On 4/19/07, Benjamin Goertzel <[EMAIL PROTECTED]> wrote:
I don't disagree with "awesome compression abilities" as a test for
"advanced AGI"
However, I think that trying to achieve awesome compression by
incrementally improving current compressors, is sorta like trying to
reach the moon by incrementally improving current pogo sticks ;-)
A different sort of architecture is needed to achieve awesome levels
of compression, than to incrementally improve on the current levels...
I've been thinking along the same lines with regard to principles of
"effective action" in the context of morality, and realized that a
mathematical description would necessarily correspond to a compression
of symbols representing increasingly complex actions. Does anyone
have pointers to any work in the optimum selection of features
representing meaningful actions in our world -- not just the motions,
but meaningful in terms of subject, object, transformation etc?
I realize this is a very open-ended question, but I'm trying to focus
at the moment on just what the feature set would look like. Maybe Cyc
would be useful in this regard.
On a related note, did anyone see any useful response to the query
about lists of inductive biases?
- Jef
-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?member_id=231415&user_secret=fabd7936
