I don't have any confidence that I will be able to understand Ben's paper based
on the language that he used in the abstract. If the term metagraph refers to
an abstract principle that can act on methods (or whatever) during runtime and
not just as a way of producing variations of functions (or producing graphics
which is another meaning) then it could be more adaptive . But it would also
tend to be more unfocused. Traditional programming used data types as
abstractions, but if there is an effective way to use variations of the
abstract functions for the AI program to learn to learn to adapt to new
situations then it could its range of learning could be extended. But it would
also be less focused.
------------------------------------------
Artificial General Intelligence List: AGI
Permalink:
https://agi.topicbox.com/groups/agi/Ta5ed5d0d0e4de96d-Mf08548410e3b32ad37c12789
Delivery options: https://agi.topicbox.com/groups/agi/subscription
