash, rinse, repeat, transcend.
- --
Psy-Kosh
OpenPGP KeyID: 39A6DB9B71AD88B3
ICQ#47621625
"The scientist who would rather refute than comprehend demonstrates
he has chosen the wrong calling." -The Forever Machine
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1-nr1 (Windows 98) - GPGOE 0.4.1
iQCVAwUBPhuQwJzw44XQRmJ0AQLV/gP/SZ4NkGqcWCxmdO/2Z//pLV/KtPpo6YAt
tc5gH3sEXqbZ0UEv6CN8relXAW1hRrH8oxYxPC6g80CBu1v24JvCNbTCPf5oMmce
i2Y3IGJln7Git7qKxq/svCqTcM+XOIoT9bZmzqUy7p7KzPf/jij1FIzbuAowbnK8
gfa5PrltZKM=
=BR8S
-----END PGP SIGNATURE-----
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:41 MDT